• No results found

Studies on some issues for design and planning of cellular manufacturing systems

N/A
N/A
Protected

Academic year: 2023

Share "Studies on some issues for design and planning of cellular manufacturing systems"

Copied!
12
0
0

Loading.... (view fulltext now)

Full text

(1)

STUDIES ON SOME ISSUES FOR DESIGN AND PLANNING OF CELLULAR

MANUFACTURING SYSTEMS

By

ALI ABDULMAJEED ALI

Department of Mechanical Engineering

Submitted

in the fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPIIV to the

INDIAN INSTITUTE OF TECHNOLOGY, DELHI

July 1998

(2)

S n the name of 4llaL

(

7Ze gracious, the (A4erciful

Dedicated to Late

MR. ABDULMAJEED ALI and DR. MOH'D ALI ALSHAHARY

gh

eir fond memories have always been a source of

S

nspiration and acted as riding spirit in this endeavour in pursuit

of excellence.

(3)

CERTIFICATE

This is to certify that the thesis entitled " STUDIES ON SOME ISSUES FOR DESIGN AND PLANNING OF CELLULAR MANUFACTURING SYSTEMS " being submitted by Mr. ALI ABDULMAJEED ALI AHMED to the Indian Institute of Technology, Delhi, for the award of the degree of ' Doctor of Philosophy' in the Mechanical Engineering Department, is a record of bonafide research work carried out by him.

Mr. ALI ABDULMAJEED ALI AHMED has worked under my guidance and supervision and has fulfilled the requirements for the submission of this thesis, which to my knowledge, has reached the requisite standard.

The results contained in this work have not been submitted in part or full, to any other University or Institute for the award of any degree.

July 20, 1998. ( Dr. PREM VRAT )

Professor Industrial Engineering and Management.

Mechanical Engineering Department Indian Institute of Technology, Delhi

Hauz Khas, New Delhi 110016

(4)

ACKNOWLEDGEMENT

The author would like to express his deep sense of gratitude to Dr. Prem Vrat, Professor, Industrial Engineering and Management, Department of Mechanical Engineering, Indian Institute of Technology Delhi, for his initiative, valuable guidance and supervision, keen interest and educative discussions throughout the course of this work. The author is indebted to his generous help and constant encouragement. His thinking and wisdom have had a permanent effect on my life. Thanks are due to Dr. S.G. Deshmukh, Associate professor, Industrial Engineering, Department of Mechanical Engineering, for his generous help and encouragement in every stage in all possible manner. His valuable advice is highly appreciated. Many thanks to Professor, R.S.

Agarwal, Head Mechanical Engineering Department, Professor, N.K. Tewari, Chairman Student Research Committee and all the members of SRC.

Many faculty members helped me directly or indirectly to complete my work and I thank them all. I would like to express my thanlullness to, Professor D.K. Banwet, Professor Arun Kanda, Professor U.R.K. Rao, Dr. S. Wadhwa, and Shri A.D. Gupta. Hearfful thanks to my best friends Hason Bin Hason Gubran for his kind help and Ayoob Ahmed Wali for his comradely.

Thanks are due to Professor Z. Michalewicz, and Professor A.L. Tits for providing the Genocop and FFSQP Software which are used in chapters 4 & 5.

I owe it to my wife and kids for their patience during my research work

Before all I owe it to my parents for their encouragement which gave courage and confidence to materialize my dreams. Lastly I wish to thank my friends Ashutosh Kumar, Amit Garg and Rajat Bahagwat and all others who helped me in completing this research.

I would like to extend my thanks to the Government of Yemen and Government of India in particular the Indian Council for Cultural Relations (ICCR) for their financial support.

Thanks are also due to staff Computer laboratory, Mechanical Engineering Department to extend their helping hands while working in the laboratory.

July 20, 1998. Ali Aitigiqc lAji Ahmed

(5)

ABSTRACT

Cellular manufacturing has been recognized as an effective model for improvement of system performance in the current manufacturing environment. The concept of cell can be considered as an attempt to obtain the advantages of flow line efficiency and job shop flexibility.

