基于DGA的QPSO-BP模型变压器故障诊断方法研究  

Dissolved Gas Analysis Based QPSO-BP Model for Transformer Fault Diagnosis

作　　者：程加堂[1] 段志梅[1] 熊燕[1] 艾莉[1] 

机构地区：[1]红河学院工学院,云南蒙自661199

出　　处：《高压电器》2016年第2期57-61,共5页High Voltage Apparatus

基　　金：云南省教育厅科研基金资助项目(2012Y450)~~

摘　　要：为了提高变压器故障诊断的准确率,提出一种基于量子粒子群优化BP神经网络(quantum particle swarm optimized BP neural network,QPSO-BP)的故障诊断模型。在该算法中,用量子位的概率幅表示种群中各粒子的当前位置,用量子旋转门实现粒子位置的移动,用量子非门进行变异操作,以获取BP神经网络的权、阈值优化参数,最终实现了变压器故障诊断模型的构建。对故障DGA样本的诊断实例表明,与粒子群优化BP网络(particle swarm optimized BP neural network,PSO-BP)法、BPNN法以及IEC三比值法相比,QPSO-BP算法具有更高的诊断正确率,从而实现了变压器故障模式的有效识别。In order to improve the accuracy of transformer fault diagnosis, a model is proposed based on quantum particle swarm optimized BP neural network（QPSO-BP）. In this algorithm, the current position of the particle is achieved with probability amplitude of quantum bits, quantum rotating gate is employed in moving position, then the mutation operation is performed by quantum non-gate, in order to acquire the optimization parameters of weights and thresholds of BP neural network, ultimately the transformer fault diagnosis model is established. The simulation results of DGA samples show that the QPSO-BP algorithm has higher diagnosis accuracy compared with particle swarm optimized BP neural network（PSO-BP）, BP neural network and IEC three ratio, and achieves the effective recognition of transformer failure mode.

关 键 词：量子粒子群算法 神经网络 变压器 故障诊断 溶解气体分析 

分 类 号：TM407[电气工程—电器]