RBF神经网络式自适应单相重合闸的应用研究

Study of Adaptive Single-Phase Reclosing Based on RBF Neural Network

为解决微机式自适应重合闸在判别瞬时性故障和永久性故障存在的缺陷 ,提出一种基于RBF神经网络的智能型单相自动重合闸的方案 ,将RBF神经网络运用于其中 ,建立一个三层前向网络模型。利用EMTP及MATLAB进行大量的仿真试验 ,验证了其可行性 ,克服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。

The adaptive single-phase reclosing scheme based on intelligent control is presented in order to overcome the problems of the adaptive single-phase reclosing in distinguishing the instantaneous faults from permanent ones. A scheme based on three-layer RBF neural network is put forward. The simulation results of EMTP and MATLAB indicate its feasibility.