Cellular manufacturing system is an advanced manufacturing technology, provides greater responsiveness to changes in customer demand and higher degree of product mix variety. The main objective in the design of cellular manufacturing is the creation of mutually independent manufacturing cells, which is essentially the formations of part families and machine groups.

But practically, the creation of exclusive cells is not possible, due to the need of some parts to share several operations on machines in different cells.

In this thesis the problem of design and planning of cellular manufacturing systems has been addressed. Three efficient neural network based artificial intelligent approaches are developed for cell formation design. Due to their massive parallelism, high computation rate, retrieval of closest matching data even if there is no exact match to the requested information, neural network based approach has been adopted. The approaches are (i) the Hopfield neural network is used to solve the design problem presented by a generalized assignment model with the objective of minimizing the dissimilarity measure (Euclidean distance) between parts and between machines, (ii) the Quickprop neural network approach developed to solve the cells design problem, in which the machine part incidence matrix is an input to the network of three layers (input-hidden-output layers), where the number of rows (machines) represent the input patterns, the number of columns (parts) represent the number of units (nodes) in the input layer, and the required number of cells (predetermined) are represented by the output layer, and (iii) the cascade correlation neural network, in which the hidden layers with the number of units, generated dynamically depends on the problem complexity, for large scale problems with ill structured matrix, the number of hidden units is more than those if the problem is small and well structured matrix.

Real life factors such as processing time and production volumes are incorporated in a nonlinear mathematical model with the objective of minimizing the machine idle time. This is

iii

(6)

solved by developing a Genetic Algorithm based approach. Unlike the traditional mathematical approaches. Genetic algorithm evaluates the objective function directly with no need foi differentiability. It searches for the optimum variables in a population of points. Another salient feature of genetic algorithm is that, it codes the variables, uses probabilistic rules to move one population of solution to another.

Genetic algorithm approach is also used to solve the part tool grouping problem with the objective of maximizing the total processing time.

A pragmatic approach to the design of cellular manufacturing system is often driven by multiple conflicting non-linear objectives and set of multiple linear and non-linear constraints.

Here we develop a model of multiple objectives to examine the routing flexibility with the presence of alternative process plans. In addition to identifying simultaneously part families and machine groups in the cell formation problem, optimum process plan for each part has to be selected, quantity to be produced through the plans selected, machine type to perform each operation as per processing time and machine capacity constraints. The feasible sequential quadratic programming approach is used to solve this problem.

A comparative Evaluation of traditional and automated manufacturing systems is represented by traditional manufacturing system with conventional machines and with NC machines, and flexible manufacturing system on one hand and other traditional and automatic cellular manufacturing systems represented by, manned and unmanned cellular manufacturing system, virtual cellular manufacturing system, and flexible manufacturing cells and virtual flexible manufacturing cells on the other hand. Analytic Hierarchy Process (AHP) is used to rank on various alternative manufacturing systems due to its flexibility and ability to analyse the complex problem and its simplicity to implement. To reduce the computation time of ARP we proposed a simple relationship based on the scaling of the pairwise comparison matrix, which generates the comparison matrix given the first row (n-1) elements. The generated pairwise comparison judgement produces a very small inconsistency ratio (less than 0.02). AHP with the proposed relation can solve large industrial problems that have many attributes, alternatives with large number of hierarchy levels effectively.

