PSO改进RBPNN在变压器故障诊断中的应用.pdfVIP

Vo l.44 No .l7 tJJ 靠拢保护岛拉4时 第44 卷第17 期 Power System Protection and Control Sep.l , 2016 2016 年9 月 1 日 DOI: 10.76671PSPC160081 PSO 改进RBPNN 在变压器故障诊断中的应用 施陶山 1 ，马宏忠 1 ，张琳1 ，李凯2 ，许洪华2 ，陈冰冰2 (1.河海大学能源与电气学院，江苏南京 211100; 2. 江苏省电力公司南京供电公司，江苏南京 210008) 摘要:针对概率神经网络σNN)及遗传算法(GA)在变压器内部故障诊断中存在的不足，提出了一种基于粒子群算 法(PSO)改进径向基概率神经网络(RBPNN)的故障诊断方法。首先，引入阻PNN，选取反向传播作为学习算法以 及泊中溶解气体含量比值作为故障特征量。然后，由于该模型受网络结构和初值影响较大，故拟用 GA、 PSO 和 改进的 PSO 对网络优化并削试。通过对比分析，得出改进的 PSO 在确定拓扑结构、降低误差精度、加快收敛速 度和提高预测准确度上更占优势的结论，同时证明了所提万法在故障诊断中的正确性和可行性。 关键词2 粒子群算注:径向基概率神经网络:反向传播:变压器:故障诊断 Application of RBPNN improved by PSO in fault diagnosis of transformers 1 1 1 2 2 2 SHI Xunshan , MAHongzhong , ZHANG Lin , LI Kai , XU Honghua , CHEN Bingbing (1. College ofEnergy and Electrical Engineering ,Hohai University,N叫ing 211100 , China; 2. Nanjing Power Supply Company, Jiangsu Electric Power Company,Nanjing 210008 , China) Abstract: Aiming at the existing deficiencies of probabilistic neural network (PNN) and genetic algorithm (GA) in internal faults of tr宿lSformers， a fault diagnosis method based on radial basis probabilistic neural network (RBPNN) improved by particle swarm optimization (PSO) is proposed. FirstIy, this paper introduces RBPNN and selects back-propagation as the learning algorithm as well as the content ratio of dissolved gases in oil as the characteristic quantity of fault. Then, since the network structure and the initial value have a great impact on RBPNN , this model is optimized and tested with GA, PSO and improved PSO. The comparison results show 由at improved PSO has more