2023.6Vol.47No.6研究与设计收稿日期:2022-11-26基金项目:吉林省教育厅项目(2018103)作者简介:马志强(1998—),男,山西省人,硕士,主要研究方向为光伏发电与并网控制。通信作者:张建民基于改进花授粉算法的光伏MPPT研究马志强,张建民(北华大学电气与信息工程学院,吉林吉林132013)摘要:局部遮阴情况下,光伏阵列的P-V曲线存在多个极值,传统最大功率跟踪方法容易陷入局部最优而追踪失败。针对多峰值P-V曲线,提出一种改进型花授粉算法来实现全局最大功率跟踪。该算法在标准花授粉算法的基础上引入了自适应转化概率来平衡全局和局部搜索,同时用Jaya算法的位置更新公式替代原算法的局部搜索公式,有效避免了原局部搜索策略易陷入局部最优的不足。通过MATLAB/Simulink进行仿真验证,并与粒子群算法、花授粉算法进行比较,结果表明改进花授粉算法具有更高的追踪精度和更快的追踪速度。关键词:局部最优;最大功率跟踪;自适应转化概率;Jaya算法;改进花授粉算法中图分类号:TM615文献标识码:A文章编号:1002-087X(2023)06-0795-05DOI:10.3969/j.issn.1002-087X.2023.06.023PhotovoltaicMPPTresearchbasedonimprovedflowerpollinationalgorithmMAZhiqiang,ZHANGJianmin(CollegeofElectricalandInformationEngineering,BeihuaUniversity,JilinJilin132013,China)Abstract:Inthecaseofpartialshading,theP-Vcurveofthephotovoltaicarrayhasmultipleextremevalues,andthetraditionalmaximumpowertrackingmethodsareeasytofallintothelocaloptimumandthetrackingfails.Aimingatthemulti-peakP-Vcurve,animprovedflowerpollinationalgorithmwasproposedtoachieveglobalmaximumpowertracking.Basedonthestandardflowerpollinationalgorithm,theadaptivetransformationprobabilitywasintroducedtobalancetheglobalandlocalsearch,andthelocalsearchformulaoftheoriginalalgorithmwasreplacedwiththepositionupdateformulaoftheJayaalgorithm,effectivelyavoidingthedisadvantagethattheoriginallocalsearchstrategyiseasytofallintothelocaloptimum.ThesimulationverificationwascarriedoutthroughMATLAB/Simulink,andthealgorithmwascomparedwiththeflowerpollinationalgorithmandparticleswarmoptimization.Theresultsshowthattheimprovedflowerpollinationalgorithmhashighertrackingaccuracyandfastertrackingspeed.Keywords:localoptimum;maxim...