基于无编码比值法的天然酯绝缘油变压器故障诊断方法研究

Research on Diagnosis Method of Natural Ester Insulating Oil Transformer Based on Non-code Ratio Method

本文以山茶籽绝缘油中溶解气体数据和特性为例,建立了基于无编码比值法的天然酯绝缘油变压器故障诊断方法。首先通过采用相关系数矩阵分析方法对H_(2)、CH_(4)、C_(2)H_(4)、C_(2)H_(6)和C_(2)H_(2)五种特征气体构成的10组特征气体比例关系进行降维优化处理,提出不具有显著相关性的5组特征量。然后采用模糊C均值聚类分析(FCM)算法提取能够区分各种故障类型的最优特征量,并根据油中溶解气体数据确定故障诊断边界条件。结果表明,本文提出的无编码比值法准确率为88%,能够较为准确的区分和诊断天然酯绝缘油变压器故障。

In this paper,a fault diagnosis method for natural ester insulating oil transformers based on non-coded ratio method was established,taking dissolved gas data and characteristics of camellia insulating oil as an example.Firstly,dimension reduction optimization treatment was made to analyzing 10groups of proportion which consists of five kinds of characteristic gases(H_(2),CH_(4),C_(2)H_(4),C_(2)H_(6)and C_(2)H_(2))by using correlation coefficient matrix method,and five groups of characteristic parameters with no significant correlation were proposed.Then optimal characteristic parameters which could distinguish fault types were extracted by using fuzzy c_(m) eans clustering analysis(FCM)algorithm.And boundary conditions for fault diagnosis were determined according to the dissolved gas data.The results showed that the accuracy of non_(c) ode ratio method proposed in this paper was 88%,which can distinguish and diagnose the faults of natural ester insulated oil transformer accurately.