Design and Realization of Transformer Fault Diagnostic Expert System Based on Drools

This paper presents a transformer fault diagnostic expert system based on Drools to solve the issues as knowledge stereotyping, which commonly exists in the current transformer fault diagnostic system, resulting in the modification of stereotyping knowledge in actual application, and thus involving the modification of whole system, as well as the bad application effect, etc. First, Drools frame should be used to build transformer fault diagnostic expert system, so as to separate rule file and logic file. Second, combine the actual case and industrial authoritative expert experience, and start with four major types of faults as overheated fault, discharge fault, insulation fault and winding deformation, select the state quantity of mechanical property, electrical property and thermal property, which can best reflect the status changes of transformer, complete the preparation of rule base by using the knowledge rules table. Finally, develop transformer fault diagnostic expert system by using Java code, combining with rule base and Drools. Practice has proven that the system has faster reaction capacity and accurate diagnostic capability, so it has very good engineering promotional value.