一种改进的物元模型在变压器故障诊断中的应用

鉴于物元理论在构建诊断模型时,忽略了分界值的不确定性,使诊断结果偏离了实际情况,将物元理论与云模型相结合,建立了新的故障诊断模型,能合理地解决边界不确定问题。同时,使用油中溶解气体的浓度、产气率、气体间比值等数据,对所提模型进行了测试,得出基于物元理论和云模型的变压器故障诊断方法具有更高的正确率。

The application of improved matter-element model in power transformers fault diagnosis

Because the matter-element theory ignores the uncertainty of boundary values when setting up a fault diagnosis model,the diagnosing results are deviated from the actual situation.Considering that cloud model theory can be applied to represent the uncertainty of the boundary,the cloud model theory is combined with matter-element in the paper and a new diagnosing model is proposed.Meanwhile,the proposed diagnosing model is tested by using data such as the concentration of dissolved gases,gas production rate and the ratio between gases in oil.Compared with other techniques for transformer diagnosis,the new proposed model has a higher diagnosing precision.