Control Engineering in Robotics and Industrial Automation: Malaysian Society for Automatic Control Engineers (MACE) Technical Series 2018 (Studies in Systems, Decision and Control, 371) [1st ed. 2022] 3030745392, 9783030745394

Table of contents :
Preface
Contents
Notation for MACE Technical Series 2018
Design and Control of a 3D Robot-Assisted Rehabilitation Device for Post-Stroke
Wet Scrubber Design
Development of Intelligent Controller for Pollution Monitoring and Control
Optimal Tuning of Fractional-Order PID Controller for Electric Power-Assisted Steering (EPAS) System Using Particle Swarm Optimization (PSO)
Forward Navigation for Autonomous Unmanned Vehicle in Inter-Row Planted Agriculture Field
Hardware Design and Development of Contactless Sensor System for Piano Playing
The Validation of Virtual Impact Tests Using LabVIEW Instrumentation Techniques
Instrumentation in Underwater Environment
BER Performance Evaluation of M-PSK and M-QAM Schemes in AWGN, Rayleigh and Rician Fading Channels
Camera Calibration and Video Stabilization Framework for Robot Localization
Introduction
Design and Control of a 3D Robot-Assisted Rehabilitation Device for Post-Stroke
1 Introduction
2 Background
3 System Description
3.1 System Modeling—Forward Kinematics
3.2 System Modeling—Inverse Kinematics
3.3 System Modeling—Velocity Kinematics: The Jacobian
3.4 System Modeling—the Robot Dynamics
3.5 Control Architecture—Sensors and Actuation System
3.6 Control Architecture—Control Hardware
3.7 Control Architecture—Feedback Control Linearization Strategy
3.8 Control Architecture—Adaptive Control
4 Results and Discussion
4.1 Trajectory Generation and Position Control
4.2 Force Tracking
5 Conclusion
References
Wet Scrubber Design
1 Introduction
2 Wet Scrubber Design
2.1 Sizing the Scrubber System
2.2 Determining of the Scrubber Wall Thickness
2.3 Determination of Quantity of Water Used in Scrubbing
2.4 Number of Nozzles in the Scrubber System
2.5 Pipe Network Design
2.6 Determination of Duct Diameter
2.7 Hood Design
2.8 Head Losses Within the Pipe Network
2.9 Rate of Energy Gained by the Scrubbing Liquid
2.10 Mechanical Power Delivered to the Pump
2.11 Temperature Rise of the Scrubbing Liquid
3 Hydraulic Similitude Design of the Scrubber System
3.1 Similitude Model and Scaling
4 Optimization of the Design Using Computational Fluid Dynamics
4.1 Gas and PM Model Description
4.2 CFD Pre-processing Stage
4.3 CFD SetUp and Solution Stage
4.4 CFD Post-processing Stage
4.5 Results Analysis
5 Conclusion
References
Development of Intelligent Controller for Pollution Monitoring and Control
1 Introduction
2 Choice of DSP Chip
2.1 Architecture
2.2 Arithmetic Format
2.3 Speed
2.4 Memory
2.5 Data Width
2.6 Power Consumption and Management
2.7 Cost
3 Features of the Selected DSP Chip
4 System Description and Intelligent Controller Development
4.1 Mathematical Model of Wet Scrubber System
4.2 Intelligent Controllers
4.3 Neuro Fuzzy
4.4 ANFIS Controller Development
5 Simulation Results
5.1 Fuzzy Logic Controller
5.2 ANFIS Controller
5.3 Real-Time Implementation of the Proposed System
6 Experimental Result and Discussion
7 Conclusion
References
Swarm Robotics Behaviors and Tasks: A Technical Review
1 Introduction
2 Review Outlines
3 Hardware and Software Platform
3.1 Hardware Platform
3.2 Software Platform
4 Low-Level Task
4.1 Aggregation
4.2 Dispersion
4.3 Self-reconfigurable and Self-assembly
4.4 Pattern Formation and Flocking
4.5 Robot–Environment Interaction
4.6 Task Allocation
4.7 Robot Learning
5 High-Level Tasks
5.1 Collective Searching and Localization
5.2 Collective Mapping
5.3 Collective Foraging
5.4 Collective Transport
5.5 Collective Manipulation
5.6 Collective Tracking
6 Challenges and the Way Forward
6.1 Challenges
6.2 The Way Forward
7 Conclusion
References
Optimal Tuning of Fractional-Order PID Controller for Electric Power-Assisted Steering (EPAS) System Using Particle Swarm Optimization (PSO)
1 Introduction
1.1 System Modeling
2 EPAS Controller
3 Assist Characteristic Curves
4 Fractional-Order PID Controllers (FOPID)
5 PSO Algorithm
6 Simulation Results
7 Controlled System in Different Speeds and Different Driver Torques
8 Conclusion
References
Forward Navigation for Autonomous Unmanned Vehicle in Inter-Row Planted Agriculture Field
1 Introduction
2 Conceptual Framework
3 Methodology
3.1 Bezier Curve-Based Path Planning
3.2 Optimized Bezier Curve-Based Trajectory Planning
4 Result and Discussion
5 Conclusion
References
Hardware Design and Development of Contactless Sensor System for Piano Playing
1 Introduction
2 Methods for Contactless Capacitive Tracking
3 Electrode and Experimental Design
4 RC Oscillator
5 Electrode Design
6 Conclusion
References
The Validation of Virtual Impact Tests Using LabVIEW Instrumentation Techniques
1 Introduction
2 Conventional Analytical Methods
2.1 Hertz Contact Law
2.2 Levy Solution
3 Data Acquisition System and LabVIEW Instrumentation Techniques
3.1 Elements of a Data Acquisition System
3.2 Piezoelectric Accelerometer
3.3 Signal Conditioning
3.4 National Instrumentation—DAQ Hardware
3.5 LabVIEW Instrumentation Techniques
4 Hammer Drop Test
5 Virtual Impact Test
6 The Impact Responses of the Steel Plates
6.1 Contact Force
6.2 Hammer Impaction
6.3 Dynamic Displacement of Steel Plates Due to Impact Force
7 Conclusions
References
Instrumentation in Underwater Environment
1 Introduction
2 Noise Measurement
2.1 Introduction
2.2 Instrumentation and Methodology for Measurement of Noise
2.3 Units for Measuring Noise
2.4 Parameters for Estimating Noise
3 Sound Measurement
3.1 Introduction
3.2 Standardization of Underwater Acoustic Terminology and Measurements
3.3 Measuring the Underwater Radiated Sound of Dredgers
4 Distance and Direction Measurements
4.1 Introduction
4.2 Basic Principle
4.3 Position and Direction Measurement Calculation
5 Fish Population Estimation
5.1 Introduction
5.2 Underwater Video Measurement
5.3 Fish Tagging and Marking Techniques
5.4 Depletion Estimates
5.5 Underwater Sonar and Laser Measuring
5.6 Comparing Laser and Sonar Systems
6 Mine Detection
6.1 Relevant Parameters
6.2 Stakeholder Analysis
6.3 Detection of Mine
6.4 Mine Detection Technique
7 Using Symbolic Pattern Analysis
8 Conclusion
References
BER Performance Evaluation of M-PSK and M-QAM Schemes in AWGN, Rayleigh and Rician Fading Channels
1 Introduction
2 Proposed Method
3 Results and Simulation
4 Conclusion
References
Camera Calibration and Video Stabilization Framework for Robot Localization
1 Introduction
2 Overview
2.1 Laser Range Finder
2.2 Sensor Network
2.3 Vision Localization Based on Images from Camera
3 Dataset Structure for Robot Localization
3.1 Dataset for Evaluations and Comparison from Humanoid Robot and Goal
3.2 Evaluation of Robot Localization
4 Proposed Framework for Humanoid Robot Localization
4.1 Proposed Experimental Setup for Framework for Humanoid Robot Localization
5 Results Localization Following by Stereo and Mono Vision Method Camera
5.1 Compare Popular Method in Computing Distance with Proposed FCC
6 Summary
6.1 Discussion
References

Citation preview

Studies in Systems, Decision and Control 371

Muralindran Mariappan Mohd Rizal Arshad Rini Akmeliawati Chong Shin Chong   Editors

Control Engineering in Robotics and Industrial Automation Malaysian Society for Automatic Control Engineers (MACE) Technical Series 2018

Studies in Systems, Decision and Control Volume 371

Series Editor Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland

The series “Studies in Systems, Decision and Control” (SSDC) covers both new developments and advances, as well as the state of the art, in the various areas of broadly perceived systems, decision making and control–quickly, up to date and with a high quality. The intent is to cover the theory, applications, and perspectives on the state of the art and future developments relevant to systems, decision making, control, complex processes and related areas, as embedded in the fields of engineering, computer science, physics, economics, social and life sciences, as well as the paradigms and methodologies behind them. The series contains monographs, textbooks, lecture notes and edited volumes in systems, decision making and control spanning the areas of Cyber-Physical Systems, Autonomous Systems, Sensor Networks, Control Systems, Energy Systems, Automotive Systems, Biological Systems, Vehicular Networking and Connected Vehicles, Aerospace Systems, Automation, Manufacturing, Smart Grids, Nonlinear Systems, Power Systems, Robotics, Social Systems, Economic Systems and other. Of particular value to both the contributors and the readership are the short publication timeframe and the world-wide distribution and exposure which enable both a wide and rapid dissemination of research output. Indexed by SCOPUS, DBLP, WTI Frankfurt eG, zbMATH, SCImago. All books published in the series are submitted for consideration in Web of Science.

More information about this series at http://www.springer.com/series/13304

Muralindran Mariappan · Mohd Rizal Arshad · Rini Akmeliawati · Chong Shin Chong Editors

Control Engineering in Robotics and Industrial Automation Malaysian Society for Automatic Control Engineers (MACE) Technical Series 2018

Editors Muralindran Mariappan Faculty of Engineering Universiti of Malaysia Sabah Kota Kinabalu, Malaysia Rini Akmeliawati School of Mechanical Engineering The University of Adelaide Adelaide, Australia

Mohd Rizal Arshad Faculty of Electrical Engineering Technology Universiti Malaysia Perlis Perlis, Malaysia Chong Shin Chong Faculty of Electrical Engineering Universiti Teknikal Malaysia Melaka Durian Tunggal, Malaysia

ISSN 2198-4182 ISSN 2198-4190 (electronic) Studies in Systems, Decision and Control ISBN 978-3-030-74539-4 ISBN 978-3-030-74540-0 (eBook) https://doi.org/10.1007/978-3-030-74540-0 © Springer Nature Switzerland AG 2022 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

One of the main objectives of the Malaysian Society for Automatic Control Engineers (MACE) is to promote the science and technology of automatic control engineering in the broadest sense in all systems whether, for example, engineering, physical, biological, social, or economic, in both theory and application. For that, MACE organizes and sponsors technical meetings such as congresses, conferences, symposia, and workshops. And, this book is about the collection of all the technical series presented in the last 2 years. Many books have been written on control engineering, describing new methods for controlling systems and better ways of mathematically formulating existing techniques to solve the ever-increasing complex problems faced by practicing engineers. However, few books fully address the application aspects of control engineering. It is the intention of this new series to redress this situation. This series will stress application issues, and not just the mathematics of control engineering. It will provide texts that present not only both new and well-established techniques but also detailed examples of the applications of these methods to the solution of real-world problems. There are many exciting examples of the applications of control techniques in the established fields of electrical, mechanical, and biomedical engineering that discussed in this book. Among the applications, it includes the applications in the brake-by-wire system, assisted rehabilitation device for post-stroke patient, pollution monitoring and control, agriculture field, piano, virtual impact tests, underwater, and camera system. This series presents books that draw on expertise from both the academic world and the applications domains and will be useful not only as academically recommended course texts but also as handbooks for practitioners in many applications domains. Control Engineering in Robotics and Industrial Automation—Malaysian

v

vi

Preface

Society for Automatic Control Engineers (MACE) Technical Series 2018 is another outstanding entry in Springer Technical Series. Kota Kinabalu, Malaysia Perlis, Malaysia Adelaide, Australia Durian Tunggal, Malaysia

Muralindran Mariappan Mohd Rizal Arshad Rini Akmeliawati Chong Shin Chong

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rini Akmeliawati Design and Control of a 3D Robot-Assisted Rehabilitation Device for Post-Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shahrul Na’im Sidek and Sado Fatai 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 System Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 System Modeling—Forward Kinematics . . . . . . . . . . . . . . . . . . . . . . 3.2 System Modeling—Inverse Kinematics . . . . . . . . . . . . . . . . . . . . . . . 3.3 System Modeling—Velocity Kinematics: The Jacobian . . . . . . . . . 3.4 System Modeling—the Robot Dynamics . . . . . . . . . . . . . . . . . . . . . . 3.5 Control Architecture—Sensors and Actuation System . . . . . . . . . . 3.6 Control Architecture—Control Hardware . . . . . . . . . . . . . . . . . . . . . 3.7 Control Architecture—Feedback Control Linearization Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.8 Control Architecture—Adaptive Control . . . . . . . . . . . . . . . . . . . . . . 4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Trajectory Generation and Position Control . . . . . . . . . . . . . . . . . . . 4.2 Force Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wet Scrubber Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. A. Danzomo, Sambo A. Umar, and Momoh Jimoh E. Salami 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Wet Scrubber Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Sizing the Scrubber System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Determining of the Scrubber Wall Thickness . . . . . . . . . . . . . . . . . . 2.3 Determination of Quantity of Water Used in Scrubbing . . . . . . . . . 2.4 Number of Nozzles in the Scrubber System . . . . . . . . . . . . . . . . . . . 2.5 Pipe Network Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

5 5 6 7 7 8 9 10 10 11 12 13 14 15 16 17 17 19 20 22 22 24 25 25 26 vii

viii

Contents

2.5.1 Diameter of the Supply Pipe . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5.2 Spray Pipe Diameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Determination of Duct Diameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7 Hood Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8 Head Losses Within the Pipe Network . . . . . . . . . . . . . . . . . . . . . . . . 2.8.1 Supply Line and Spray Line Head Losses . . . . . . . . . . . . . . . 2.8.2 Losses Due to Sudden Contraction . . . . . . . . . . . . . . . . . . . . 2.8.3 Losses at the Nozzles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8.4 The Overall Head Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.9 Rate of Energy Gained by the Scrubbing Liquid . . . . . . . . . . . . . . . 2.10 Mechanical Power Delivered to the Pump . . . . . . . . . . . . . . . . . . . . . 2.11 Temperature Rise of the Scrubbing Liquid . . . . . . . . . . . . . . . . . . . . 3 Hydraulic Similitude Design of the Scrubber System . . . . . . . . . . . . . . . . . 3.1 Similitude Model and Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1 Geometric Similitude Scaling . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.2 Kinematic Similitude Scaling . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.3 Dynamic Similitude Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Optimization of the Design Using Computational Fluid Dynamics . . . . . . 4.1 Gas and PM Model Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.1 Gas-Phase Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.2 Particle Phase Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 CFD Pre-processing Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 CFD SetUp and Solution Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 CFD Post-processing Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1 Graphical Display of the Gas Flow Velocities . . . . . . . . . . . 4.4.2 Graphical Display of Pressure and Density Due to Gas Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Results Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.1 Analysis of the Velocity Contours . . . . . . . . . . . . . . . . . . . . . 4.5.2 Analysis of Velocity Profile . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.3 Analysis of the Drag Force for the Gas and Gravity Force for the Particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.4 Analysis of the Pressure Drop Across the Scrubbing Chamber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.5 Analysis of the Total Pressure . . . . . . . . . . . . . . . . . . . . . . . . . 5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Development of Intelligent Controller for Pollution Monitoring and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sambo A. Umar and Momoh Jimoh E. Salami 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Choice of DSP Chip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Arithmetic Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