(7)

CONTENTS

Certificate

Acknowledgement ii

Abstract iii

Contents

List of Tables xi

List of Figures xiv

CHAPTER 1: INTRODUCTION 1-20

1.0 INTRODUCTION 1

1.1 FEATURES OF MANUFACTURING SYSTEM 1

1.2 AUTOMATION IN MANUFACTURING 3

1.3 GROUP TECHNOLOGY 4

1.4 CELLULAR MANUFACTURING SYSTEM 5

1.5 ADVANTAGES OF GT / CMS 6

1.6 DISADVANTAGES OF GT / CMS 6

1.7 TRADITIONAL MANUFACTURING SYSTEMS 8 1.8 COMPARISON OF CELLULAR MANUFACTURING AND

TRADITIONAL MANUFACTURING SYSTEMS 10 1.9 FACTORS EFFECTING THE ADOPTION OF GT / CMS 11 1.10 SOME RELEVANT ISSUES RELATED TO CMS 13 1.11 CURRENT SCENARIO OF CMS RESEARCH 14

1.12 MOTIVATIONS 14

1.13 OBJECTIVES OF THIS RESEARCH 16 1.14 OVERVIEW OF THE RESEARCH WORK 17

(8)

1.1!; ORGANIZATION OF THE THESIS 18

CHAPTER 2: LITERATUTE REVIEW 21-58

2.0 INTRODUCTION 21

2.1 HISTORICAL REVIEW 21

2.2 CLASSIFICATION OF GT / CMS TECHNIQUES 22

2.2.1 Visual Inspection 22

2.2.2 Classification and Coding 25 2.2.3 Cluster Analysis Based Methods 26 22.3.1 Production Flow Analysis 26

2.2.3.2 Matrix Formulation 27

2.2.3.3 Graph Theoretic Approach 32 2.2.3.4 Artificial Intelligent Based Algorithms 33

2.2.3.5 Heuristics 38

22.3.6 Mathematical Programming 42

2.3 STATE OF THE ART 57

2.4 SUMMARY 58

CHAPTER 3: APPLICATION OF NEURAL NETWORK BASED

59-108 ALGORITHMS FOR i CELLFORMATTON PROBLEM

3.0 INTRODUCTION 59

3.1 NEURAL NETWORK 60

3.2 APPLICATION OF NEURAL NETWORK 62

3.3 BIOLOGICAL AND SIMULATED NEURONS 63

3.4 NEURAL NETWORK ARCHITECTURE 66

3.4.1 Network Learning Phase 68

(9)

3.4.2 Recall Phase 69

3.5 STOPPING CRITERIA 69

3.6 CELL FORMATION: A HOPFIELD NEURAL NETWORK 70 3.6.1 Formulation of the Problem 72 3.6.2 Modeling the Neural Network 76

3.6.3 Algorithm 77

3.6.4 Computational Experience 77

3.7 QUICK ALGORITHM FOR MANUFACTURING CELL DESIGN 82

3.7.1 Algorithm 82

3.7.2 Computational Results 87

3.7.3 Effect of Momentum Rate Factor 89 3.8 CELL FORMATION: A CASCADE CORRELATION NEURAL

NETWORK 97

3.8.1 Algorithm 97

3.8.2 Computational results 101

3.9 SUMMARY AND CONCLUSIONS 108

CHAPTER 4 : GENETIC ALGORITHM FOR CONCURRENT

MACHINE CELL PART FAMILY FORMATION 110-133

4.0 INTRODUCTION 110

4.1 INTRODUCTION TO GENETIC ALGORITHM 111 4.2 MECHANICS OF GENETIC ALGORITHM 114

4.3 MINIMIZING MACHINE IDLE TIME 117

4.4 MATHEMATICAL FORMULATION 119

4.4.1 Model I 120

4.5 ALGORITHM 121

vii

(10)

4.3.3 Computational Experience 123

4.3.4 Discussion 126

4.4 PART TOOL GROUPING PROBLEM 126

4.4.1 Model 11 127

4.4.2 Problem Formulation 127

4.4.3 Illustration 130

4.5 SUMMARY 133

CHAPTER 5: DESIGN AND PLANNING OF CMS :

