【摘要】 在分析粒子群参数特征的基础上,提出自适应粒子群优化算法,使用自适应粒子群优化BP神经网络,建立基于自适应粒子群优化BP神经网络(PSO-BP)的变压器故障诊断系统.通过对52组训练样本和28组测试样本的仿真实验,可知自适应PSO-BP法能提高变压器故障诊断的准确率,有效减小网络的误差精度.更多还原

【Abstract】 Based on the analysis of the particle swarm parameter characteristic,adaptive particle swarm optimization algorithm is put forward.Using adaptive particle swarm optimization back propagation(PSO-BP) neural network,transformer fault diagnosis system is built up based on adaptive particle swarm optimization BP neural network.By the simulation experiment using 52 groups of training samples and 28 groups of test samples,it can be seen that the adaptive PSO-BP method can improve the transformer fault diagnosis accuracy and reduce the network error precision effectively.更多还原