26 26 27 28 29 29 31 31 32 33 33 34 34 35 35 36 36 38 40 40 41 43 46 47 48 49 52 54 56 58 59 60 61 61 63 64 66 68 69

Contents

ix

2.3 Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Data Width . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Power Consumption and Management . . . . . . . . . . . . . . . . . . . . . . . . 2.7 Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Features of the Selected DSP Chip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 System Description and Intelligent Controller Development . . . . . . . . . . . 4.1 Mathematical Model of Wet Scrubber System . . . . . . . . . . . . . . . . . 4.2 Intelligent Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1 Fuzzy Logic Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Neuro Fuzzy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 ANFIS Controller Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Fuzzy Logic Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 ANFIS Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Real-Time Implementation of the Proposed System . . . . . . . . . . . . 6 Experimental Result and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69 70 70 70 70 71 71 71 75 75 79 82 83 83 83 85 91 94 95

Swarm Robotics Behaviors and Tasks: A Technical Review . . . . . . . . . . . . M. H. A. Majid, M. R. Arshad, and R. M. Mokhtar 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Review Outlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Hardware and Software Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Hardware Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Software Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Low-Level Task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Aggregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.1 Definition and Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.2 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.3 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Dispersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1 Definition and Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.3 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Self-reconfigurable and Self-assembly . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Definition and Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Pattern Formation and Flocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1 Definition and Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.2 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.3 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Robot–Environment Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

99 100 104 105 105 106 106 109 110 110 111 115 115 115 116 117 118 118 119 121 121 122 123 128

x

Contents

4.5.1 Definition and Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.2 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.3 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Task Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6.1 Definition and Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6.2 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6.3 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7 Robot Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7.1 Definition and Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7.2 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7.3 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 High-Level Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Collective Searching and Localization . . . . . . . . . . . . . . . . . . . . . . . . 5.1.1 Related Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Collective Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 Related Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.2 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Collective Foraging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 Related Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Collective Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1 Related Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.2 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 Collective Manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.1 Related Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.2 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 Collective Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.1 Related Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.2 Methods and Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Challenges and the Way Forward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2 Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.3 Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.4 Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.5 Actuators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.6 Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 The Way Forward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

128 129 129 132 133 133 134 135 135 136 136 137 137 138 138 142 142 142 143 144 144 145 146 146 146 147 147 147 148 148 148 148 149 149 150 150 150 151 151 152 153

Contents

Optimal Tuning of Fractional-Order PID Controller for Electric Power-Assisted Steering (EPAS) System Using Particle Swarm Optimization (PSO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mohd Khair Hassan, Adel Amiri, Hamiruce Marhaban, and Asnor Juraiza 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 System Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 EPAS Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Assist Characteristic Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Fractional-Order PID Controllers (FOPID) . . . . . . . . . . . . . . . . . . . . . . . . . . 5 PSO Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Controlled System in Different Speeds and Different Driver Torques . . . . 8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Forward Navigation for Autonomous Unmanned Vehicle in Inter-Row Planted Agriculture Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Norashikin M. Thamrin, Nor Hashim Mohd. Arshad, Ramli Adnan, and Rosidah Sam 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Conceptual Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Bezier Curve-Based Path Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Optimized Bezier Curve-Based Trajectory Planning . . . . . . . . . . . . 4 Result and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hardware Design and Development of Contactless Sensor System for Piano Playing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Choo Chee Wee and Muralindran Mariappan 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Methods for Contactless Capacitive Tracking . . . . . . . . . . . . . . . . . . . . . . . . 3 Electrode and Experimental Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 RC Oscillator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Electrode Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Validation of Virtual Impact Tests Using LabVIEW Instrumentation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chee-Siang Chong, Nittala Surya Venkata Kameswara Rao, Muralindran Mariappan, and Wei Leong Khong 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Conventional Analytical Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Hertz Contact Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

169 169 171 172 173 173 174 176 178 181 181 183

184 185 188 188 190 194 197 197 199 200 202 203 203 205 207 207 209

209 212 212

xii

Contents

2.2 Levy Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Data Acquisition System and LabVIEW Instrumentation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Elements of a Data Acquisition System . . . . . . . . . . . . . . . . . . . . . . . 3.2 Piezoelectric Accelerometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Signal Conditioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 National Instrumentation—DAQ Hardware . . . . . . . . . . . . . . . . . . . 3.5 LabVIEW Instrumentation Techniques . . . . . . . . . . . . . . . . . . . . . . . 4 Hammer Drop Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Virtual Impact Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 The Impact Responses of the Steel Plates . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Contact Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Hammer Impaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Dynamic Displacement of Steel Plates Due to Impact Force . . . . . 7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Instrumentation in Underwater Environment . . . . . . . . . . . . . . . . . . . . . . . . M. M. Rashid and Raju Ahamed 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Noise Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Instrumentation and Methodology for Measurement of Noise . . . . 2.3 Units for Measuring Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Parameters for Estimating Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Sound Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Standardization of Underwater Acoustic Terminology and Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Measuring the Underwater Radiated Sound of Dredgers . . . . . . . . . 4 Distance and Direction Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Basic Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Position and Direction Measurement Calculation . . . . . . . . . . . . . . . 5 Fish Population Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Underwater Video Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Fish Tagging and Marking Techniques . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Depletion Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 Underwater Sonar and Laser Measuring . . . . . . . . . . . . . . . . . . . . . . 5.6 Comparing Laser and Sonar Systems . . . . . . . . . . . . . . . . . . . . . . . . . 6 Mine Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Relevant Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Stakeholder Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Detection of Mine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

213 215 215 216 216 218 220 222 223 225 225 229 230 230 236 239 239 240 240 241 241 241 243 243 243 244 244 244 245 246 246 246 247 247 248 249 250 250 251 252 252

Contents

6.4

Mine Detection Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.1 Wavelet Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.2 Using Neural Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.3 Using Augmented Reality . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Using Symbolic Pattern Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . BER Performance Evaluation of M-PSK and M-QAM Schemes in AWGN, Rayleigh and Rician Fading Channels . . . . . . . . . . . . . . . . . . . . . Ali Farzamnia, Muralindran Mariappan, Ervin Moung, and Ramanan Thangasalvam 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Proposed Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Results and Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Camera Calibration and Video Stabilization Framework for Robot Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Farshid Pirahansiah, Shahnorbanun Sahran, and Siti Norul Huda Sheikh Abdullah 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Laser Range Finder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Sensor Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Vision Localization Based on Images from Camera . . . . . . . . . . . . . 3 Dataset Structure for Robot Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Dataset for Evaluations and Comparison from Humanoid Robot and Goal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Evaluation of Robot Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Proposed Framework for Humanoid Robot Localization . . . . . . . . . . . . . . 4.1 Proposed Experimental Setup for Framework for Humanoid Robot Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Results Localization Following by Stereo and Mono Vision Method Camera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Compare Popular Method in Computing Distance with Proposed FCC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xiii

252 252 254 254 254 255 255 257

257 258 260 265 266 267

268 268 269 270 270 274 274 276 276 277 279 281 282 284 285

Notation for MACE Technical Series 2018

Design and Control of a 3D Robot-Assisted Rehabilitation Device for Post-Stroke a α d θ T s c x3 y3 z3 x0 y0 z0 ui q v q˙ J J (q) k τ L K P Mr Iact C(q, q) ˙ G(q)

Link length Link twist Link offset Joint angle Homogeneous transformation matrix Sine of the angle Cosine of the angle End-effector frame in x-direction End-effector frame in y-direction End-effector frame in z-direction Base frame in x-direction Base frame in y-direction Base frame in z-direction Tangent of any half angle Joint variables/3x1 vector of generalized joint coordinates Linear velocity Angular velocity Jacobian Jacobian matrix First derivative of the forward kinematics equation 3x1 vector of generalized input actuator forces/torque Lagrangian Kinematic energy of the robotic system Potential energy of the system Robot mass matrix Actuator inertia Coriolis and centrifugal terms Gravity term xv

xvi

τe fe M(q) Mm Bm Km S fd xd

Notation for MACE Technical Series 2018

Vector of environment reaction torque Task space contact force Sum of the total robot mass matrix and the actuators inertia Robot virtual impedance parameter of mass Robot virtual impedance parameter of damping Robot virtual impedance parameter of stiffness factors Selector matrix for force and position control Reference force Reference position

Wet Scrubber Design ρ A U1 U2 Ug m D Z Pe L K1 K2 E t Q q v a N Ucrit dsup Q spray dspray dduct Qh f H Re hD f ε kC

Density Cross sectional area Inlet velocity Outlet velocity Maximum gas velocity Mass flow rate of the exhaust gas Cylindrical spay chamber diameter Height of the scrubber system Collapsing pressure Length Numerical coefficients depending on the length to diameter ratio Numerical coefficients depending on the diameter to thickness ratio Modulus of elasticity Thickness Quantity of liquid/discharge Quantity of discharge of each droplet Nozzle critical velocity Nozzle orifice Number of nozzles Critical velocity Diameter of the supply pipe Quantity of liquid of the spray pipe Diameter of the spray pipe Duct diameter Volume of gas flowing through the frustum in one second Height of the hood Reynolds number Head loss due to friction in turbulent flow Darcy friction factor Roughness of a pipe/turbulent kinetic energy dissipation rate Coefficient of contraction

Notation for MACE Technical Series 2018

θ kL N hT L N E Ppump η pump Pelectric ηmotor Eloss T λ λL λT λU λa λQ λF λρ T Tp Tm a Fr Fm Fp EU i j k μt δk δε c1 c2 φj Uj Ddi f f Fg FD Stk τj τf P Psystem Pgauge

xvii

Cone angle Frictional loss coefficient Total losses due to nozzles Rate of energy gain Mechanical power delivered to the pump Efficiency of the pump Electric power of the motor Efficiency of the electric motor Lost mechanical energy Temperature rise Scale factor Length scale factor for the scrubber diameter, height and thickness Time scale factor Velocity scale factor Acceleration scale factor Discharge scale factor Force scale factor Density scale factor Ratio of time Ratio of time required for homologous particles to travel in a prototype Ratio of time required for homologous particles to travel in a model Acceleration Scaled ratio of homologous forces Scaled ratio of homologous forces in the model Scaled ratio of homologous forces in the prototype Euler number Direction vector Direction vector Turbulent kinetic energy Viscosity coefficient of turbulent flow Constant of 1.0 Constant of 1.3 Constant of 1.44 Constant of 1.92 Concentration of chemical specie and j is the particle matter particle size Particle matter velocity Diffusivity of a scalar in the fluid Gravitational force Drag forces Particle stokes number Relaxation time Relevant fluid time scale Pressure drop System pressure Gauge pressure

xviii

Patm

Notation for MACE Technical Series 2018

Atmospheric pressure

Development of Intelligent Controller for Pollution Monitoring and Control y p ( j, 0) Q L /Q G vr vg z dD ηsep ψ Cc f μg vtd ρp ρD ρg Kn λ Tg P Mg R g CD Re vtd N M L Vcal Vo

Particulate matter concentration at the scrubber inlet Liquid to gas ratio Relative velocity between liquid droplets and dust particles Gas velocity Height of the scrubber Droplet diameter Gas-particle separation efficiency Impaction parameter Cunningham slip-correction factor Gas viscosity Terminal settling velocity of the liquid droplets Particle density Particle diameter Gas density Knudsen number Gas mean free path Gas temperature Atmospheric pressure Molecular weight of the gas Universal gas constant Gravitational acceleration Drag coefficient Reynolds number Terminal settling velocity Number of membership functions for inputs and outputs Number of membership functions for inputs and outputs Number of membership functions for inputs and outputs Calculated voltage Sensor analog voltage

Optimal Tuning of Fractional-Order PID Controller for Electric Power-Assisted Steering (EPAS) System Using Particle Swarm Optimization (PSO) Jc

Moment of inertia of the column

Notation for MACE Technical Series 2018

Bc Ts θc Td Kc Xr Rs Tm Jm Bm Ka R L Kb U ia θm Mr Br Rs Kr K (v) Imax C(S) Kp Ki Kd λ μ vi w xi rand C1 C2 Pbest G best W Wmax Wmin iter itermax Ts Tr Mp

Column damping Torque sensor Rotation angle of the driving wheel Torque of driving wheel Steering column stiffness Rack horizontal displacement Radius of pinion steering Motor output Motor inertia Motor damping Motor torque Resistance Inductance Anti-EMF coefficient Motor voltage Motor current Angle of motor Steering tie rod mass Steering tie rod damping coefficient Radius of pinion steering Tire spring rate Ratio coefficient Maximum current Transfer function of the controller Proportional gain Integral time constant Derivative time constant Fractional power Fractional power Velocity of particle i Inertia weight factor Particle position Random function Acceleration constant Acceleration constant Best position of the ith particle Best position among all particles in the swarm Weight function Initial weight Final weight Current iteration time Maximum number of iterations Setting time Rise time Overshoot

xix

xx

Notation for MACE Technical Series 2018

ess I SE I T SE I AE a

Steady-state error Integral of squared-error Time-weighted-squared-error Integrated absolute error Weighting factor

Forward Navigation for Autonomous Unmanned Vehicle in Inter-Row Planted Agriculture Field Bi C Pn x y H V δt

Bernstein function Control vector Point at x-plane Point at y-plane Distance Velocity Total elapsed time

Hardware Design and Development of Contactless Sensor System for Piano Playing C ε A d

Capacitance Dielectric constant of the material between the two conducting plates Area of both conducting plates Distance between two conducting plates

The Validation of Virtual Impact Tests Using LabVIEW Instrumentation Techniques P(t) Pmax R t τc K V1 e m1 v1

Impact force respect to time Maximum force Radii curvature Time after the impaction Contact period Constant regarding the material and dimension properties of the surfaces Initial impact velocity of the hammer Coefficient of restitution Mass of the hammer Poisson’s ratio of the hammer surface

Notation for MACE Technical Series 2018

v2 E∗ E1 E2 w(x, y) q(t) x y

xxi

Poisson’s ratio of the plate surface Relative elastic moduli of the contacting surfaces Elastic moduli of the hammer surface Elastic moduli of the plate surface Transverse displacement of the plate Temporal equation that deals with frequency of the natural vibration Impacted point X Impacted point Y

Instrumentation in Underwater Environment SPL SL N R a

Sound pressure level Source level Transmission loss Distance from the source Coefficient of absorption

BER Performance Evaluation of M-PSK and M-QAM Schemes in AWGN, Rayleigh and Rician Fading Channels Q K Eb No P N M

Q-function Rician factor Ratio of energy per bit Spectral noise density Equation of bit error rate for binary phase-shift keying Rician fading channel diversity M value of M-ary Quadrature Amplitude Modulation (M-QAM) modulation scheme in additive white Gaussian Noise

Camera Calibration and Video Stabilization Framework for Robot Localization f b z ε

Focal length Baseline Distance computed Error

Introduction Rini Akmeliawati

Abstract The series of research collections, CONTROL ENGINEERING IN ROBOTICS AND INDUSTRIAL AUTOMATION, presents research outcomes in the fields of, but not limited to, control engineering, mechatronics, robotics and automation. The book’s eleven chapters demonstrate the current state-of-the-art technology in the aforementioned fields. Research work on a rehabilitation robotic device, web-scrubber systems, swarm robots, an electric power-assisted steering system (EPAS), a navigation system for an unmanned vehicle as part of agricultural technology, a sensor-based piano-playing analyzer, a steel-plate virtual impact tester, a review work on underwater instrumentations, an efficient wireless communication system for sensor data and vision-based robot localization are presented in this book. These diverse contribution will be beneficial to researchers, industrialists and whoever else is interested in the topics.

