基于三对比值的变压器故障诊断识别

Transformer Fault Diagnosis and ldentification Based on Three Contrast Values

变压器作为电力系统发、输、变、配电环节重要的电气设备,其故障能否被快速、精准的诊断对电力系统稳定运行起到至关重要的作用。目前溶解于变压器的气体的三对比值法(DGA)被广泛应用于变压器内部故障诊断。本文利用三比值法对变压器进行诊断,在广泛收集种变压器故障的基础上,首先利用ReliefF算法对三比值特征的重要性进行评价,并利用KNN分类器进行故障识别。实验结果表明,该方法可以有效的识别变压器故障类型。

Transformer is important electrical equipment in power system generation,transmission,transformation,and distribution.Whether its fault can be diagnosed quickly and accurately plays a vital role in the stable operation of the power system.At present,the three-contrast value method(DGA)of gases dissolved in transformers is widely used in transformer internal fault diagnosis.This paper uses the three-ratio method to diagnose the transformer.Based on the extensive collection of five types of transformer fauts,the ReliefF algrithm is first used to evaluate the importance of the three-ratio feature,and the KNN classfier is used for fault identification.Experimental results show that this method can eftectively identify the type of transformer fault.