Dehgan Haghighi - EPRC
Jahed Motlagh - IUST
Monsef

In this paper we have discussed the development of an expert system pattern for boiler fault diagnosis. This system is a model-rule-based expert system which used the possiblity theory. Fuzzy logic operators have been used, for representing certainty or uncertainty. Here we are faced with two main problems in the inference. The first one is searching through model network which is solved by using backward chaining strategy. The second is determining the remedial actions based upon the results which got at the first step. This step is done by using forward chaining strategy. Also, we used in the first step from Hill-climbing and Depth irst methods, and in the second step from list processing method and pattern of certainty factors which is done by using fuzzy logic.Although the developed expert system can be used in the off-line phase, with minimum changes in the source of program can be used also in on-line In this research work we have used from expert informations of Touss Power Plant control room. If it is used on-line , it will help power plant operators to do in optimum form, which fulfils vital role in the power plant system efficiency. This expert system is written in C++.