基于CQPSO的LCL滤波器设计方法

LCL filter design methods based on chaos quantum particle swarm optimization

针对光伏并网逆变系统中LCL滤波器参数设计的困难,提出采用混沌量子粒子群算法对滤波器参数进行寻优.利用混沌优化方法具有的随机性、有界性、遍历性等特性,来扩大算法的搜索范围和提高收敛速度.对LCL滤波器进行分析后,确定约束条件,结合算法来对其参数进行优化设计.对三相光伏并网逆变器进行仿真实验,将混沌量子粒子群算法与自适应遗传算法进行比较,结果表明该算法比自适应遗传算法有更好的滤波效果,谐波畸变率也更小.

Because the design of LCL filter of PV grid-connected inverter system was very difficult,chaotic quantum particle swarm optimization algorithm（CQPSO）was used to optimize the filter parameters in this paper.The chaos optimization method with the characteristics of randomness,boundedness and ergodicity could improve the search scope and convergence rate of the quantum particle swarm optimization.After the analysis of LCL filter,the parameters of LCL filter were designed and optimized by CQPSO with decided constraints.Compared with the adaptive genetic algorithm（AGA）,the results from simulation experiment of the three-phase photovoltaic grid-connected inverter showed that the CQPSO had a better filtering effect and smaller THD.