2基于IQPSO-FCM的变压器故障诊断-电测与仪表.doc

基于改进量子粒子群优化模糊聚类的 变压器故障诊断方法* 王永强，李敏，谢军，律方成 （华北电力大学 河北省输变电设备安全防御重点实验室，河北 保定 071003） 摘要：对电力变压器进行高效准确的故障诊断可有效保障电力系统安全、稳定运行。为提高变压器故障诊断正确率，提出了一种基于改进量子粒子群优化模糊聚类的变压器故障诊断方法。采用遗传算法杂交概率的思想改进量子粒子群算法提高算法收敛速度、防止陷入局部极值，克服模糊聚类算法易受初始值影响的不足，进而实现对变压器高效、准确的故障诊断。以变压器油中典型气体作为故障特征量，选取68组数据建立故障集，采用改进量子粒子群算法寻找最佳初始聚类中心，并将其应用于3种不同数据组进行验证，实验结果表明本文所提方法的有效性。 关键词：电力变压器；改进QPSO-FCM；油中溶解气体；故障诊断 中图分类号：TM835.4 文献标识码：A 文章编号：1001-1390（2015）00-0000-00 Research on transformer fault diagnosis method based on improved QPSO-FCM algorithm Wang Yongqiang, Li Min, Xie Jun, Lv Fangcheng (Hebei Provincial Key Laboratory of Power Transmission Equipment Security Defense, North China Electric Power University, Baoding 071003, Hebei, China) Abstract：Having an efficient and accurate for power transformers fault is a key aspect for the safety and stability of the power system. To improve the accuracy rate of transformer fault diagnosis, a transformer fault diagnosis method based on improved quantum-behaved particle swarm optimization fuzzy clustering (IQPSO-FCM) is proposed. To overcome the shortcomings for sensitivity to initial value in fuzzy clustering method, the QPSO is improved by genetic hybrid algorithm to increase the convergence speed and prevent falling into local extremes. Thus, an efficient and rapid fault diagnosis result for transformers was obtained. In this article, the dissolved gas in oil was taken as the characteristic quantity of fault, the fault set was composed of 68 groups of data, and IQPSO was adopted to gain the optimal initial cluster centers for verifying the 3 different data sets. Experimental results show that the effectiveness of the proposed method. Keywords：power transformer, improved IQPSO-FCM, dissolved gas in oil, fault diagnosis 0　引 言 电力变压器是电力系统最重要的设备之一，对变压器故障实现准确判定与识别可有效保障电力系统的安全稳定运行[1-2]。提高变压器故障诊断的正确率具有重要意义。 目前，油中溶解气体分析（Dissolved gas analysis,DGA）的三比值法仍是最主要的故障诊断方法，但其在应用过程中，存在误判、错判等问题[3-5]。针对这些问题，专家系统[1]，支持向量机[2,6]，人工神经网络[7]，模糊聚类(Fuzzy Clustering Me