基于小波神经网络的自适应单相重合闸

Adaptive single-phase reclosing based on wavelet neural network

自适应重合闸的首要问题是正确判别瞬时性故障和永久性故障,针对故障时丰富的暂态信息,提出了一种基于小波神经网络式单相自动重合闸的方案,通过小波变换算法提取故障暂态信号特征量后,再利用ANN的非线性拟合和自学习能力进行故障判别,确定正确的重合闸策略.通过EMTP及MATLAB进行了大量的仿真试验,验证了该方案的可行性,并解决了自适应单相重合闸的电压判据在判别瞬时性故障与永久性故障时存在的缺陷.

An adaptive single-phase reclosing scheme based on wavelet neural network is presented to overcome the fault of voltage criteria of adaptive single-phase reclosing in distinguishing the instantaneous faults from permanent fault. And a scheme based on three-layer wavelet neural network has been put forward. The simulated results of electric magic transient program(EMTP) and MATLAB show the feasibility of this method.