Control technology is an engineering field that involves multiple disciplines in its development and applications; however, it is often ‘hidden’ within various other branches of technology. It is in fact a branch of the engineering field that becomes a key enabler to several technological applications and is considered the ‘brain’ of any mechatronics systems. Control technology often comes together with sensor and measurement technology, as we will see in this volume. In this first series of research collections, various applications of control engineering, sensors, and instrumentation technology in robotics and industrial automation, and other mechatronics, navigation and communication systems, are presented. Control applications for an end-effector robotic device for upper-extremity rehabilitation is presented by Sidek and Fatai. Such a robotic device is aimed to help hemiparetic post-stroke patients. It offers three-degrees-of-freedom motions for elbow and shoulder exercise. The controller is designed based on an adaptive hybrid impedance R. Akmeliawati (B) School of Mechanical Engineering, The University of Adelaide, Adelaide, SA 5005, Australia e-mail: [email protected] © Springer Nature Switzerland AG 2022 M. Mariappan et al. (eds.), Control Engineering in Robotics and Industrial Automation, Studies in Systems, Decision and Control 371, https://doi.org/10.1007/978-3-030-74540-0_1

1

2

R. Akmeliawati

framework for safe robot–patient dynamic interactions during repetitive exercises and monitors the patients’ motor recovery. An intelligent control system finds an interesting application in web scrubber systems, as presented by Umar and Salami. The controller is used to adjust the flow rate of the liquid pump to deliver the appropriate droplet size for scrubbing the Particulate Matter (PM) contaminants based on their detected concentration level. As a result of the implementation of the designed intelligent controller, the web scrubber system can maintain the emission size below the recommended value in less than 10seconds. The details of the designed web scrubber system is provided in the preceding chapter by Danzomo et al. Majid et al. present an interesting technical review on swarm robotics’ behaviors and tasks. The review provides an overall overview of research in swarm robotics tasks, which are classified into two types: low-level and high-level tasks. The chapter identifies and shows the correlation between the two types of tasks. The review also includes the software and hardware platforms used for simulation and experimentation in swarm robotics research. Still related to swarm behaviors, Hassan et al. show the application of a particle swarm optimization algorithm as a tuning mechanism for Fractional-Order PID (FOPID) controllers for the electric motor of an Electric Power-Assisted Steering system (EPAS). The performance of the overall system is evaluated under different speeds and driver torques. The optimal FOPID with PSO has demonstrated a more efficient performance than a classical PID controller to the system. An autonomous navigation system of a small-scaled unmanned vehicle for agriculture applications has been proposed by Thamrin et al. Forward and headland turn navigation based on the Bezier curve provides optimized trajectory planning for the autonomous vehicle to perform narrow inter-tree navigation in an agricultural field. This is essential for precision agriculture systems. The application of sensor technology is not only limited to the engineering field but can also be found in music. Choo and Mariappan propose the design of a contactless sensor system to study the finger positions of musicians while playing the piano. A non-intrusive and long-range capacitive sensor is developed and placed under the keyboard area to sense the position of the player’s fingers on five piano keys. An artificial neural network is used to process the data such that the pianists can store and analyze their playing techniques. The application of sensor and instrumentation technology is also found in impact testing, as described by Chong et al., who propose the LabVIEW-based instrumentation to validate virtual impact tests of steel plates with various boundary conditions. The piezoelectric accelerometer is used as the primary sensor in the instrumentation, providing excellent means of dynamics measurements of the impact incidents. The LabVIEW-based instrumentation was used for hammer drop tests to investigate the impact response of the steel plate with different hammer heights and various boundary conditions. Rashid and Ahmad present a review on state-of-the-art underwater instrumentation to measure environmental underwater noise. Underwater measurements are affected by various parameters, such as distance, fish population, sound, and other

Introduction

3

uncertainties due to oceanic activities. Advances in sensors and measurement techniques are improving the accuracy and quality of the measurement data under such conditions. The review describes the construction, operation and performance of various underwater instrumentation and measurement techniques for specific applications, such as noise, sound, distance and direction measurements, fish population estimation, and mine detection. Controllers and their sensors are linked through communication systems. Efficient wireless communication systems are essential in transmitting information, such as sensor data. Minimum errors can be achieved with proper channel conditions and modulation techniques. Farzamnia et al. investigate and compare the Bit Error Rate (BER) performance of two modulation techniques which are used in communications systems, the M-ary Phase Shift Keying (M-PSK) and M-ary Quadrature Amplitude Modulation (M-QAM), under the Additive White Gaussian Noise (AWGN), Rayleigh and Rician fading channels. Another application of control and sensor technology can be found in robot localization problems. In the last chapter of this volume, Sahran et al. investigate two major issues in vision-based localization: camera calibration and video stabilization. A stereo-vision method using Fuzzy Camera Calibration (FCC), Fuzzy Optical Flow (FOF) and Fuzzy Gaussian Pyramid methods is proposed to improve robot localization. As described, 11 chapters which focus on control and sensor technology and their related systems are included in this volume. We hope that the contributions can enlighten the readers with various applications of control and sensor technology and the cutting-edge research in this field. The chapters will be beneficial to researchers, graduates, academics, and industrialists who have interests in and/or are dealing with such technology, whether directly or indirectly.

Design and Control of a 3D Robot-Assisted Rehabilitation Device for Post-Stroke Shahrul Na’im Sidek and Sado Fatai

Abstract With the ever-increasing population of stroke patients requiring rehabilitation therapy compared with the few available therapists, what is now crucial is an adaptive system that can complement closely the role of an expert therapist by sensing the patients’ muscle tone, physical recovery condition, or sensorimotor control performance to specify appropriate therapy and to provide an assessment. A “high-level” adaptive hybrid impedance controller based on Modified Ashworth Scale (MAS) assessment criteria for rehabilitation of the upper extremity of poststroke patients was therefore proposed and discussed in this chapter. An end-effectorbased 3 degree-of-freedom (3-DOF) rehabilitation platform was developed with the proposed control strategy with the emphasis on proper joint coordination and control to actualize effective trajectory tracking and consequently effective therapy.

1 Introduction Rehabilitation therapy post-stroke is crucial in helping patients to regain as much possible use of their paretic limbs in the activities of daily living. The major challenge, however, in contemporary post-stroke rehabilitation therapy especially in the post-acute phase of stroke recovery is that the therapy is time-demanding and laborintensive. Therefore, expensive with a consequent reduction in the amount of training sessions required for optimal therapeutic outcome. In recent years, the use of robotic devices for rehabilitation therapy has been widely favored. Robot-assisted rehabilitation therapy is cost-effective, fatigue-free, and has the potential to improve the efficiency of the rehabilitation process. More so, positive outcomes of improved motor control abilities for patients undergoing robot-assisted therapy have been widely recorded through proper developed exercise programs that are usually task-specific and intensive, and require progression of difficulty. S. N. Sidek (B) · S. Fatai Department of Mechatronics Engineering, International Islamic University Malaysia, P.O. Box 10, 50728 Kuala Lumpur, Malaysia e-mail: [email protected] © Springer Nature Switzerland AG 2022 M. Mariappan et al. (eds.), Control Engineering in Robotics and Industrial Automation, Studies in Systems, Decision and Control 371, https://doi.org/10.1007/978-3-030-74540-0_2

5

6

S. N. Sidek and S. Fatai

This chapter presents the development and control of a portable 3-DOF endeffector-type robotic device for upper-extremity rehabilitation for hemi-paretic poststroke patients. The device has three active DOFs consisting of two revolute joints and one prismatic joint (R-R-P) designed to allow three-dimensional range of motion (ROM) exercise for elbow and shoulder rehabilitation. An adaptive hybrid impedance control framework has also been developed for the device to allow safe robot–patient dynamic interaction during the planned repetitive range of motion exercises and to keep track of patients’ motor recovery based on an embedded Modified Ashworth Scale (MAS) muscle assessment criteria. Experimental results performed, using a healthy subject, to test and evaluate the ability of the device to track a planned simple flexion/extension range of motion exercise for the elbow joint showed the possibility of use of the device for real patients.

2 Background In the last few decades, the increasing cases of upper-extremity disabilities resulting from stroke, spinal cord injuries (SCI), and other related illness have favored the use of robotic devices to provide assistance and support to patients undergoing rehabilitation therapy [11, 14]. The effectiveness of the devices in extending the therapy session and in providing repetitive exercises aimed at inducing motor plasticity have been widely reported [8, 11]. Robotic devices have the potentials of allowing repeatability and automation of therapeutic procedures with significant improvement on the efficiency of the rehabilitation process [13]. With the increasing demand, for robotic devices, the need for portable, lightweight, and safety devices have created new challenges ranging from the design of compact robot mechanical/actuation systems to the development of effective control algorithms. A survey on robotic devices in the recent past showed that most of the earlier robotic devices and some of the recent commercially available ones are mechanically bulky and suited for use only in rehabilitation centers [2, 4, 5]. Besides, most are less autonomous to the extent that they require constant assistance of a human therapist for effective usage, for task specification, and for difficulty level adjustment of therapy [5]. This implies a considerable amount of valuable therapist time in programming the robot, monitoring the patients, analyzing data from the robot, and also assessing the progress of the patients. Krebs et al. [6, 10] earlier proposed a 3-DOF robotic device that allowed the horizontal motion and a 6-DOF Mirror Image Movement Enabler (MIME) robotic device that allowed three-dimensional spatial motion respectively for upper-limb rehabilitation which was generally bulky and suited only for rehabilitation centers. The devices were less autonomous and require the manual monitoring of patients’ progress by an expert therapist. A 9-DOF cabledriven robotic device was also proposed by Loureiro et al. [9] based on GENTLE/S system. The device was effectively used for reach and grasp therapy for post-stroke patients but was equally bulky and non-adaptive to patients’ recovery. A more recent commercially available robotic device referred to as Armeo Power [12] with 7-DOF

Design and Control of a 3D Robot-Assisted …

7

for elbow and shoulder rehabilitation and additional DOFs for wrist and fingers flexion/extension rehabilitation exercise which has been successfully used in many rehabilitation centers is also seen to be largely stationary and not lightweight. As the development of robotic devices for rehabilitation exercises continues, the need for portable and lightweight devices capable of autonomous guidance and monitoring of patients’ progress, therefore, remains a critical driving factor of consideration [5]. In this chapter, the development and control of a 3-DOF end-effector robotic device for autonomous rehabilitation of the shoulder and elbow region of patients with hemi-paretic upper-limb impairment is presented. The device is made portable and lightweight to allow flexibility of usage and for possible use at homes. In addition, a novel adaptive control framework as reported in Sado et al. [15] is developed for the device to allow safe robot–patients dynamic interaction and effective tracking of a planned range of motion exercise, and independent monitoring of patients’ physical recovery progress.

3 System Description 3.1 System Modeling—Forward Kinematics The robotic device has three active DOFs arranged in a revolute–revolute–prismatic (R–R–P) configuration as shown in Fig. 1. The three joints allow for the rehabilitation (flexion, extension, abduction, and adduction) of the elbow and shoulder of the patients in three-dimensional space. The relationship between the joints variables and the position and orientation of the robot’s end-effector is derived from the four Denavit–Hartenberg (DH) parameters given in Table 1. The four parameters ai , α i , d i , θ i are generally known as the link length, link twist, link offset, and joint angles respectively [16]. The link transformation matrices which relate any two successive frames attached to the joints are derived from the homogenous transformation matrix [16] given by Eq. (1). ⎤ −sθi 0 ai−1 cθi ⎢ sθi cαi−1 cθi cαi−1 −sαi−1 −sαi−1 di ⎥ i−1 ⎥ ⎢ i T =⎣ sθi sαi−1 cθi sαi−1 cαi−1 cαi−1 di ⎦ 0 0 0 1 ⎡

(1)

where s and c stands for sine and cosine of the angles. The homogenous transformation matrix relating the end-effector frame (x 3 , y3 , z3 ) to the base frame (x 0 , y0 , z0 ), see Fig. 1, is therefore obtained by multiplying the individual link transformation matrices as shown in Eq. (2). 0 3T

= 01 T × 12 T × 23 T

(2)

8

S. N. Sidek and S. Fatai

Fig. 1 The robot kinematic model

Table 1 The DH-parameters Link

ai

αi

di

θi

1

0

−90

0

θ1

2

l

90

0

θ2

3

0

0

d3

0

Finally, the forward kinematics equation derived from Eq. (2) is expressed as Eq. (3). ⎤ ⎡ ⎤ ⎡ x cθ1 (lcθ2 + d3 sθ2 ) ⎣ y ⎦ = ⎣ sθ1 (lcθ2 + d3 sθ2 ) ⎦ z

(3)

d3 cθ2 − lsθ2

3.2 System Modeling—Inverse Kinematics The inverse kinematics of the rehabilitation robot is derived in order to determine the joint angles or variables in terms of the end-effector position and orientation. It

Design and Control of a 3D Robot-Assisted …

9

is however more computationally demanding than the forward kinematics analysis due to the non-linearity in the forward kinematics equation. Since the robotic device is designed with two-axis intersecting, the inverse kinematic analysis becomes fairly easy with a possible closed-form solution. Observing that the forward kinematics equations are transcendental, the following substitutions are made: u i = tan

θi , cos θi = 2

1−u i2 1+u i2

, sin θi =

2u i 1+u i2

(4)

where ui denotes the tangent of any half angle, and i = 1, 2. By substituting Eq. (4) in Eq. (3) and solving for the joint variables q = [θ1 θ2 d3 ], the inverse kinematics equations for the robotic system is obtained as follows: ⎫ y θ1 = Atan2(± ⎪ ⎪  x , 1)

⎬ d3 −z 2 θ2 = Atan2(−l ± l + d3 +z , 1) ⎪ ⎪  ⎭ d3 = ± x 2 + y 2 + z 2 − l 2

(5)

The solution is however not unique since there is the possibility of different configurations (some unreachable) for a given solution. Therefore, the use of inverse kinematics for controller development is avoided as will be seen in the next section.

