Diagnostic reasoning based on structure and behavior
Artificial Intelligence - Special volume on qualitative reasoning about physical systems
Explorations in parallel distributed processing: a handbook of models, programs, and exercises
Explorations in parallel distributed processing: a handbook of models, programs, and exercises
IEA/AIE '88 Proceedings of the 1st international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 2
Learning internal representations by error propagation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
The logic of activation functions
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
IEA/AIE '88 Proceedings of the 1st international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 2
IEA/AIE '89 Proceedings of the 2nd international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 1
An object-oriented expert system for coal-fired MHD power plant fault monitoring and diagnosis
IEA/AIE '89 Proceedings of the 2nd international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 1
Integration of neural networks and expert systems for process fault diagnosis
IJCAI'91 Proceedings of the 12th international joint conference on Artificial intelligence - Volume 2
| Hi-index | 0.00 |
The detection and diagnosis of faults in real time are active areas of research in knowledge-based expert systems. Several methods of diagnosis have been applied to a variety of physical systems. Rule-based approaches have been applied successfully to some domains. However, encoding knowledge in rule bases raises many difficult knowledge acquisition issues; in addition, rule-based systems are often too slow to be effectively applied in a real-time environment. More advanced diagnostic systems may incorporate a simulation of the physical system in the knowledge base. Although simulation-based expert systems can exhibit powerful capabilities, simulating the domain properly may be difficult and too computationally intensive for real-time diagnosis.An effort is underway at The University of Tennessee Space Institute to develop diagnostic expert system methodologies based on the analysis of patterns of behavior of physical mechanisms. In this approach, fault diagnosis is conceptualized as the mapping or association of patterns of input data (e.g., from instrumentation) to patterns representing one or more fault conditions. Associative memories and neural networks are being investigated as a means of storing and retrieving fault scenarios, as they offer several powerful and useful features, including 1) general mapping capabilities, 2) resistance to noisy input data, 3) the ability to be trained in a supervised learning mode, and 4) the capability of operation with incomplete input.Pattern-based fault diagnosis and detection methodologies are currently being applied to jet and rocket engines. These domains are characterized by failure scenarios which may be catastrophic, and may occur over very short time periods. A requirement of the present study is that diagnoses be performed in real time, in order to allow time for effective action to be taken prior to possible engine destruction. This paper 1) outlines an architecture for a real-time pattern-based diagnostic expert system capable of accommodating noisy, incomplete, and possibly erroneous input data, and 2) presents results from prototype systems applied to jet and rocket engine fault diagnosis.