基于LF-ATSO算法在光伏系统MPPT中的研究  

Research on photovoltaic system MPPT based on LF-ATSO algorithm

作　　者：李嘉轩 于惠钧[1] 马凡烁 刘紫英 LI Jiaxuan;YU Huijun;MA Fanshuo;LIU Ziying(College of Electrical and Information Engineering,Hunan University of Technology,Zhuzhou 412007,China)

机构地区：[1]湖南工业大学电气与信息工程学院,湖南株洲412007

出　　处：《现代电子技术》2024年第21期149-155,共7页Modern Electronics Technique

基　　金：湖南省自然科学基金项目(2021JJ50052);湖南省教育厅科学研究项目(21A0354)。

摘　　要：在复杂遮荫工况下,传统MPPT策略在面对多峰值现象时易陷入局部最大功率点(LMPP),而基于元启发式算法的最大功率点跟踪策略也存在寻优精度不高、追踪时间慢等问题。为解决上述问题,文中构建Lévy飞行-自适应金枪鱼群算法(LF-ATSO)。首先,采用基于Circle混沌映射的反向学习策略合理分配初始化种群以提高种群遍历性;其次,改进参数a用以调整最优个体和前一个体的比重,提高收敛速度;然后,嵌入Lévy flight策略提高算法全局搜索能力,帮助其跳出局部最优;最后,加入算法重启机制以应对复杂变化工况。将改进后的TSO算法与未改进TSO算法、PSO算法、改进GWO算法进行仿真对比,实验结果表明,改进后的TSO算法在静态、动态复杂遮荫工况下均能够更快、更精准地追踪到全局最大功率点(GMPP)。In complex shading conditions,the traditional MPPT(maximum power point tracking)strategy is prone to falling into the local maximum power point(LMPP)in the face of multiple peaks.In addition,the MPPT strategy based on meta-heuristic algorithm has deficiencies,such as low optimizing accuracy and slow tracking time.In view of the above,a Lévy flight-adaptive tuna swarm optimization(LF-ATSO)algorithm is constructed.Firstly,the reverse learning strategy based on Circle chaotic mapping is used to allocate the initialized population reasonably to improve the population ergodicity.Secondly,the improved parameter a is used to adjust the proportion of the optimal individual to the previous individual,so as to improve the convergence speed.Then,the Lévy flight strategy is embedded to improve the global search ability of the algorithm to help the algorithm get rid of the local optimum.Finally,the algorithm restart mechanism is added to cope with the complex shading conditions.The improved TSO algorithm is simulated and compared with the unimproved TSO algorithm,PSO(particle swarm optimization)algorithm and improved GWO(grey wolf optimization)algorithm.The improved TSO algorithm demonstrates superior performance in swiftly and precisely tracking the global maximum power point(GMPP)under both static and dynamic complex changing conditions,as indicated by the experimental findings.

关 键 词：光伏系统 最大功率点跟踪 局部遮荫 金枪鱼群算法 Lévy flight策略 Circle混沌映射 

分 类 号：TN209-34[电子电信—物理电子学]