Paper Info
Title
Fault Detection and Diagnosis Using the Fuzzy Min-Max Neural Network with Rule Extraction
Abstract
In this paper, a symbolic rule learning and extraction algorithm is proposed for integration with the Fuzzy Min-Max neural network (FMM). With the rule extraction capability, the network is able to overcome the "black-box" phenomenon by justifying its predictions with fuzzy IF-THEN rules that are comprehensible to the domain users. To assess the applicability of the resulting network, a data set comprising real sensor measurements for detecting and diagnosing the heat transfer conditions of a Circulating Water (CW) system in a power generation plant is collected. The rules extracted from the network are found to be compatible with the domain knowledge as well as the opinions of domain experts who are involved in the maintenance of the CW system. Implication of the FMM neural network with the rule extraction capability as a useful fault detection and diagnosis tool is discussed.
Year
DOI
Venue
2004
10.1007/978-3-540-30134-9_48
Lecture Notes in Artificial Intelligence
Keywords
Field
DocType
heat transfer,neural network,domain knowledge,power generation
Data mining,Neuro-fuzzy,Domain knowledge,Computer science,Extraction algorithm,Fault detection and isolation,Fuzzy logic,Artificial intelligence,Adaptive neuro fuzzy inference system,Artificial neural network,Machine learning,Distributed computing
Conference
Volume
ISSN
Citations 
3215
0302-9743
2
PageRank 
References 
Authors
0.35
5
3
Name
Order
Citations
PageRank
Kok Yeng Chen1131.52
Chee Peng Lim21459122.04
Weng-kin Lai3335.49