Intelligent Condition Monitoring and Diagnosis System: A Computational Intelligent Approach

by K. Wang

Overview

Human beings used to use their own senses and experience to monitor the condition of machinery, but they have a habit of forming labor unions, so a substitute is being sought in computer intelligence. Wang (production and quality engineering, Norwegian U. of Science and Technology), describes one approach to using advanced instrumentation to quantify the health or condition of industrial equipment, so that problems can be diagnosed early in their development and corrected by suitable maintenance, before they become serious enough to cause plant breakdown. His example is an intelligent monitoring system for a centrifugal pump system using multiple layered feed forward artificial neural networks, which effectively avoids the necessity for manually setting up accurate mathematical descriptions. Annotation ©2004 Book News, Inc., Portland, OR

Product Details

ISBN-13:	9781586033125
Publisher:	IOS Press, Incorporated
Publication date:	04/28/2003
Series:	Frontiers in Artificial Intelligence and Applications Series
Pages:	122
Product dimensions:	6.30(w) x 9.50(h) x 0.40(d)

Table of Contents

	Preface	v
Chapter 1	Introduction	1
1.1	Condition monitoring and diagnosis	1
1.2	Computational Intelligence	3
1.2.1	Expert Systems	4
1.2.2	Artificial Neural Networks	4
1.2.3	Fuzzy Logic Systems	5
1.2.4	Genetic Algorithms	5
1.2.5	Hybrid CI systems (HCIS)	6
1.3	Aims and objectives	7
	References	10
Chapter 2	Artificial Neural Networks	11
2.1	A brief introduction to the Artificial Neural Networks	11
2.1.1	Basic concepts	12
2.1.2	Network learning strategies, classifications and characteristics	15
2.2	Back propagation neural network	18
2.3	Associative memory networks	21
2.3.1	Structure	21
2.3.2	Learning methods	22
2.4	Other commonly used networks	22
2.4.1	Hopfield network and Boltzmann machine	22
2.4.2	ART network	24
2.4.3	Kohonen network	24
2.5	Practical considerations of using neural networks	25
2.6	Summary	26
	References	27
Chapter 3	Using ANNs to Monitor Mechanical Equipment	29
3.1	A brief introduction to condition monitoring	29
3.1.1	Basic concepts	29
3.1.2	Condition monitoring techniques	30
3.1.3	Transducers	31
3.1.4	Vibration analysis for rotational equipment	32
3.1.5	Common machine faults detectable by vibration	34
3.2	Using ANN to monitor centrifugal pump	37
3.2.1	Experimental set-up	38
3.2.2	Data collection	40
3.2.3	Feature selection and transformation	43
3.2.4	Using multiple-layered feed forward ANN to monitor centrifugal pump	45
3.3	Summary	50
	References	51
Chapter 4	Genetic Algorithms	53
4.1	Introduction to GAs	53
4.1.1	A brief introduction to GAs	53
4.1.2	Schema theorem	55
4.2	Using GAs to select ANN structure	56
4.2.1	Chromosomes and fitness function	57
4.2.2	Selection	58
4.2.3	Crossover	58
4.2.4	Mutation	59
4.2.5	Inversion	59
4.3	Implementation	60
4.3.1	Experimental results and discussions	62
4.4	Summary	62
	References	63
Chapter 5	Using Fuzzy Logic Systems for Monitoring and Diagnosing	65
5.1	Introduction	65
5.2	Classical logic and fuzzy logic	66
5.2.1	Fuzzy sets	67
5.2.2	Membership function	67
5.2.3	Fuzzy operations	68
5.2.4	Defuzzification	70
5.3	Using fuzzy logic to monitor seal ring problem	70
5.4	Using B-Spline network to extract fuzzy relationships	76
5.4.1	B-Spline network	77
5.4.2	Learning method	81
5.4.3	Extracting fuzzy relationships	81
5.5	Summary	84
	References	84
Chapter 6	Expert Systems for Maintenance Management	85
6.1	The role of Expert Systems in maintenance	85
6.2	Expert Systems	85
6.2.1	A brief introduction to Expert Systems	85
6.2.2	Characteristics	86
6.3	Basic structures of an Expert System	89
6.4	Using rule-based Expert Systems to diagnose pump fault	91
6.4.1	Fault phenomenon, causes and location	92
6.4.2	Inference mechanism	95
6.4.3	Development tool - Knowledge Pro	97
6.5	Summary	101
	References	102
Chapter 7	An Architecture of Intelligent Monitoring and Diagnosis Systems	103
7.1	A generic structure	103
7.1.1	The sensor array	103
7.1.2	The signal processing and feature extraction sub-system	104
7.1.3	The monitoring and diagnosis sub-system	105
7.2	Intelligent hybrid monitoring and diagnosis system	107
7.3	Conclusions and future works	114
7.3.1	Summary	114
7.3.2	Future works	116
	References	118
	Index	119