引入竞争机制的改进AR谱估计间谐波分析方法

Inter-harmonic Analysis Method of Improved AR Spectral Estimation by Introducing Competitive Mechanism

随着我国各种大功率电力电子负载的广泛应用,导致了公用电网谐波污染日益严重,使得对应的电能质量问题受到越来越多的关注,因此准确分析间谐波特征对提高电能质量具有重要意义。提出了一种具有竞争机制的改进混合遗传粒子群算法(HGAPSO),通过引入一系列淘汰及精英学习的竞争策略使粒子群算法(PSO)具有更快的搜索速度以及更强的全局搜索能力,从而显著提高自回归(AR)谱估计中的参数估计精度。通过AR谱估计获取检测信号中的间谐波阶数及频率初值,并基于间谐波信号模型对谐波进行转换,将转换后的幅值编码于改进HGAPSO中对间谐波的幅值和相位进行参数估计。仿真结果表明,该算法较HGAPSO,PSO和遗传算法(GA)能够获得较高精度的间谐波参数值,同时具有更强的抗干扰能力。

With the wide application of various high-power power electronic loads in China,the harmonic pollution of public power grids is becoming more and more serious,and the corresponding power quality problems have attracted more and more attention.Therefore,it is of great significance to improve power quality by accurate analysis of inter-harmonic characteristics.An improved hybrid genetic algorithm and particle swarm optimization(HGAPSO)with a competitive mechanism was proposed.By introducing a series of competitive strategies of elimination and elite learning,the particle swarm optimization(PSO)has a faster search speed and stronger global search capability,thereby the accuracy of parameter estimation in autoregressive(AR)spectral estimation is significantly improved.The AR spectral estimation was used to obtain the inter-harmonic order and initial frequency in the detection signal,and the harmonics were converted based on the inter-harmonic signal model.The converted amplitude was encoded in the improved HGAPSO to estimate the amplitude and phase of inter-harmonics.The simulation results indicates that the algorithm can get higher precision inter-harmonic parameters compared with HGAPSO,PSO and genetic algorithm(GA),and has a stronger anti-interference ability.