摘要
利用暂态故障信号实现配电网的故障定位是当前的一个研究热点,对配电网的快速恢复起着至关重要的作用。本文研究了小波神经网络的原理,针对小波神经网络容易陷入局部最优且收敛速度慢的缺点,利用粒子群算法对其进行改进,给出了具体的小波神经网络特征提取过程,以及粒子适应度函数与权值的计算。将改进算法用于配电网故障定位,结果表明,改进算法的定位精度明显优于小波神经网络的定位精度,为实际电网故障定位系统的设计提供重要参考。
Using the transient fault signal to realize the power grid fault location is a hotspot,which plays an important role in rapid recovery of power grid.This paper studies the principle of the wavelet neural network.Wavelet neural network is easy to fall into local optimum and has the disadvantage of slow convergence rate,which is improved by the use of particle swarm algorithm.Wavelet neural network feature extraction process is given,as well as the particle fitness function and weight calculation is worked out.The improved algorithm is applied to fault location of distribution network,the results show that the accuracy of improved algorithm is obviously superior to accuracy of wavelet neural network.It can provide important reference for the actual power grid fault positioning system.
出处
《科技通报》
北大核心
2013年第6期59-61,共3页
Bulletin of Science and Technology
关键词
故障定位
小波神经网络
粒子群
定位精度
fault location
wavelet neural network
particle swarm
positioning accuracy
