基于粗糙集理论的变压器故障诊断专家系统

在传统的变压器故障诊断专家系统的基础上,为解决专家系统较难获取完备知识的瓶颈问题,提出了一种基于粗糙集理论的故障诊断专家系统。该系统从历史故障数据所形成的决策表出发,运用粗糙集理论进行约简,构建专家系统知识库模型,通过计算诊断规则粗糙度,确定诊断规则的置信程度,利用推理机和故障事例库,实现对知识库的动态维护。

Expert System for Transformer Faults Diagnosis Based on Rough Set Theory

This paper puts forward a fault diagnosis expert system based on rough set theory,in order to solve the bottleneck problem in the expert system,which is based on traditional transformer fault diagnosis,and therefore is difficult in acquiring complete knowledge.The system starts from the policy sheet formed by historical fault data,and reduced by rough set theory,then construct the model for expert system knowledge.After that it implements dynamical maintenance of the knowledge database by calculating diagnosis rule roughness,by determining the confidence degree and by using the reasoning mechanism and fault cases database.