3.3 System Modeling—Velocity Kinematics: The Jacobian The velocity kinematics equation relates the linear velocity, v and angular velocity, q of the end-effector to the joint velocities of the robotic device. Similar to the forward kinematics equation, it defines the map between the Cartesian space and joint space. This relationship is expressed by the Jacobian time-varying linear transformation given by Eq. (6). v = J (q)q˙

(6)

Taking the first derivative of the forward kinematics equations (refer to Eq. 3) and using the techniques described in Spong et al. [16], the Jacobian, J is obtained and given as

10

S. N. Sidek and S. Fatai

⎡

−s1 (lc2 + d3 s2 ) −c1 (lc2 + d3 s2 ) ⎢ c (lc + d s ) −s (lc + d s ) 3 2 1 2 3 2 ⎢ 1 2 ⎢ 0 k ⎢ J =⎢ ⎢ 0 −s1 ⎢ ⎣ 0 −c1 1 0

⎤ c1 s2 s1 s2 ⎥ ⎥ ⎥ c2 ⎥ ⎥ 0 ⎥ ⎥ 0 ⎦ 0

(7)

where k =− c21 (lc2 + d 3 s2 ) − s21 (lc2 + d 3 s2 )

3.4 System Modeling—the Robot Dynamics To obtain the robot dynamic model, the Lagrange equation of motion for a conservative system is adopted and given by Eq. (8) ∂L d ∂L − =τ dt ∂ q˙ ∂q

(8)

where q is the 3 × 1 vector of generalized joint coordinates, τ is the 3 × 1 vector of generalized input actuator forces/torque, and L the Lagrangian. The Lagrangian is given by the expression L=K−P

(9)

where K is the kinematic energy of the robotic system and P is the potential energy of the system. Using Eqs. (8) and (9), the overall dynamic model of the robotic device is derived and expressed by Eq. (10) which is a requisite for developing the control algorithm. ˙ q˙ + G(q) = τ Mr (q)q¨ + Iact q¨ + C(q, q)

(10)

˙ M r represents the robot mass matrix, I act represents the actuator inertia, C(q, q) represents the Coriolis and centrifugal terms, and G(q) represents the gravity term. Since the joint frictions are small for the robotic device, these components are assumed negligible in the robot dynamic model.

3.5 Control Architecture—Sensors and Actuation System A DC motor coupled at the prismatic joint, at frame (x 2 , y2 , z2 ), provide linear actuation by means of a lead screw mechanism to allow flexion and extension of the

Design and Control of a 3D Robot-Assisted …

11

Fig. 2 The mechanical structure of the 3-DOF robotic device

elbow joint along 90˚ to the supine (other configurations are possible when actuation is done in conjunction with other joints actuation), see Figs. 1 and 2. The DC motor is equipped with an incremental encoder which gives a precise measurement of linear position, velocity, and acceleration of the end-effector. The other two revolute joints at frame (x 0 , y0 , z0 ), denoted base frame, and frame (x 1 , y1 , z1 ) are actuated separately by means of brushless DC motors with coupled incremental encoders to allow precise measurement of joint angular positions, velocities, and accelerations along with their respective frames. Actuation of the base frame allows for rotational motion along the vertical axis (z0 ) which assists shoulder rehabilitation, and in conjunction with the second joint actuation allows abduction and adduction ROM exercises of the shoulder. The actuation of the second joint at frame (x 1 , y1 , z1 ) allows tilting or rotational motion along the horizontal axis (z1 ). In addition, contact force measurement is achieved at the end-effector location by means of a three-axis load cell custom fabricated into the end-effector (see Fig. 2).

3.6 Control Architecture—Control Hardware A hardware control unit designed for the robotic device consists of a central microcontroller (Arduino Mega), see Fig. 3. The microcontroller provides an interface between the Matlab/Simulink control model on the PC side and the robotic device. The unit also consists of motor drivers which provide an interface between the microcontroller and the motors, including the PWM to analog voltage converters which output an analog voltage corresponding to the control torque on the PC side for tuning of the motor speed.

12

S. N. Sidek and S. Fatai

Fig. 3 Control hardware and interfacing

3.7 Control Architecture—Feedback Control Linearization Strategy Since the robot will interact with an environment (patients or healthy subjects in this case) with 3 × 1 joint space reaction torques given by τ e , the robot-patient dynamic motion given by Eq. (10) is modified as ˙ q˙ + G(q) = τ − τe M(q)q¨ + Iact q¨ + C(q, q)

(11)

The vector of environment reaction torque τ e in joint space is related to the task space contact force, f e, by the Jacobian matrix, J(q), given by τe = J T (q) f e

(12)

Equation (11) can thus be rewritten as τ = M(q)q¨ + C(q, q) ˙ q˙ + G(q) + J T (q) f e

(13)

where M(q) represents the sum of the total robot mass matrix and the inertia of the actuator (M r (q)+ I act ). The control objective is achieved by linearizing the non-linear dynamic equation given by Eq. (13) using the concept of inner/outer loop control strategy [1] as shown in Fig. 4. The inner loop represents the non-linear feedback linearization or inverse dynamics control (given by Eq. 14) and the outer loop represents a hybrid impedance control strategies (given by Eq. 15). ˙ + N (q, q) ˙ + J T (q) f e τ = M(q)J −1 (q)(a − J˙(q)q) a = S Mm−1 (( f e − f d ) − Bm x) ˙ + Km ˙ +M (xd − x) − (I − S)(x¨d + MBmm (x˙d − x) m

1 Mm

fe )

(14)

(15)

Design and Control of a 3D Robot-Assisted …

13

Fig. 4 Inner and outer loop control strategies

where M m , Bm , and K m represent the robot virtual impedance parameters (mass, damping, and stiffness factors), S represents the selector matrix for force and position control, f d represents the reference force, and x d represents the reference position.

3.8 Control Architecture—Adaptive Control An adaptive control framework is developed (to allow task space switching of position and force control as well as monitoring of patients’ physical recovery progress) using a novel hybrid automata (HA) framework and an autoregressive exogenous recursive polynomial model estimator (ARX- RPME) as reported in earlier works [15], see Fig. 5 for the adaptive framework. The RPME implements an ARX model of the upper

Fig. 5 The ARX-RPME and the HA model

14

S. N. Sidek and S. Fatai

Fig. 6 The overall adaptive control framework

limb to estimate the mechanical impedance parameters of the limb as a measure of muscle tone. The estimated parameters serve as the inputs to the HA framework by means of which the automata accomplish two tasks. First, allows transition between its seven-state (S 0 … S 6 ) automaton (including a default transition state) based on the level of muscle tone (impairment) as classified by the MAS muscle assessment criteria [3, 15], and second, allows switching, concurrently, of task space between force and position controls, based on the changing upper-limb mechanical impedance and the interpretation of the duality principle [7, 16]. Figure 6 gives the overall control framework.

4 Results and Discussion In order to experiment with the proposed control algorithm on the device, its ability to follow a planned reference ROM trajectory of the end-effector, and also on its ability to maintain a certain reference interactive force at the end-effector/human contact point is analyzed.

Design and Control of a 3D Robot-Assisted …

15

4.1 Trajectory Generation and Position Control To actualize the position control, a simple elbow flexion range of motion exercise of the end-effector along the positive x, y, z directions is recorded (using a healthy subject) and the reference trajectory (from the data) is derived using a second-order polynomial (see Eq. 16). Figure 7 shows the position tracking results for the actual and desired position of the end-effector along the x, y, z directions with RMSE of 0.023, 0.024, and 0.013 m respectively. ⎤ ⎡ ⎤ −0.0005t 2 − 0.0009t + 0.1521 xd (t) ⎣ yd (t) ⎦ = ⎣ 0.0004t 2 − 0.0164t + 0.0078 ⎦ z d (t) 0.0000t 2 − 0.0144t + 0.0012 ⎡

(16)

Fig. 7 Experimental results for position tracking in the a z-direction b y-direction c x-direction

16

S. N. Sidek and S. Fatai

4.2 Force Tracking To test the robotic device/controller on the ability to maintain a certain amount of contact force (exerted by the end-effector on the environment/subject’s limb), a reference force trajectory using a fourth-order polynomial is planned such that the robot end-effector is made to exert force (by actuating the individual joints) on the subject’s upper limb steadily from 0 to a maximum of 50 N, and thereafter reduce the exerted force gradually to zero over a computer time of 21 s. The generated reference trajectory along the positive z-direction of the end-effector is given by Eq. (17). Figure 8 shows the force tracking results by the controller along the z-direction with an RMSE of 0.343 N. f dz (t) = 0.0009t 4 − 0.0347t 3 + 0.3535t 2 − 0.517t − 0.7265

(17)

As can be seen from Fig. 8, there is fairly good tracking of the reference force trajectory due to noise from the custom-made force sensor. As a modification, for future studies, a signal conditioning circuit will be adopted therefore to eliminate noisy signals. Regarding stability, no analysis for controller stability is given in this work since the stability of control frameworks based on the standard computed torque control, or inverse dynamic control, has been generally established. However, mechanical stability is guaranteed using worm gears and a firm tripod support.

Fig. 8 Experimental results for force tracking along the z-direction

Design and Control of a 3D Robot-Assisted …

17

5 Conclusion A new portable 3-DOF end-effector robotic device for autonomous rehabilitation of the shoulder and elbow region of patients with hemiparetic upper-limb impairment has been presented. A dedicated control framework is proposed and developed in the previous study, by the authors, and is adopted to allow safe robot–patients dynamic interaction and effective tracking of a planned range of motion exercise, as well as independent monitoring of patients’ physical recovery progress. Experimental results have shown the feasibility of the use of the device for trajectory motion tracking of a reference ROM exercise. Future works however include experimentation to test the adaptability of the robot control scheme to changing the upper-limb mechanical impedance and final experimentation on real patients.

References 1. Anderson, R., Spong, M.W.: Hybrid impedance control of robotic manipulators. IEEE J. Robot. Autom. 4, 549–556 (1988) 2. Bogue, R.: Exoskeletons and robotic prosthetics: a review of recent developments. Ind. Robot Int. J. 36, 421–427 (2009) 3. Bohannon, R.W., Smith, M.B.: Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys. Ther. 67, 206–207 (1987) 4. Dellon, B., Matsuoka, Y.: Prosthetics, exoskeletons, and rehabilitation. IEEE Robot. Autom. Mag. 14, 30 (2007) 5. Díaz, I., Gil, J.J., Sánchez, E.: Lower-limb robotic rehabilitation: literature review and challenges. J. Robot. 11 (2011) 6. Krebs, H.I., Ferraro, M., Buerger, S.P., Newbery, M.J., Makiyama, A., Sandmann, M.: Rehabilitation robotics: pilot trial of a spatial extension for MIT-Manus. J. Neuroeng. Rehab. 1, 5 (2004) 7. Lewis, F.L., Dawson, D.M., Abdallah, C.T.: Robot Manipulator Control: Theory and Practice. CRC Press (2003) 8. Lo, H.S., Xie, S.Q.: Exoskeleton robots for upper-limb rehabilitation: state of the art and future prospects. Med. Eng. Phys. 34, 261–268 (2012) 9. Loureiro, R.C., Harwin, W.S.: Reach and grasp therapy: design and control of a 9-DOF robotic neuro-rehabilitation system. In: Rehabilitation Robotics, 2007. ICORR 2007. IEEE 10th International Conference on, pp. 757–763 (2007) 10. Lum, P.S., Burgar, C.G., Van der Loos, M., Shor, P.C., Majmundar, M., Yap, R.: MIME robotic device for upper-limb neurorehabilitation in subacute stroke subjects: a follow-up study. J. Rehabil. Res. Dev. 43, 631 (2006) 11. Maciejasz, P., Eschweiler, J., Gerlach-Hahn, K., Jansen-Troy, A., Leonhardt, S.: A survey on robotic devices for upper limb rehabilitation. J. Neuroeng. Rehab. 11 (2014) 12. Nef, T., Guidali, M.V., Klamroth-Margansk, A.V., Riener, R.: ARMin-exoskeleton robot for stroke rehabilitation. In: World Congress on Medical Physics and Biomedical Engineering, pp. 127–130. Munich, Germany (2009) 13. Nef, T., Mihelj, M., Riener, R.: ARMin: a robot for patient-cooperative arm therapy. Med. Biol. Eng. Compu. 45, 887–900 (2007) 14. O’Donnell, M.J., Xavier, D., Liu, L., Zhang, H., Chin, S.L., Rao-Melacini, P.: Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22 countries (the INTERSTROKE study): a case- control study. The Lancet 376, 112–123 (2010)

18

S. N. Sidek and S. Fatai

15. Sado, F., Sidek, S.N., Yusof, H.M.: Adaptive hybrid impedance control for a 3DOF upper limb rehabilitation robot using hybrid automata. In: 2014 IEEE Conference on Biomedical Engineering and Sciences (IECBES),pp. 596–601. Miri, Sarawak (2014) 16. Spong, M.W., Hutchinson, S., Vidyasagar, M.: Robot Modeling and Control, vol. 3. Wiley, New York (2006)

Wet Scrubber Design B. A. Danzomo, Sambo A. Umar, and Momoh Jimoh E. Salami

Abstract Wet scrubber systems have inherent advantages over other air pollution control devices as they have the ability to absorb gaseous pollutants, remove flammable and explosive dust particles safely. Though various types of wet scrubbers systems exist, spray towers based are preferred due to their simplicity in design, least energy consumption, cheaper to construct and maintain, less space requirements and operation with slight pressure drop, ability to handle a large volume of gases as well as engaged for the dual purpose of absorbing gaseous pollutants while removing particle contaminants. The mechanisms for the separation of the particle pollutant from the gas stream include impaction, interception, and diffusion, an inertia impaction mechanism is used in the spray tower wet scrubber system. This mechanism limits the size of particle contaminants control to Particle Matter (PM) size ≥ 5 μm, whereas PM10 and PM2.5 ≤ 5 μm pollutants from industrial sources constitute a great danger to human health. Consequently, several attempts have been made to improve the performance of spray tower wet scrubber for the pollutant control of PM10 and PM2.5 ≤ 5 μm. Two approaches for improving the design of wet scrubber systems, namely, similitude model design and computational fluid dynamics approaches have been discussed in this chapter. A pilot scrubber system for PM2.5 and PM10 control has been designed using PM data obtained from Ashaka cement industry in Gombe State, Nigeria. A Hydraulic Similitude approach has been employed to design a scaled model of the scrubber system. The airflow velocity and pressure fields within the scrubber system were simulated using ANSYS Fluent software to obtain optimum design of the system, improve efficiency, shorten B. A. Danzomo Department of Mechatronics Engineering, College of Engineering, Hussaini Adamu Federal Polytechnic, Kazaure PMB 5004, Jigawa State, Nigeria S. A. Umar (B) Faculty of Engineering and Engineering Technology, Department of Mechatronics and Systems Engineering, Abubakar Tafawa Balewa University, Bauchi PMB 0248, Bauchi State, Nigeria e-mail: [email protected] M. J. E. Salami Department of Electrical and Electronics Engineering, Elizade University, Ilara Mokin, Ondo State, Nigeria © Springer Nature Switzerland AG 2022 M. Mariappan et al. (eds.), Control Engineering in Robotics and Industrial Automation, Studies in Systems, Decision and Control 371, https://doi.org/10.1007/978-3-030-74540-0_3

19

20

B. A. Danzomo et al.

experimental period, and avoid dead zone. Some simulated results are presented to justify the essence of the designed approaches.

