Development of a knowledge based plant diagnosis and operation guidance system

A prototype system of the plant diagnosis and operation guidance system using knowledge engineering has been developed. The prototype system is one step toward the establishment of a man-machine communication system to support operators' work under abnormal plant conditions. The plant diagnosis and operation guidance system consists of an anomaly detection subsystem, a diagnosis subsystem, an operation guidance subsystem, a knowledge base, a numerical processing part, and man-machine interactive devices. A hierarchical structure composed from module programs, and a meta-knowledge representation of inference procedures are adopted to simplify corrections of the program and the inference procedures for diagnosis or operation guidance. Each module program is coded in a symbolic data form and converted into Lisp functions by a developed compiler to realize high speed inference. The prototype system was successfully tested for a simulated accident