粒子群改进算法在配电网故障定位中的应用

为了满足含分布式电源配电网故障定位的要求,对传统二进制粒子群算法进行改进,利用改进二进制粒子群算法(BPSO)解决配电网故障定位问题。改进BPSO初始化随机数采用均匀分布,同时引入收缩因子和线性变换的惯性权重来提升算法收敛于最优解的能力,避免陷入局部最优,提升故障定位的精确性。对算例配电网中的多种故障情形进行仿真分析,包含少量故障信息畸变的情况,诊断结论全部正确。仿真结果表明,改进算法在精确性和收敛速度上均优于传统粒子群算法,对含分布式电源的配电网故障定位具有一定的有效性和容错性。改进BPSO可以满足电网定位对准确和实时性的要求。

Application of improved binary particle swarm optimization for fault location in distribution network with distributed generation

This paper is focused specifically on seeking a solution to the fault location in distribution network by improving the traditional binary particle swarm optimization （BPSO） algorithm, as a part of our efforts to fulfill the requirement of fault location in distribution network with distributed generations. The improved BPSO algorithm operates by initializing random number using uniform distribution, introdu- cing the constriction factor and the inertia weight of linear transformation for improving algorithm＇ s ability to keep convergence on the optimal solution, and thereby improving the accuracy of fault location, thanks to the prevention of local optimum. The improved BPSO is validated by the ultimate diagnosis due to a simulation analysis of a variety of faults, including fault information distortion, occurring in distribution network. The simulation shows that the improved algorithm is superior to the conventional BPSO in terms of accuracy and convergence speed and thus demonstrates a greater effectiveness and fault tolerance for the distribution network fault location with distributed generations.