基于扩展粒子群优化算法的同步发电机参数辨识

GENERATOR PARAMETER IDENTIFICATION BASED ON EXTENDED PARTICLE SWARM OPTIMIZATION METHOD

提出一种新的扩展粒子群优化(EPSO)算法并应用于同步发电机参数辨识。在粒子群优化(PSO)算法的基础上,EPSO算法采用更多粒子的位置值信息进行变异操作,并且提出根据各粒子的适应值大小确定算法控制参数的方法,保证了扩展后算法的收敛性,EPSO算法模型更具有通用性。仿真算例结果表明了在系统受到较大干扰的情况下,EPSO算法比EP算法和PSO算法具有更精确的参数综合辨识能力,并且EPSO算法比EP算法具有更高的收敛效率。

A new method of extended particle swarm optimization (EPSO) is presented and used to the generator parameter identification. Based on the original particle swarm optimization(PSO) method, EPSO method uses more particles' information to control the mutation operation and employs coefficients through the comparison of particles' fitness values, thus the convergence of the extended method is ensured. Under the condition of measurements of generator which are highly contaminated by noise, EPSO method possesses stronger capability of parameter identification than EP and PSO methods and has better integrated identification capability of parameters than EP method. Numerical simulation results demonstrate the conclusion.