A-MULTI-OBJECTIVE MODELLING USING FEASIBLE SEQUENTIAL QUADRATIC

PROGRAMMING (FSQP) 134-152

5.0 INTRODUCTION 134

5.1 MULTIPLE OBJECTIVES IN CMS 135

5.1.1 Maximize Machine Utilization 5.1.2 Minimize Inter-cell Movement 5.1.3 Minimize total processing cost

5.1.4 Minimize Total Inter-cell Workload Imbalance 5.2 PROBLEM FORMULATION

5.3 COMPUTATIONAL EXPERIENCE 5.4 SUMMARY

CHAPTER 6: COMPARATIVE EVALUATION OF ALTERNATIVE MANUFACTURING SYSTEM

6.0 INTRODUCTION

6.1 ANALYTIC HIERARCHY PROCESS

136 137 137 137 138 142 152

153-190 153 154

(11)

6.2 COMPUTATIONAL EFFORT REQUIRED FOR AHP 155

6.3 ALGORITHM OF AHP 156

6.4 IDENTIFICATION OF ALTERNATIVES 160 6.4. I Traditional Manufacturing System with Conventional Machines 160 6.4.2 Traditional Manufacturing System with NC Machines 161 6.4.3 Unmanned Cellular Manufacturing System 162 6.4.4 Manned Cellular Manufacturing System 163 6.4.5 Virtual Cellular Manufacturing System 164 6.4.6 Flexible Manufacturing System 165 6.4.7 Flexible Manufacturing Cells 166 6.4.8 Virtual Flexible Manufacturing Cells 166

6.5 PROBLEM FORMULATION 167

6.5.1 Hierarchical Structure of the Problem 167

6.5.2 Objectives 169

6.5.2.1 Minimization of total production cost 169 6.5.2.2 Maximization of Product mix variety 169 6.5.2.3 Maintain delivery date 171

6 5 2 4 Improve Quality 171

6.6 IDENTIFICATION OF CRITERIA 171

6.6.1 Investment 172

6.6.2 Total operation cost 172

6.6.3 Product mix 173

6.6.4 Adaptation to New Products 174

6.6.5 Flexibility 175

6.6.6 Material Handling 175

6.6.7 Throughput Time 176

ix

(12)

6.6.8 Fixtures 177

6.6.9 Quick Response 177

6.6.10 Labour 178

6.6.11 Quality Management 1 79

6.6.12 Scheduling and Planning 180

6.6.13 System Utilization 181

6.6.14 Reliability 181

6.7 IDENTIFICATION OF SUB-CRITERIA 182

6.8 ILLUSTRATION 188

6.5.1 Discussion 189

6.9 SUMMARY 190

CHAPTER 7: MAJOR CONCLUSIONS AND SCOPE FOR

FUTURE WORK 201-208

7.0 INTRODUCTION 201

7.1 SUMMARY OF THE WORK 201

7.2 MAJOR CONTRIBUTIONS REPORTED BY THIS RESEARCH 205

7.3 LIMITATION OF THE RESEARCH 207

7.4 SCOPE FOR FUTURE WORK 208

REFERENCES 209

Appendix A 240

Appendix B 242

References

Related documents

Angola Benin Burkina Faso Burundi Central African Republic Chad Comoros Democratic Republic of the Congo Djibouti Eritrea Ethiopia Gambia Guinea Guinea-Bissau Haiti Lesotho

The scan line algorithm which is based on the platform of calculating the coordinate of the line in the image and then finding the non background pixels in those lines and

3.6., which is a Smith Predictor based NCS (SPNCS). The plant model is considered in the minor feedback loop with a virtual time delay to compensate for networked induced

HAUZ KHAS, NEW DELHI -110016

Group scheduling is suitable for application in independent manufacturing cells; however, we assume that parts need different setup requirements. A heuristic has been proposed

Daystar Downloaded from www.worldscientific.com by INDIAN INSTITUTE OF ASTROPHYSICS BANGALORE on 02/02/21.. Re-use and distribution is strictly not permitted, except for Open

The matter has been reviewed by Pension Division and keeping in line with RBI instructions, it has been decided that all field offices may send the monthly BRS to banks in such a

The Education Cess on income-tax shall continue to be levied at the rate of two per cent on the amount of tax computed inclusive of surcharge. In addition, the amount of tax