The Research About Fault Diagnosis Of Boiler Based On Fuzzy Expert System

With the rapid development of industrialization,boilers are more and more widely used in many fields. The extent of system is lager and larger, and structure is maor and more complex. So the owner of boiler has to pay more attention to the daily maintenance and troubleshooting.To make sure the boiler running without trouble,not only can improve the utilization ratio of fuel,but also can reduce the cost of operation.It is even an guarantee for the staff of boiler.Expert system is becoming a main mode to do fault diagnosis because of its easy-operation and flexibility. But because of the problem of knowledge getting, traditional expert system is often simple and flexibility,the rate of accuracy is usually very low. In this paper, a new way of knowledge representation is proposed after analyzing the structure of boiler. The new way is made of fuzzy assertions, weight and credibility.In this paper, the system is designed to realize the system of knowledge representation, storage, and the integration of reasoning.By the tool of VC++6.0 ,CLIPS and SQL, the software can realize the knowledge getting, storage and automatic diagnosis. In the paper, the fuzzy assertions and weight are introjected to store by the way of .CLP,by the combination of neural network and fuzzy theory,the reasoning mechanism can solve the fuzziness of the boiler failure correlation, etc. In the system design in VC + + firstly written several signs of tools for signs of credibility and discriminant functions, develops the membership functions of the fault diagnosis of fuzzy expert system and modify the knowledge module and fuzzy reasoning module. To finish the CLIPS tools and knowledge repository of modification and update maintenance management. Examples of the diagnosis suggest verifies the contrast experiment, the system can accurately to common fault diagnosis in incomplete information, and fault symptoms, inductive aspects showed obviously superiority.