Crafting knowledge-based systems: expert systems made realistic
Crafting knowledge-based systems: expert systems made realistic
Pattern-based fault diagnosis using neural networks
IEA/AIE '88 Proceedings of the 1st international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 1
PISCES: an expert system for coal fired power plant monitoring and diagnostics
IEA/AIE '88 Proceedings of the 1st international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 1
A plant intelligent supervisory control expert system
ACM '86 Proceedings of 1986 ACM Fall joint computer conference
Hi-index | 0.00 |
Abnormal process behaviors observed through sensor data values are symptomatic of process faults. It has been demonstrated that pattern recognition techniques that associate faults with the symptoms they produce can be successfully applied in performing on-line automated fault detection and diagnosis [1-3]. This paper describes an object-oriented implementation of a knowledge-based expert system designed to aid power plant operators by performing automated sensor based fault detection and diagnosis. This object-oriented implementation replaces a rule-based backward chaining implementation of a Plant Intelligent Supervisory Control Expert System (PISCES) developed earlier at the University of Tennessee Space Institute [4-9]. All references to PISCES in this paper refer to the new object-oriented implementation. The earlier version will be referred to as a rule-based system.