1 Introduction The interest in wet scrubber systems has been attributed to their important advantages when compared to other air pollution control devices. The device can absorb gaseous pollutants and remove flammable and explosive dust particles safely. Wet scrubbers have been successfully used for medical waste incineration and other industrial applications such as; cement industries, fertilizer industries, meat processing industry, coal mining plant, pig and poultry farms, etc. In a scrubber system, the scrubbing liquid comes into contact with gas and particle contaminants or dust particles. The greater the contact of the gas particle and liquid streams, the higher the particle removal efficiency [7]. The first patent for a particulate matter scrubber design was issued in 1892 in Germany, and the first gas scrubber with rotating elements was developed in 1900 [7]. The first Venturi scrubber was developed in the United States after World War II and mainly represented a breakthrough in scrubber design [21]. Among various wet scrubber systems, spray towers are considered to be simple and least energy consuming of all wet scrubbers, cheaper to construct and maintain, is a system that needs small space requirements and causes a slight pressure drop, can handle a large volume of gases, and they can be engaged for the dual purpose of absorbing gaseous pollutants while removing particle contaminants. Gas–particle separations in a spray tower wet scrubber shown in Fig. 1 is the process by which the polluted gas and the particle rises upward and the particles collide with liquid droplets formed by atomization of a spray nozzle situated across the flow passage, where the flow rate of the scrubbing liquid and pressure in the atomizing nozzle control the droplet size and number. Due to the action of the inertial separation mechanisms, the particle is separated from the gas stream and a clean gas passes through a mist eliminator placed at the top of the scrubber to remove excess clean and dirty water droplets. The particle contaminants and the scrubbing liquid form slurry which accumulates at the bottom of the scrubber system and can be recycled into a usable state. According to DoE [10], Cement dust PM slurry can be used to I. II. III.

Eliminate SO2 from flue gas contaminants at very high efficiencies, Recycle dust PM back to the cement kiln, Recover K2 SO4 which is a saleable fertilizer-grade by-product.

While the slurry formed from scrubber operations for the control of ammonium nitrate particle contaminant can be (a) (b)

Poured into the soil and it will still serve its purpose as a fertilizer or Used in cold packs.

Wet Scrubber Design

21

Fig. 1 Gas–particle separations in spray tower wet scrubber system

The inertial separation mechanisms which caused the separation of the particle pollutant from the gas stream as described by Mahajan [23] include impaction, interception, and diffusion mechanism. The primary mechanism by which particle pollutants are separated in spray tower wet scrubbers is inertial impaction. This mechanism is limited to control of particle contaminants ≥5 μm in size [Kim 2001, 12, 28]. But, the particle size that represents a significant fraction of particle pollutant (particulate matter, PM) emission from most industrial sources and also forms the regulated particles that penetrate the lower respiratory tract of human lungs as shown is the size of PM10 and PM2.5 ≤ 5 μm [26]. Several attempts have been made to improve the performance of spray tower wet scrubber system for the control of PM2.5 and PM10 ≤ 5 μm such as Bozorgi et al. [4, 5, 12, 24, 28, 36, 37]. Other studies involved the application of Computational Fluid Dynamics (CFD) such as Pieloth [13, 20, 30]. In this chapter, to improve the design of the scrubber system, hydraulic similitude model design method and CFD

22

B. A. Danzomo et al.

techniques have been employed. This goes a long way to provide the optimum gas and particle velocities, liquid to gas ratio, and the height of the system for effective scrubbing.

2 Wet Scrubber Design Exhaust PM data obtained from Ashaka [1], (Tables 1, 2 and 3) in Gombe state, Nigeria were used to design the proposed wet scrubber system for Gas–PM separations. Also, a computational fluid dynamics analysis of the scrubbing chamber was performed so as to describe the flow direction of the Gas–PM within the inlet and outlet ducts and the scrubbing chamber.

2.1 Sizing the Scrubber System Fluid dynamics is one of the primary approaches used to design a wide variety of engineering devices. This chapter intends to use empirical relations which include continuity equations and other flow dynamic equations and data in Table 1 to design Table 1 Summary of kiln shop exhaust gas–PM data Air pollution control equipment used

Operating parameter

Operating data

Unit

Electrostatic precipitator (ESP)

Volume flow rate

104,885

m3 /hr

Mass flow

119,084

Kg/hr

Gas temperature

105

°C

Gas pressure

−12

mmWg

Gas density

0.82

Kg/m3

Dust burden (Inlet)

62,500

μg/m3

Dust burden (Outlet)

25,000

μg/m3

Table 2 Summary of mill shop exhaust gas–PM data Air pollution control equipment used

Operating parameter

Operating data

Unit

Electrostatic precipitator (ESP)

Volume flow rate

14,618

m3 /hr

Mass flow

22,765

Kg/hr

Gas temperature

91.7

°C

Gas pressure

17.76

mmWg

Gas density

0.932

Kg/m3

Dust burden (Inlet)

22,859,000

μg/m3

Dust burden (Outlet)

2,250,000

μg/m3

Wet Scrubber Design

23

Table 3 Summary of coal mill shop exhaust gas–PM data Air pollution control equipment used

Operating parameter

Operating data

Unit

Baghouse filter (Fabric Filter)

Volume flow rate

51,701

m3 /hr

Mass flow

68,586

Kg/hr

Gas temperature

100

°C

Gas pressure

−22

mbar

Gas density

0.95

Kg/m3

Dust burden (Inlet)

NA

μg/m3

Dust burden (Outlet)

28,000

μg/m3

the proposed spray tower wet scrubber system for the gas–PM separations. According to Jerry et al. [16], the waste gas flow rate is the most important sizing parameter in a wet scrubber. Therefore, a steady flow involving a stream of a specific fluid flowing through the spray tower scrubber at Sects. 1 and 2 is considered by Munson et al. [27]. ρ1 A1 U1 = ρ2 A2 U2

(1)

where ρ 1 and ρ 2 are the respective densities, A1 and A2 are the cross-sectional areas, while U 1 and U 2 are the inlet and outlet velocities respectively. From Table 1, the gas density is given by 0.82 kg/m3, while the mass flow rate is 119,084 m3 /hr (33.08 kg/s). Considering Eq. (1), m˙ 1 = ρ2 A2 U2

(2)

m˙ 1 = mass flow rate of the exhaust gas (scrubber inlet gas) = 33.08 kg/s The density, ρ 2 is ρ2 =

ρin + ρout 2

ρin = ρ1 = density of the inlet scrubber gas = 0.82 kg/m3 ρout = ρsat (T ) at T = 30◦ C, from table of saturated steam, ρout = 0.03056 kg/m3 0.82 + 0.03056 = 0.4253 kg/m3 ρ2 = 2 The cross-sectional area, A2 is given by the expression

(3)

24

B. A. Danzomo et al.

A2 =

π D2 4

(4)

D is the cylindrical spay chamber diameter. Considering the maximum gas velocity recommended from the simulation analysis in this chapter and literature, U g = U 2 = 1.2 m/s, the diameter of the scrubber is determined using Eq. (2) m˙ 1 = ρ2 ×π 4D × U2  4×33.08 1 ⇒ D = ρ2 4×m = 0.4253×π×1.2 = 9.08 ≈ 9.0 m ×π×U2 2

According to Garba [12], the typical height to diameter ratio of a cylindrical shell is approximately 2:1. Therefore, the height of the scrubber system, Z = 9 × 2 = 18 m.

2.2 Determining of the Scrubber Wall Thickness In designing the scrubbing chamber, consideration should be laid on the thickness of the metallic body. According to Sturm [34], the sole design basis for establishing the metal thickness of a shell or thin cylinder in a system is the external pressure acting on the cylinder wall. This external pressure is also the collapsing pressure, Pe, on a submerged submarine and vacuum tanks given by 



Pe = K 1 + K 2

D t

2 

t D

3 E

(5)

K 1 and K 2 are the numerical coefficients depending on the length to diameter ratio (L/D) and diameter to thickness ratio (D/t) respectively. An alternative expression for the determination of the metal thickness of the scrubber wall is described by Eq. (6). 

t Pe = K E D

3 (6)

A carbon steel material was selected for the design of the scrubber wall, from metals and material table, the modulus of elasticity, E =200.1 x109 N/m2 . But the collapsing pressure in the scrubbing chamber is atmospheric, therefore, Pe = 101.3 × 103 N/m2 . Assuming a factor of safety of 2, Pe = 2×(101.3 × 103 ) = 202.6 × 103 N/m2 . The numerical coefficient, K = 50 was adopted from the study conducted by Garba [12]. However, from the expression for the collapsing pressure, the thickness, t has been computed by making t the subject in Eq. (6), therefore

Wet Scrubber Design

25

 t=  t=

202.6 × 103 50 × (200.1 × 109

Pe KE

1/ 3

1/ 3

×D

(7)

× 9 = 0.0245 m = 24.5 mm

2.3 Determination of Quantity of Water Used in Scrubbing Liquid to gas ratio plays a significant role in wet scrubber performance. Considering the liquid to gas ratio recommended by Jerry et al. [16] is between 0.7 and 2.7 L/m3 , the quantity of liquid needed for effective wet scrubbing can be determined. Considering the maximum ratio of 2.7 L/m3 and Table 1, the volumetric flow of the PM into the scrubber has been determined as 29.13 m3 /s (Table 4). Then the appropriate quantity of liquid needed for an effective scrubbing process is given by Q i = 29.13 × 2.7 = 78.651 L/s

(8)

For proper scrubbing, the proposed scrubber is divided into four stages, which means that Qi = 78.651/4 = 19.66 L/s of water is discharged at every stage.

2.4 Number of Nozzles in the Scrubber System Critical velocity is the velocity for liquid atomization which depends on the gas stream, at a lower critical velocity of the gas stream, atomization may not occur resulting in liquid accumulation. Assuming a nozzle critical velocity of 4.6 m/s and Table 4 Recommended liquid to gas ratios for various wet scrubbers (Table developed by referring to the study of Jerry et al. [16])

S/N

Wet scrubber

Liquid to gas ratio L/m3

Gal/1000 ft3

4.0–5.0

3.0–40.0

1.

Venturi

2.

Spray tower

0.7–2.7

5.0–20.0

3.

Cyclonic spray

0.3–1.3

2.0–10.0

4.

Moving bed

0.4–8.0

3.0–60.0

5.

Orifice (self-induced spray)

0.07–0.7

0.5–5.0

6.

Mechanically aided (fan)

0.07–0.5

0.5–4.0

26

B. A. Danzomo et al.

a nozzle orifice of 3.2 mm [12], the quantity of discharge of each droplet is computed using Eq. (9) q = va

(9)

Therefore, q = 4.6 ×

3.142(0.0032)2 = 3.7 × 10−5 m3 /s = 3.7 × 10−2 L/s 4

Hence, the number of nozzles required discharging 78.65 L/s of liquid within the scrubbing chamber is calculated using Eq. (10) Nn = Nn =

Qi q

(10)

78.651 = 2, 126 nozzles 3.7 × 10−2

2.5 Pipe Network Design 2.5.1

Diameter of the Supply Pipe

Assuming the critical velocity of 4.6 m/s, the diameter of the supply pipe is determined by considering the quantity of liquid used for scrubbing as follows: Q i = Ucrit Asup = Usup  dsup =

2.5.2

4 × Qi = π × Ucrit



2 π dsup

4

(11)

4 × (78.65 × 10−3 ) = 0.15 m π × 4.6

Spray Pipe Diameter

Assuming the scrubber is divided into four sections, the diameter of the spray pipe in each section was determined by dividing the quantity of liquid needed for the scrubbing Q spray =

Qi 78.65 × 10−3 = = 19.66 × 10−3 m3 /s 4 4

Wet Scrubber Design

27

The diameter is given by  dspray =

4 × Q spray = π × Ucrit



4 × (19.66 × 10−3 ) = 0.074 m π × 4.6

2.6 Determination of Duct Diameter When a particulate contaminant is captured by a hood system and enters the ductwork, a minimum transport velocity must be maintained to keep the contaminants from settling out of the gas stream and building up deposits in the ductwork. This leads to decreased hood capture efficiencies and increased fugitive emissions. Proper duct diameter is a key element when addressing minimum transport velocity. If a section of ductwork has a larger than necessary diameter, then settling out will most likely occur. If a section of ductwork is too small, the pressure drop will increase across this section, thus requiring the fan to handle more static pressure. To design the duct system for delivering the PM particle into the scrubber system, an equation that relates the transport velocity and the volumetric flow rate of the gas–PM is considered and this is given by Q G = Aduct Utrans

(12)

The duct diameter can be determined by using the volumetric flow rate of the gas– PM (gas–cement dust) from kiln data in Table 1 (QG = 104,855 m3 /hr = 29.13 m3 /s) and the recommended transport velocity of different contaminants described by Table 5. From the Table, the minimum and maximum range of the transport velocities for the PM contaminants considered in this chapter (PM2.5 and PM10 ) are given as 5.08–20.32 m/s. Using the average of this velocity range (12.7 m/s), the duct diameter has been determined as Table 5 Transport velocities for different contaminants (Table developed by referring to the study of Jerry et al. [16])

S/N

Contaminant

Transport velocity m/s

1.

Vapors, gases, smoke

5.08–10.16

2.

Fume

7.11–10.16

3.

Very fine and light dust

10.16–12.70

4.

Dry dust and powders

12.70–17.78

5.

Average industrial dust

17.78–20.32

6.

Heavy dusts

20.32–22.86

7.

