Design and Implementation of Inference Engine in Safety Risk Assessment Expert System in Petrochemical Industry Based on Fault Tree

The project in petrochemical industry is complex and risky. For this feature, we established a safety risk assessment (SRA) expert system based on fault tree in petrochemical industry. In this paper, we studied the design and implementation of infer engine in SRA expert system. we adopted the method of fault tree analysis (FTA) to acquire expert knowledge and the fault tree established is to be the basis of inference. The knowledge in petrochemical industry was divided into shallow knowledge and deep knowledge and the method of KR(knowledge representation) adopted in this paper is production rule combined with the framework. On the basis of good representation and organization of knowledge, we adopted infer control strategy of forward reason combined with depth-first search which improve the validity and accuracy of the SRA expert system to a certain extent.