基于D-S证据理论的油浸式电力变压器故障诊断

Oil-immersed Power Transformer Faults Diagnosis Based on D-S's Theory of Evidence

当前对变压器进行故障诊断,主要是利用从分散的传感器收集到的检测信息,但使用单独的传感器采集数据进行判别,由于缺乏对数据的融合分析,易造成诊断中的误判或漏判。提出了一种基于D-S证据理论的油浸式电力变压器故障诊断方法,该方法通过融合数据来处理故障诊断中的不确定度。建立了故障诊断的融合模型,给出了方法的具体实现步骤并应用实例进行计算。结果表明,本方法可以减少诊断中的不确定性,提高故障诊断的准确性。

A novel method on oil-immersed power transformer faults diagnosis is suggested based on D-S Theory of Evidence, which can fuse the observation data and deal with the uncertainty in the decision. The basic concept of D-S Theory of Evidence and Dempster coalition rule are introduced briefly. A fault decision model is proposed. The detailed processes of the method and the experiment are given. The results show that it can reduce uncertainty efficiently and improve the accuracy of fault diagnosis.