Very heavy or moist dust

>22.86

28

B. A. Danzomo et al.

Aduct =

π dd2 QG QG also, = Ug 4 Ug

Thus  dduct =

4Q G = πUg



4 × 29.13 = 1.7089 ∼ = 2.0 m π × 12.7

2.7 Hood Design Considering the fact that the scrubber gas outflow is the subject of concern in this chapter, a curved transition inform of a hood should be provided so as to obtain shape and good flow characteristics of the gas out of the scrubber with minimum friction loss. To achieve this, a class of circular canopy hoods which draws air or contaminants in tanks or furnaces described by Schnelle [33] has been considered to be installed at the top of the cylindrical spray tower wet scrubber system. From a geometrical point of view, this class of hood is termed as a frustum which includes a portion of a right circular cone that lies between two parallel planes cutting it, in which each of the plane section is a floor or base of the frustum as shown in Fig. 2. Mathematically, the volume flow rate of a circular frustum for 1 s is given as Q frustum =

π 2 H R2 + R1 R2 + R12 3

(13)

Thus Q frustum =

π π 2 H 4.5 + 1 × 4.5 + 12 = H 25.75 = 26.965 H 3 3

Considering the volume of gas flowing through the frustum in 1 s, Qhf Q h f = 26.965 H (m3 /s) Fig. 2 Canopy hood for the scrubber system

(14)

Wet Scrubber Design

29

But volumetric flow rate of the gas–PM from kiln data in Table 1, QG = 29.13 m3 /s, by substituting Qhf = QG . Therefore, the height of the hood, H can be determined as 26.965 H = 29.13 ∴ H = 1.08 m

2.8 Head Losses Within the Pipe Network The head loss in a pipe circuit falls into two categories: i. ii. 2.8.1

Loss in a straight pipe due to viscous resistance extended throughout the total length of the circuit, Loss due to localized effects such as sudden changes in area of flow and bends. Supply Line and Spray Line Head Losses

This represents loss in a straight pipe of diameter d and length L. In order to determine pressure losses in these pipes, friction coefficient is a convenient idea. Assuming the two pipes to be made from galvanized steel length of lengths 16 and 8 m respectively. The diameter of the two pipes has already been calculated to be 0.15 and 0.074 m, while the velocity of flow within the pipes is 4.6 m/s. Considering the scrubbing liquid to be water at 20 °C, for flow through the two pipes, it is well known that the flow goes from laminar to sinuous to turbulent as the Reynolds number increases with the transition to turbulence occurring approximately for Re > 2 × 103 [15]. To determine the frictional losses, the nature of the flow should be known using an expression for Reynolds number: Re =

ρUcrit dsup μ

For the supply pipe, Resup =

998.2 × 4.6 × 0.15 = 6.87 × 106 1.002 × 10−3

Respray =

998.2 × 4.6 × 0.074 = 3.39 × 106 1.002 × 10−3

For the Spray pipe,

(15)

30

B. A. Danzomo et al.

The Darcy–Weisbach Equation [27] is a theoretical equation that predicts the frictional energy loss in a pipe based on the velocity of the fluid and the resistance due to friction. It is used almost exclusively to calculate head loss due to friction in turbulent flow. This is given by hD =

f LU 2 D2g

(16)

The Darcy friction factor, f is usually selected from a chart known as the Moody diagram. The Moody diagram is a family of curves that relate the friction factor, f to Reynolds number, Re and the relative roughness of a pipe, E. This relative roughness is given by the equation ε=

k D

(17)

The frictional coefficient for galvanized steel, k = 0.15 × 10−3 m [3]. The relative roughness for the supply and spray pipes is 0.15 × 10−3 = 0.0010 0.15

εsup = εspray =

0.15 × 10−3 = 0.00196 ∼ = 0.002 0.074

From Moody [25] chart, f = 0.005, when εsup = 0.001 f = 0.006, when εspray = 0.002 Therefore, the Darcy head losses for the two pipes are h D(sup) =

0.005 × 16 × 32 = 0.233 m 2 × 9.81 × 0.15

h D(spray) =

0.006 × 8 × 32 = 0.287 m 2 × 9.81 × 0.074

Since there are four spray lines, the total losses from a spray line will be h T D(spray) = 4 × 0.287 = 1.148 m

Wet Scrubber Design

31

Fig. 3 Flow through a sudden contraction in supply and spray pipes

2.8.2

Losses Due to Sudden Contraction

Sudden contraction in the pipes occurs when the area of the pipe reduces suddenly along the length of the pipe; the downstream velocity, U 2 will be higher than the upstream velocity, U 1 as shown in Fig. 3. Head loss due to this sudden contraction in the pipes can occur and it can be related to the downstream flow velocity and turbulence. An equation for the determination of the head loss due to contraction in relation to the spray pipe downstream velocity, U 2 = 4.6 m/s is given as h LC = kc kc =

U22 2g

(18)

dspray d2 0.074 = 0.4933 ∼ = = = 0.50 d1 dsup 0.15

The value for the coefficient of contraction, k c was interpolated using the data obtained by Garba [12]. This was determined to be 0.335, therefore, the head loss due to sudden contraction in the supply and spray pipes is h LC

 = 0.335

4.62 2 × 9.81

 = 0.36 m

Considering the four spray lines, hTLC = 4 × 0.36 = 1.445 m.

2.8.3

Losses at the Nozzles

There exist a head loss due to localized effects such as sudden changes in the area of flow as in a nozzle having a cone angle, θ and bends. Assume that the nozzle possesses a half cone angle of 20˚ as shown in Fig. 4. Considering the sudden change in area of the nozzle circuits, the head loss is given by an expression for head loss due sudden contraction as

32

B. A. Danzomo et al.

Fig. 4 Flow through a sudden contraction in spray nozzle

h L N = kL N

U22 2g

(19)

From the chart for frictional coefficient in nozzle and diffusers, at a half angle of 20˚, the frictional loss coefficient is approximately k LN = 0.033, while the critical velocity, U 2 = 4.6 m/s, therefore h L N = 0.33

4.62 = 0.0356 m 2 × 9.81

In the scrubber, there are 2126 nozzles; therefore, total losses due to nozzles is given by h T L N = 0.0356 × 2126 = 75.69 m

2.8.4

The Overall Head Loss

The overall head loss is the combination of both the frictional loss, loss due to sudden contraction, and nozzle losses, given as h L = h D(sup) + h T D(spray) + h T LC + h T L N

(20)

h L = 0.233 + 1.148 + 1.445 + 75.69 = 78.516 m However, the total head loss is determined by adding the supply pipe length with the overall head loss. h T = h L + 16 = 78.516 + 16 = 94.516 m

Wet Scrubber Design

33

2.9 Rate of Energy Gained by the Scrubbing Liquid The rate of energy gained by the scrubbing liquid is a measure of the rate of mechanical energy of the liquid as it flows through the pump. An equation relating the hydraulic power and pressure head loss has been adopted from Cengel [6].   p ˙  E = m˙ ρ

(21)

m˙ = ρ L × Q i = 0.9982 × 78.651 L/s = 78.5094 kg/s p = 330, 000 mm H2 O (kg/m2 ) = 3236.19 kpa ρ = ρ L = 998.2 kg/m3 Therefore, the rate of energy gain becomes  E˙ = 78.5094 ×



3236.19 998.2

 = 254.53 kw

2.10 Mechanical Power Delivered to the Pump When the pump operates at 85% efficiency during the wet scrubbing process, then an expression for the pumping efficiency can be used to determine the mechanical power delivered to the pump, Ppump  Ppump = Ppump =

  E˙ ηpump

(22)

254.53 × 103 = 299.45 × 103 = 299.45 kw 0.85

Assuming the efficiency of the electric motor is 90%, then the electric power of the motor, Pelectric becomes   Ppump (23) Pelectric = ηmotor Pelectric =

299.45 × 103 = 332.72 × 103 = 332.72 kw 0.9

34

B. A. Danzomo et al.

2.11 Temperature Rise of the Scrubbing Liquid Considering the 299.45 kw of mechanical power delivered to the pump, only 254.53 kw is gained by the liquid as mechanical energy. The remaining 44.92 kw is converted to thermal energy due to frictional effects, and this lost mechanical energy, E loss, manifests itself as a heating effect in the scrubbing liquid. E loss = Ppump −  E˙ mech

(24)

E loss = 299.45 kw − 254.53 kw = 44.92 kw This relationship can be used to determine the temperature rise of the liquid water using the thermal energy balance [6] ˙ p T E loss = mC

(25)

Since the scrubbing liquid is flowing at 20°C, the specific heat of water at that temperature is 4.1 kJ/kg°C, while the mass flow rate is calculated to be 78.5094 kg/s. Substituting Eq. (25) for the thermal balance, the temperature rise is determined to be   E loss (26) T = mC ˙ p T =

44.92 × 103 ≈ 0.137 ◦ C 78.5094 × 4.18 × 103

This indicates that the scrubbing liquid will experience a temperature rise of 0.137 °C due to mechanical inefficiency, which is very small. However, in an ideal situation, the temperature rise should be less since part of the heat generated will be transferred to the pump casing and to the surrounding air.

3 Hydraulic Similitude Design of the Scrubber System Similitude is the theory and art of predicting prototype performance from model observations using dimensional analysis. It is the main indicator of a known relationship between a model and a prototype. Similitude’s main applications are hydraulic and aerospace modeling. As hydraulic modeling becomes more rigorous, in order to improve modeling accuracy, similitude requirements emerge from the normalization of the mathematical expressions describing the process [32]. Dimensional analysis of pertinent variables may provide a shortcut to identify many if not all of the similitude parameters associated with the process.

Wet Scrubber Design

35

The use of small models for studying the prototype hydraulic design can be dated at least to Leonardo da Vinci (1452–1519). But the method developed for using the results of experiments conducted on a scale model to predict quantitatively the behavior of a full-scale hydraulic structure (or prototype) was realized only after the turn of the century. The principle on which the model design is based comprises the theory of hydraulic similitude [12]. The analysis of the basic relationship of the various physical quantities involved in the static and dynamic behaviors of water flow in a hydraulic structure is known as dimensional analysis. As indicated by Garba [11, 12], all the important hydraulic structures now designed and built after certain preliminary model studies have been completed.

3.1 Similitude Model and Scaling To achieve similarity between model and prototype behavior, all the corresponding pi-terms must be equated between the model and the prototype. However, a model is said to have similitude with the real application if it shares geometric, kinematic, and dynamic similarities between model and prototype.

3.1.1

Geometric Similitude Scaling

The basic and the most obvious requirement of similitude is that the model is an exact geometric replica of the prototype. In this form, the model is a geometrical reduction of the prototype and is accomplished by maintaining a fixed ratio for all homologous geometric dimensions between the model and the prototype by applying the scale factor, λ for the prototype, p and model, m. The pi-terms involved in the geometric similitude include ratios of the important lengths, L or height, H thickness, t diameter, d area, A velocity, U and volume, Q of the model scrubber system. 

= ϕ(L , H, t, d, A, U, Q)

(27)

1m

This is represented by the equation dp Hp tp Lp = = = = λL Lm dm Hm tm

(28)

where λL is the length scale factor for the scrubber diameter, height and thickness.

36

B. A. Danzomo et al.

3.1.2

Kinematic Similitude Scaling

Kinematic similitude implies similarity in motion. The similarity between a model and the prototype is attained if the homologous moving particles have the same velocity ratio along geometrically similar paths. The kinematic similarity involves the scale of time as well as length. The ratio of time, T required for homologous particles to travel homologous distances in a model, T m and its prototype, T P is given by Tp = λT Tm

(29)

The velocity is defined in terms of distance per unit time; thus, the ratio of velocity, U can be expressed as Up = Um

Lp T p2

=

Lm Tm2

Lp Lm Tm Tm

=

=

λL = λa λ2T

λL = λU λT

(30)

The acceleration, a is given as Lp T p2

ap = am

Lm Tm2

(31)

The discharge, Q is expressed in terms of volume per unit time that is Qp = Qm

L 3p Tp

=

L 3m Tm

L 3p L 3m Tp Tm

=

λ3L = λQ λT

(32)

where λT , λU , λa and λQ are the time, velocity, acceleration, and discharge scale factors respectively.

3.1.3

Dynamic Similitude Scaling

Dynamic similitude implies similarity in forces involved in motion such as the viscous, inertia, and pressure forces. The similarity between a model and its prototype is attained if the scaled ratio of homologous forces, F r in the model, F m and prototype, F P is kept at a constant value given by Fp = λF Fm

(33)

Wet Scrubber Design

37

Many hydrodynamic phenomena may involve several different kinds of forces in action. Since models are usually built to simulate the prototype or reduced scales, they usually are not capable of simulating all forces simultaneously. In practice, a model is designed to study the effects of only a few dominant forces.  2 ρ p L 3p T 2p Mpap ρ p 3 λL Fp p 2 λL = = = λ = λ λ = λ p λ2L λU = λ F p L Fm M m am ρm L λ2T λT ρm L 3m LTm2 L

(34)

m

where λF , λρ , and λU are the force, density, and velocity scale factors respectively. The above relation occurs when the controlling dimensionless group on the righthand side of the defining equation is the same for prototype and model. Hence, for a complete dynamic similarity to exist between the prototype and the model, the Reynolds number, Re and the Euler number, Eu have to be the same for the two (prototype and model). Thus ρ p L pU p ρm L m Um = μp μm

(35)

Pp Pm = ρ p U p2 ρm Um2

(36)

According to Bedford [2], in most studies, the scale factor for hydraulic similitude models is chosen as even numbers described as 1:144, 1:72, 1:48, 1:32, 1:24, 1:16, 1:8, and 1:4 respectively. In this chapter, the scaled factor, λ between the designed scrubber (prototype) and the proposed model was chosen to be 1:16 due to smaller space requirements. Thus, the estimated parameters for the proposed model spray tower wet scrubber system have been computed by scaling down the dimensions of the prototype spray tower wet scrubber using hydraulic similitude design. Summary of the design results for the prototype and model spray tower wet scrubber and their corresponding estimated parameters are described in Table 6. Schematic diagram of the designed scrubber system shown in Fig. 5 indicated that the wet scrubber comprises the spray chamber which includes a nozzle spray row and the inlet gas– PM. Considering the fact that the scrubber gas outflow is the subject of concern in this chapter, a curved transition as shown in the figure was provided so as to provide shape and good flow characteristics of the gas out of the scrubber with minimum friction loss. However, this transition was not provided at the slurry exit, because its flow characteristics will not be considered in this chapter.

38

B. A. Danzomo et al.

Table 6 Summary of the design specifications Spray tower wet scrubber parameter

Prototype dimension

Similitude model dimension

Unit

Scrubber diameter

9.0

0.5

m

Scrubber height

18.0

1.0

m

Scrubber wall thickness

24.5

1.53

mm

Supply pipe diameter

0.15

0.0094

m

Supply pipe length

16.0

1.0

m

Spray pipe diameter

0.074

0.0046

m

Spray pipe length

8.0

0.5

m

Scrubber duct diameter

2.0

0.2

m

Hood transition height

1.08

0.1

m

Transport velocity

5.08–20.32

0.32–1.27

m/s

Quantity of liquid for scrubbing

78.65

0.192

L/s

Overall pipe head loss

94.56

1.000139

m

Scrubber pressure drop

0.12–0.75

0.47–2.93

kpa

Number of Nozzles

2,126

5

–

Power delivered to the liquid 72,920

2.0

W

Pump power

91,150

2.5

W

Electric power of the pump motor

113,940

3.0

W

4 Optimization of the Design Using Computational Fluid Dynamics CFD is a way of modeling a complex flow by breaking down geometry into cells that comprise a mesh so as to develop a dynamic model representation of a given system. At each cell, an algorithm is applied to compute the fluid flow for the individual cell. The flow direction of the gas transporting the PM into the scrubbing chamber and its emissions from the chamber into the atmosphere is a key factor which should be considered in developing a wet scrubber system for gas–PM separations. If the drag force of the gas is not sufficient enough to transport the particles within the inlet and outlet ducts and into the scrubbing chamber, the particles may deposit along with the duct ways and the chamber and the gas may develop another swirling action within the chamber. Sequel to this, a CFD analysis is employed so as to analyze this flow and optimize the design effectively. Steps used for the CFD analysis in this chapter are shown in Fig. 6. As indicated by Product Corporation and Kennedy et al. [18, 31], the CFD approach has promising applications in both dry and spray tower scrubber systems. It has the benefit of developing a virtual prototype which can evaluate and compare

Wet Scrubber Design

Fig. 5 Geometric sketch of the scrubber system

Fig. 6 Steps for the CFD analysis

39

40

B. A. Danzomo et al.

design alternatives without material, construction, and testing cost, the advantage of enabling and improving engineering design with tight margins, simulating complex fluid flow phenomena at a reduced cost, and demonstrating project feasibility prior to construction. The CFD governing equations are the conservation equations of mass, momentum, and energy equations. As shown in the wet scrubbing process, the PM particle was transported by a carrier gas into the scrubber and accelerates upward. As the particles are very small, they are strongly coupled with the gas phase. The scrubbing liquid is injected through nozzles and accelerated by the drag force acting on the liquid droplets. Due to relative velocity between the droplets and the particles, a collision occurs, and as result, the particle is captured by the droplet.

