基于改进的粗糙集理论与贝叶斯网络的变压器设备故障诊断模型

Fault diagnosis model of transformer equipment based on improved rough set theory and Bayesian network

电力变压器是电力系统最为关键的一部分,快速准确地判断变压器故障类型对变压器安全运行具有重要作用,传统的粗糙集理论进行故障诊断往往存在两个问题,一是忽视属于NP-hard问题的最佳属性约简;二是未充分考虑故障诊断的概率属性。基于此,提出了一种基于改进的粗糙集理论与贝叶斯网络的变压器设备故障诊断模型,通过构建可分辨二进制矩阵,引入启发式属性约简算法求解最佳属性约简,并基于最佳属性约简构建最小诊断规则,用贝叶斯网络完成概率建模实现变压器故障诊断。通过案例分析,该模型能有效地判断变压器设备故障类型,且与油色谱三比值法相比,准确率有所提升。

Power transformer is one of the most important equipment in the power system.It is very important to judge the type of transformer fault quickly and accurately for the safe operation of transformer.There are two problems in the traditional rough set theory for fault diagnosis,one is to solve the problem of optimal attribute reduction which belongs to NP hard,the other is to not fully consider the probability attribute of fault diagnosis.In this paper,a fault diagnosis model of transformer equipment based on improved rough set theory and Bayesian network is proposed.By constructing recognizable binary matrix,heuristic attribute reduction algorithm is introduced to solve the optimal attribute reduction,and the minimum diagnosis rule is constructed based on the optimal attribute reduction,and probability reasoning is realized by Bayesian network to realize the transformer fault diagnosis.through case analysis,Compared with the three ratio method of oil chromatogram,the accuracy of the model is improved.