4.1 Gas and PM Model Description Particulate matter and gas flow, also referred to as the particle-laden flow, is a class of two-phase fluid flow in which one of the phases is continuously connected (the carrier phase or gas phase) and the other phase is made up of immiscible particle (particulate matter phase).

4.1.1

Gas-Phase Model

The gas phase is air at near-standard temperature and pressure. The appropriate governing equations for the gas flow are the incompressible Navier–Stokes Equation (NSE) for a single-phase flow obtained from the conservation of mass and momentum described by Li et al. []. Continuity Equation ∂(Ui ) =0 ∂ xi

(38)

Momentum Equation

j ρg ∂ Ugi Ug ∂ xi

   j ∂Ugi ∂P ∂ ∂Ug =− + + μeff + ρg gi ∂ xi ∂x j ∂x j ∂ xi

(39)

where i and j are the direction vectors for the three coordinates (x, y, and z), while j U ig and U g are the velocity components (u, v, w) of the gas flow, μeff is the effective dynamic viscosity of gas, ρ g is the gas density, P is the gas pressure, and ρgi is the gas volume force in i-direction. To avoid unnecessary consumption of the computer time for the solution of the full scale, turbulence model for predicting the effects

Wet Scrubber Design

41

of turbulence in the gas phase should be considered. Averaging is often used to simplify the solution of the governing equations of turbulence, but models are needed to represent scales of the flow that are not resolved. One of the most effective viscosity models for the simulations of the turbulent flow is the Harlow–Nakayama k-ε model of the turbulent flow, shown in Eqs. (40) and (41). The model provides the timeaveraged values of velocities and pressure of the gas or air throughout the system. k-Equation  ∂ ∂k i) μ+ ρ ∂(kU + ρU = j ∂x j ∂ xi

∂ xi ∂U j ∂U j ∂Ui μt ∂ x j ∂ x j + ∂ xi − ρε

μi δk

∂k ∂ xi

 +

(40)

ε-Equation ∂ ∂(εUi ) ρ = ∂ xi

 μ+

μt δε

∂ xi

∂ε ∂ xi

 c1 ε ∂Ui μt + k ∂x j



∂U j ∂Ui + ∂x j ∂ xi

 − c2 ρ

ε2 k

(41)

where k is the turbulent kinetic energy, ε is the turbulent kinetic energy dissipation rate, μt is viscosity coefficient of turbulent flow, while δ k ,δ ε , c1 , c2 are the constants; c1 = 1.44, c2 = 1.92, δ k = 1.0, δ ε = 1.3. The standard k-ε model was based on the hypothesis of the isotropic eddy-viscosity, which was modeled through the flow fields of the turbulent kinetic energy and the specific dissipation rate. According to Karthik [17], the model leads to stable calculations that converge easily and allow reasonable predictions for many flows. Also, Li et al. indicated that k-ε is the most commonly used turbulence model which has many advantages; its concept is simple and it has been implemented in many commercial CFD codes. The model has demonstrated the capability to simulate many industrial processes effectively as in Goniva et al. [13, 14, 29, 35].

4.1.2

Particle Phase Model

The particles are injected as a discrete phase in the scrubber system. They are modeled by dividing the dust phase into dust fractions each representing a certain diameter with the concentration of each fraction chosen by considering size distribution of the particle phase. The following assumptions are made for the particle phase modeling such as i. ii.

The dust particle’s mass fraction is negligibly small which can be treated as an additional passive phase having no influence on the gas or droplets. The movement is strongly coupled with fast flow due to a very low relaxation time of the dust particles.

42

B. A. Danzomo et al.

iii.

Due to the effect of gravitational and centrifugal drag forces, the dust particles are allowed to drift with respect to the gas phase.

These assumptions lead to the transport equations for the PM concentration, ϕ j, given by Goniva et al. [13]:

∂φ j + ∇. φ j U j = Ddiff ∇ 2 φ j ∂t

(42)

where U j is the PM particle velocity (U g + U drift ), Ddiff is the diffusivity of a scalar in the fluid which is analogous to the thermal and momentum diffusivity, ϕ j is the concentration of chemical species which is the scalar, and j is the PM particle size. The drift velocity can be determined using the equation for single PM particle by considering a force balance describing the effect of gravity, F g and drag, F D forces. mj But, C D = Aj =

πd 2p ( j) 4

2 dU j A j ρ j C D Udrift = mjg − dt 2 24 Re j

Re j =

ρ p Udrift d p ( j) μg

m j = ρp

(43) πd 3p ( j) 6

Therefore, Eq. (43) becomes 18μg Udrift dU j =g− dt ρ p d 2p ( j)

(44)

dU j Udrift =g− dt τj

(45)

Thus

where τ j is the relaxation time. Also, particle motion can be defined by considering both local and convective particle acceleration. ∂Ug dU j = + Ug (∇.U j ) dt ∂t

(46)

For turbulent particle-laden flows, the particle relaxation time can provide an indication of the responsiveness of the particle to fluid velocity fluctuations. Particles with long relaxation time tend to respond only to the most energetic turbulent eddies with large time scales. Likewise, particles with very short relaxation times closely adhere to the local fluid motion. The ratio of the particle relaxation time, τ j to a relevant fluid time scale, τ f is known as the particle stokes number given by (Timothy et al. 2002).

Wet Scrubber Design

43

Stk = ψ =

τj τf

(47)

4.2 CFD Pre-processing Stage The pre-processing CFD analysis involves identification of the flow region of interest, geometric representation of the region, meshing, and definition of flow physics. Steps for the pre-processing applied in this chapter are shown in (Figs. 7 and 8). Having generated the 3D view, the geometry was imported into ANSYS fluent work-bench (Fig. 9) for the CFD pre-processing process. From the Figure, the pipe geometry for the scrubbing liquid spray and the slurry outlet duct were suppressed from the main geometry so as to analyze only the gas flows across the scrubbing chamber. The geometry was then used to develop volume fill flow which provides flow passages for the gas and the scrubbing liquid across the control volumes of the scrubber system as shown in Fig. 10.

Fig. 7 Pre-processing steps for CFD simulations

44

B. A. Danzomo et al.

Fig. 8 3D geometry of the scrubber system

Fig. 9 3D geometry in ANSYS fluent workbench

Having generated the fluid flow volume from the solid geometry, the solid part of the two geometries were then suppressed leaving only the flow volume which was then saved and exported to mesh workbench where a boundary condition was employed on the volume to define the regions of inflow, outflow, and walls. After the boundary has been created, a mesh was generated across the volume part and an efficient meshing procedure has been applied by adjusting the meshing parameters:

Wet Scrubber Design

45

Fig. 10 Volume fill for the scrubber flow region

the growth rate, the relevance center, and the refinement. The boundary condition and mesh profile for the scrubber geometry are shown in Fig. 11. To obtain an optimum meshing across the scrubber geometry, a growth rate of 1.4 and a refinement of 3 were used at the scrubber inlets, outlets, and the wall. Also, a slow transition and maximum tetrahedral sizes of 0.360 were chosen. This yielded a statistics of 91,051 nodes and 459,870 elements, an aspect ratio of 5.98832, and a skewness of 0.6078 for the first geometry with the suppressed pipe and slurry duct and 183,205 nodes and 96,260 elements, an aspect ratio of 4.21718, and a skewness of 0.52019 for the unsuppressed geometry. However, these values are within the

Fig. 11 Mesh profile and boundary condition for the scrubber geometry

46

B. A. Danzomo et al.

required range of mesh skewness ≤1.0 and aspect ratio ≤10 respectively. After the mesh has been generated for the two geometries, they were saved and exported into the ANSYS Fluent setup and solution for simulation analysis.

4.3 CFD SetUp and Solution Stage The setup and solution stage was performed using ANSYS 13.0 and ANSYS 12.0 In the setup and solution, a pressure-based solver and a gravity of 9.81 m/s2 along the z-axis were selected. The simulation procedures are shown in Fig. 12. Since the main objective of the simulation involves the analysis of gas flow within the system is based on the change in some operating parameters of the scrubber system, air fluid was chosen in the material selection option. Viscous standard k-ε model which is one of the most commonly used turbulence models that gives good results in fluid flow analysis for wall-bounded and internal flows was considered. According to Karthik [17], the model leads to stable calculations that converge easily and allow reasonable predictions for many flows. Also, Li et al. indicated that k-ε is

Fig. 12 Flow diagram of CFD solution procedure

Wet Scrubber Design

47

the most commonly used turbulence model which has many advantages; its concept is simple and it has been implemented in many commercial CFD codes. The model has demonstrated the capability to simulate many industrial processes effectively. For the boundary condition, the velocity inlet has been selected at the gas–PM inlet, while the pressure outlet was chosen for the gas outlet. The value for the inlet values was obtained from the range of recommended gas transport velocities using kinematic similitude model design: 0.318, 0.794, and 1.270 m/s. The simulation starts by identifying and applying conditions at the domain boundaries and then solving the governing equations on the computational mesh iteratively as a steady state or transient. Since this is a pressure-based solution method, the PISO (Pressure Implicit with Splitting of Operators) algorithm recommended by ANSYS® for solving both transient and steady flow calculations (especially when the solution involves turbulence model) has been used. Also, ANSYS® indicated that generally for triangular and tetrahedral meshes, a more accurate result is obtained by using the second-order discretization. There are six differential equations to be solved in 3D and there are six residuals for convergence: continuity, x-velocity, y-velocity, z-velocity, and k-epsilon turbulent equations respectively. These residuals represent a kind of average error in the solution and they were set at 10−6 from the solution monitors option. An absolute initialization was set to compute from the gas inlet. Considering the fact that, the flow field in the wet scrubber is complicated. The gas flow CFD simulation of the scrubber system is based on the following assumptions: i. ii. iii. iv.

The gas-phase flow was only considered in the scrubber and no liquid phase is taken into account. The flow is stable and isothermal, so heat exchange among the phase is not considered. The gas is regarded as incompressible. Three velocities are considered from kinematic similitude design: 0.318, 0.794, and 1.270 m/s.

4.4 CFD Post-processing Stage The CFD solution involves the gas transport velocities obtained from the kinematic similitude design: 0.318, 0.794, and 1.270 m/s which were calculated using an efficient residual of 10−6 starting from 50 iterations. The number of iterations was increased gradually depending on the delay for the solutions to converge. The plot of the residuals and the results of the simulation in velocity, pressure, and density contour displays are subsequently presented.

48

4.4.1

B. A. Danzomo et al.

Graphical Display of the Gas Flow Velocities

Considering the initial velocity of 0.318 m/s, the iteration was set at 50 and then increased up to 1000. But the solution converged after approximately 580 iterations as shown in Fig. 13. Contour display in X- and Y-Coordinates for the scrubber geometry shows clearly the flow of the gas within the chamber as shown in Figs. 14 and 15. Considering 0.794 m/s velocity, the residuals converged after approximately 600 iterations as shown in Fig. 16.

Fig. 13 Iterative residuals at 0.318 m/s

Fig. 14 Gas flow contour at 0.318 m/s in X-coordinate

Wet Scrubber Design

49

Fig. 15 Gas flow contour at 0.318 m/s in Y-coordinate

Fig. 16 Iterative residuals at 0.794 m/s

For the gas flow across the geometry at 0.794 m/s, the contour display of the simulation describing flow is shown in Figs. 17 and 18 using X- and Y-Coordinates. Considering 1.270 m/s velocity, the residual converged after approximately 550 iterations as shown in Fig. 19. For the gas flow across the geometry 1.270 m/s, the contour display of the flow simulation is shown in Figs. 20 and 21 using X- and Y-Coordinates.

4.4.2

Graphical Display of Pressure and Density Due to Gas Flow

Pressure contour describing the static and total pressure in the scrubber system is described below in Figs. 22, 23, 24, 25, 26 and 27 at velocities of 0.318, 0.794, and 1.270 m/s respectively.

50

Fig. 17 Gas flow contour at 0.794 m/s in Y-coordinate

Fig. 18 Gas flow contour at 0.794 m/s in X-coordinate

Fig. 19 Iterative residuals at 1.270 m/s

B. A. Danzomo et al.

Wet Scrubber Design

51

Fig. 20 Gas flow contour at 1.270 m/s in Y-coordinate

Fig. 21 Gas flow contour at 1.270 m/s in X-Coordinate

While the density of the gas flowing through the scrubber geometry has been found to be constant for the three velocities, however, only one diagram is used to describe the contour as shown in Fig. 28.

52

B. A. Danzomo et al.

Fig. 22 Contour of static pressure at 0.318 m/s

Fig. 23 Contour of static pressure at 0.794 m/s

4.5 Results Analysis In this section, the computed result for the flow properties is compared with the results from analytical, computational, or experimental studies to establish their validity. The

Wet Scrubber Design

53

Fig. 24 Contour of static pressure at 1.270 m/s

Fig. 25 Contour of total pressure at 0.318 m/s

sensitivity of the computed results is examined to understand the possible differences in the accuracy of results or performance of the computation with respect to dimensionality, flow conditions, initial and boundary conditions.

54

B. A. Danzomo et al.

Fig. 26 Contour of total pressure at 0.794 m/s

Fig. 27 Contour of total pressure at 1.270 m/s

4.5.1

Analysis of the Velocity Contours

The velocity contours in X- and Y-coordinates for 0.318, 0.794, and 1.270 m/s velocities described the gas flow across the scrubber which includes the spray chamber

Wet Scrubber Design

55

Fig. 28 Contour of density at 0.318, 0.794, and 1.270 m/s

and the inlet and outlet ducts. The velocity vector in Fig. 29 indicates a distributed flow across the spray chamber. The contour plots are further supported by the velocity vector displays across the scrubber inlet and the outlet ducts shown in Figs. 30 and 31 indicated a streamline flow which has conformed to Bernoulli’s theorem.

Fig. 29 Velocity vector of the gas flow within the scrubbing chamber

56

B. A. Danzomo et al.

Fig. 30 Velocity vector of the gas flow along the scrubber inlet

Fig. 31 Velocity vector of the gas flow along the scrubber exit

4.5.2

Analysis of Velocity Profile

Not all fluid particles travel at the same velocity within a pipe. The shape of the velocity curve (the velocity profile across any given section of the pipe) depends upon whether the flow is laminar or turbulent as shown in Fig. 32 [6]. Fig. 32 Velocity profile across a cylindrical pipe for laminar and turbulent flows

Wet Scrubber Design

57

As shown in the diagram in Fig. 32 [6], if the flow in a pipe is laminar, the velocity distribution at a cross section will be parabolic in shape with the maximum velocity at the center being about twice the average velocity in the pipe. In turbulent flow, a fairly flat velocity distribution exists across the section of pipe, with the result that the entire fluid flows at a given single value. The velocity of the fluid in contact with the pipe wall is essentially zero and increases further away from the wall. To validate the gas flow into the scrubber system, velocity profiles for the minimum, average, and maximum velocities across the gas inlet duct were plotted as shown in Figs. 33, 34 and 35 respectively. Since the turbulent model was used in the CFD simulation, from the two figures, it can be seen that the velocity profiles have fully conformed to the recommended profile for turbulent flows in pipes and cylinders.

Fig. 33 Velocity profile for the inlet duct at 0.318 m/s

Fig. 34 Velocity profile for the inlet duct at 0.794 m/s

58

B. A. Danzomo et al.

Fig. 35 Velocity profile for the inlet duct at 1.270 m/s

4.5.3

Analysis of the Drag Force for the Gas and Gravity Force for the Particle

When PM10 and PM2.5 contaminants are subjected to the transport velocities as considered in the simulation (0.318–1.270 m/s), it is expected that the drag force by the gas is much greater than the weight (gravity force) of the particle; therefore, the gas carries along with the particle. This has been justified by considering a spherical particle of mass, m and diameter, d P shown in Fig. 36 moving through the gas under the action of the drag force by the gas and the gravity force of the particle. As shown in the diagram, the particle is carried along the gas stream at a velocity, U g and drag force, F d which opposes the particle weight or the gravity force, F g . The drag force, F d of the gas which is upward around the particle is given by Fd =

2 π C D ρg d P2 Ug 8

(48)

while the gravitational force, F g acting downward is given by the relation Fg =

Fig. 36 Particle flow along the gas stream

π gρ P d P3 6

(49)

Wet Scrubber Design

59

Using Eqs. (48) and (49), the drag force of the gas and the gravity force were computed as follows: The Drag Force: Particle size of 0.1 μm Fd =





2 π × 0.45 × 1.205 × 10−3 × 0.1 × 10−4 × (79.4)2 = 1.79E − 10 8

Particle size of 10 μm Fd =





2 π × 0.45 × 1.205 × 10−3 × 10 × 10−4 × (79.4)2 = 1.79E − 6 8

The Gravity Force: Particle size of 0.1 μm Fg =



3 π × 980 × 1.72 × 0.1 × 10−4 = 8.83E − 13 6

Particle size of 10 μm Fg =



3 π × 980 × 1.72 × 10 × 10−4 = 8.83E − 9 6

From the result, F d > > F g (as Fg /Fd > 300) for both the minimum size (0.1 μm) and the maximum size (10 μm). This indicates that the drag force by the gas is greater than 300 greater than the weight (gravity force) of the particle, hence the gas carries along with the particle and this justifies the assumption that no relative velocity exists between the gas and the particle.

4.5.4

Analysis of the Pressure Drop Across the Scrubbing Chamber

Pressure drop is an important indicator of wet scrubber performance, especially for particulate matter control. The pressure drop is a measurement of the resistance to flow as the flue gas passes from one point to another point and is simply the arithmetic difference between the static pressure at the scrubber inlet and outlet (measured at right angles to the flow). The value of the pressure drop takes into account the amount of energy loss during the wet scrubbing process. According to Box [3], the system pressure is computed using the gauge pressure (pressure measured in a system) and the atmospheric pressure given as Psystem = Pgauge + Patm

(50)

60

B. A. Danzomo et al.

Table 7 Comparison between the desired and computed pressure drop Transport velocity (m/s)

Jerry et al. [16] Recommended pressure drop (pa)

Computed pressure drop using CFD (pa)

0.318

0.47

0.30

0.794

1.7

1.32

1.270

2.93

3.03

The ANSYS Fluent code uses a pressure gauge to measure the static pressure which is also the scrubber systems pressure at the inlet and outlet as shown in Figs. 25, 26 and 27 for the three velocities 0.318, 0.794, and 1.270 m/s respectively. For 0.318 m/s, the static gauge pressure was obtained to be 0.309 pa and 0.0164 at the inlet and the outlet of the system. Also, for the 0.794 m/s, the static gauge pressure was obtained to be 1.32 pa and 0.00442 pa at the inlet and outlet. However, a static gauge pressure of 3.16 pa and 0.131 pa at the inlet and outlet for the 1.270 m/s respectively. The pressure drop, ΔP for the three flows velocities was computed using the relation P = Psystem(1) − Psystem(2)

(51)

Comparison between the dynamic similitude pressure obtained from the recommended pressure drop for spray tower scrubber and the computed pressure drop from the CFD analysis shown in Table 7 indicated a closer agreement.

4.5.5

Analysis of the Total Pressure

The study of pressure across any fluid system arises from Bernoulli’s equation which is significant in the study of an incompressible fluid flow that is fundamental to an understanding of the design and operations of most fluid-based devices. A simplified form of Bernoulli’s equation is described by the word equation as Static Pressure + Dynamic Pressure = Constant

(52)

Equation (52) indicates that the sum of flow static and dynamic pressure is constant along a streamline of fluid flow. According to Cengel [6], the sum of these two pressures is termed as the total pressure. Figures 33, 34 and 35 presented total pressure contours for the three velocities. For 0.318 m/s velocity which is the velocity at which the scrubber performed optimally in the simulation analysis described in chapter three, the total pressure is almost constant across the spray chamber. But the pressure is higher across the inlet and slightly lower along the exit. For 0.794 and 1.270 m/s velocities, the total pressure is slightly higher and approximately constant within the chamber. The pressure is higher across the inlet and slightly lower at the exit as

Wet Scrubber Design

61

in 0.318 m/s velocity. This indicated that the proposed scrubber has conformed to Bernoulli’s equation, especially at a velocity of 0.318 m/s.

5 Conclusion In this chapter, a pilot scrubber system for PM2.5 and PM10 control has been designed using PM data obtained from the Ashaka Cement Industry in Gombe State, Nigeria. A hydraulic similitude approach was employed to design a scaled model of the scrubber system. The airflow velocity and pressure fields within the scrubber system were simulated using ANSYS Fluent software so as to obtain optimum design of the system, improve efficiency, shorten the experimental period, and avoid dead zone. The velocity flow contours and vectors at the inlet, across the scrubbing chamber and the outlet shows a distributed flow and the velocity profiles have fully conformed to the recommended profile for turbulent flows in cylindrical pipes. The total pressure within the scrubber cross-section is constant which follows Bernoulli’s principle. The minimum pressure drop, ΔPmin for the scrubber system was obtained to be 0.30 pa, and the maximum, ΔPmax was 3.03 pa which has conformed to the recommended pressure drop for wet scrubbers. From the analysis, it can be deduced that the numerical simulation using CFD is an effective method to study the flow characteristics of a counter-flow wet scrubber system.

References 1. 2. 3. 4.

5. 6. 7. 8.

9.

10. 11. 12.

Ashaka.: Cement PM Laden Data. Ashaka Cement Company, Nigeria (2011) Bedford, A.: The Lego Builder’s Guide. No Starch Press, San Francisco (2011) Box, E.T.: Frictional Coefficient [Online] Bozorgi, Y., Keshavarz, P., Taheri, M., Fathikaljahi, J.: Simulation of a spray scrubber performance with Eulerian/Lagrangian approach in the aerosol removing process. J. Hazard. Mater. 137(1), 509–517 (2006) Byeon, S.H., Lee, B.K., Mohan, B.R.: Removal of ammonia and particulate matter using a modified turbulent wet scrubbing system. Separat. Purif. Technol. 98, 221–229 (2012) Cengel, Y.A.: Fluid Mechanics. Tata McGraw-Hill Education (2010) Cooper, C.D., Alley, F.C.: Air Pollution Control: A Design Approach. Waveland Press (2002) Crowder, T.M., Rosati, J.A., Schroeter, J.D., Hickey, A.J., Martonen, T.B.: Fundamental effects of particle morphology on lung delivery: predictions of Stokes’ law and the particular relevance to dry powder inhaler formulation and development. Pharm. Res. 19(3), 239–245 (2002) Crowder, T.M., et al.: Fundamental effects of particle morphology on lung delivery: predictions of Stokes’ law and the particular relevance to dry powder inhaler formulation and development. Pharmaceutical research 19.3 : 239–245 (2002) DoE, U.: Energy information administration. Assumpt. Annual Energy Outlook 69, 2000 (2001) Ettema, R.: Hydraulic Modeling: Concepts and Practice. ASCE Publications (2000) Garba, M.N.: Gas particle separations using wet scrubber method. MEng, Mechanical Engineering Bayero University Kano, Nigeria (2005)

62

B. A. Danzomo et al.

13. Goniva, C., Tukovic, Z., Feilmayr, C., Bürgler, T., Pirker, S.: Simulation of offgas scrubbing by a combined Eulerian-Lagrangian model. In: Seventh International Conference on CFD in the Minerals and Process Industries, CSIRO, pp. 09–11. Melbourne, Australia (2009) 14. Hendawi, M., Molle, B., Folton, C., Granier, J.: Measurement accuracy analysis of sprinkler irrigation rainfall in relation to collector shape. J. Irrigat. Drain. Eng. 131(5) 477–483 (2005) 15. Hetsroni, G.: Handbook of multiphase systems (1982) 16. Jerry, W.C., Tim, K., Douglas, P.H., Douglas, Smith, T.: APTI: 413 control of particulate matter. In: Environmental Protection Agency (EPA) and the National Association of Clean Air Agencies (NACAA2013) (2013) 17. Karthik, T.S.D.: Turbulence models and their applications, MSc, Department of Mechanical Engineering, Indian Institute of Technology, Madras, India (2011) 18. Kennedy, C., Diwakar, P., Leonad, J.R., Rosendall, B.: Computation based engineering of multiphase process using CFD. Bechtel. Technol. J. 3 (2011) 19. Kim, H.T., et al.: Particle removal efficiency of gravitational wet scrubber considering diffusion, interception, and impaction. Environmental engineering science 18.2 : 125–136 (2001) 20. Krishnaraj, R., Sakthivel, M., Devadassan, S., Dinesh, M., Navaneethasanthakumar, S.: Investigation of sand filtration techniques to reduce secondary pollution in wet scrubber. Eur. J. Sci. Res. 89, 384–393 (2012) 21. Lee, C.C., Lin, S.: Handbook of Environmental Engineering Calculations. McGraw-Hill Professional (2000) 22. Lim, K.S., Lee, S.H., Park, H.S.: Prediction for particle removal efficiency of a reverse jet scrubber. J. Aerosol Sci. 37(12), 1826–1839 (2006) 23. Mahajan, S.P.: Air Pollution Control. Capital Publishing Company, New Delhi, India (2006) 24. Mohebbi, A., Taheri, M., Fathikaljahi, J., Talaie, M.R.: Simulation of an orifice scrubber performance based on Eulerian/Lagrangian method. J. Hazard. Mater. 100(1) 13–25 (2003) 25. Moody, L.F.: Friction factors for pipe flow. Trans. Asme 66(8), 671–684 (1944) 26. Moore, D., Copes, R., Fisk, R., Joy, R., Chan, K., Brauer, M.: Population health effects of air quality changes due to forest fires in British Columbia in 2003: estimates from physician-visit billing data. Canad. J. Public Health Revue Canadienne de Sante’e Publique, 105–108 (2006) 27. Munson, B.R., Young, D.F., Okiishi, T.H.: Fundamentals of Fluid Mechanics. New York (1990) 28. Ngala, G.M., Sulaiman, A.I., Sani, M.U.: Air pollution control in cement factory using horizontal type wet scrubber. Appl. Sci. 3, 1–9 (2008) 29. Palma, P.C., Lee, C.W., Simmons, K.: Comparison of computational image velocimetry data for the airflow in an aero-engine bearing chamber. J. Eng. Gas Turbines Power 127, 697–703 (2005) 30. Pieloth, D., Kohnen, B., Schaldach, G., Walzel, P.: CFD-Simulation von Nasswäschern Teil 1. Grundlagen und Implementierung in ein kommerzielles CFD-Programm. Chemie Ingenieur Technik 84(1–2) 127–137 (2012) 31. Product Corporation, H.: Air cooled heat exchanger modeling with computational fluid dynamics (2011) 32. Roberson, J.A., Crowe, C.T.: Engineering Fluid Dynamics. Wiley, NY, USA (1997) 33. Schnelle Jr, K.B., Brown, C.A.: Air Pollution Control Technology Handbook. CRC press (2001) 34. Sturm, R.G.: A study of the collapsing pressure of thin-walled cylinders. Univer. Illinois Bull. 39(12) (1941) 35. Xiang, R.B., Lee, K.W.: Numerical study of flow field in cyclones of different height. Chem. Eng. Process. 44(8), 877–883 (2005) 36. Yetilmezsoy, K., Saral, A.: Stochastic modeling approaches based on neural network and linear– nonlinear regression techniques for the determination of single droplet collection efficiency of countercurrent spray towers. Environ. Model. Assess. 12(1), 13–26 (2007) 37. Zhao, B.: Modeling of particle separation in bends of rectangular cross-section. Am. J. Appl. Sci. 2(1), 394–396 (2005)

Development of Intelligent Controller for Pollution Monitoring and Control Sambo A. Umar and Momoh Jimoh E. Salami

Abstract Air pollution, which arises from industrial wastes in form of Particulate Matter (PM) and ozone, poses serious threats to human and biotic environments. Several abatement measures have been taken to alleviate the effects of these pollutants of which wet scrubber system has been very successful as it has the ability to control both PM and gaseous pollutants. Recently some design approaches, in the form of similitude model design, computational fluids dynamics, and intelligent controller design, were suggested for improving the performance of wet scrubber system to enable it to control PM contaminants that are less than 5 µm in diameter. The results obtained from the simulation studies on the pilot system indicated a good control performance and that the intelligent controller design approach provided better strategy as compared to other approaches. However, the real-time implementation of the intelligent controller presents some challenging problems such as proper selection of hardware components and sampling rate. This chapter discusses the DSP-based implementation of the wet scrubber system in order to effectively and efficiently control the PM emission rate below the allowable rate. Results obtained from the hardware implementation of the intelligent controller show that the wet scrubber system has been able to maintain emission of the cement dust contaminant below the recommended value within 9.4 s, despite the disturbance caused by the dust particles.

S. A. Umar (B) Faculty of Engineering and Engineering Technology, Department of Mechatronics and Systems Engineering, Abubakar Tafawa Balewa University, PMB0248 Bauchi, Bauchi State, Nigeria e-mail: [email protected] M. J. E. Salami Department of Electrical and Electronics Engineering, Elizade University, Ilara Mokin, Ondo State, Nigeria © Springer Nature Switzerland AG 2022 M. Mariappan et al. (eds.), Control Engineering in Robotics and Industrial Automation, Studies in Systems, Decision and Control 371, https://doi.org/10.1007/978-3-030-74540-0_4

63

64

S. A. Umar and M. J. E. Salami

1 Introduction Air pollution is a serious threat to human health today. Air pollution is the presence of particulates, biological molecules, or other harmful materials in the Earth’s atmosphere, possibly causing diseases, death to humans, and damage to other living organisms [29]. Studies revealed that around 7 million people die annually (one in eight of total global deaths) as a result of air pollution [36]. These findings show that air pollution is now the world’s largest single environmental health risk. Reducing air pollution could therefore save millions of lives. Environmental concerned agencies such as Environmental Protection Agency (EPA) and World Health Organization (WHO) enforced laws on maximum PM emission limit [24, 25]. According to Brunekreef and van Bree [5], the two major air pollutants are particulate matter (PM) and ozone, mostly emitted from industrial wastes which result in several thousands of death. The Baseline Scenario’s report shows that, even in 2020, significant risks to human health will remain from PM and ozone exposures if further stringent measures are not taken [34]. Particles