基于改进差分进化算法的微网容量优化配置  

Optimal Configuration of Microgrid Capacity Based on Improved Differential Evolution Algorithm

作　　者：张子恒 吴定会[1] 杨朝辉 马睿洁 ZHANG Zi-heng;WU Ding-hui;YANG Zhao-hui;MA Rui-jie(College of Internet of Things,Jiangnan University,Wuxi 241000,China;Wuxi Longma Technology Co.,Ltd,Wuxi 214028,China)

机构地区：[1]江南大学物联网工程学院,江苏无锡214122 [2]无锡隆玛科技股份有限公司,江苏无锡214028

出　　处：《控制工程》2023年第1期90-97,共8页Control Engineering of China

基　　金：国家重点研发计划(2020YFB1711102)。

摘　　要：针对风光柴储独立型微网的容量配置问题,综合考虑经济性、环保性和系统可靠性等要求,以总成本最低为目标,构建了容量配置模型,针对传统算法对模型求解精度不高、求解速度较慢的问题,提出了一种改进的差分进化算法。该算法在初始化种群阶段采用了反向学习策略以获得更优良的初始化种群,在变异阶段通过精英策略对种群进行划分,采用双变异策略和自适应变异参数对不同种群应用不同的变异操作,改进算法前期的全局搜索速度和后期的局部搜索精度。实验结果表明,将所提方法使用在微网容量优化配置上,可使成本下降4.9%,年缺电率控制在0.2%,算法表现良好。在更大规模的容量配置中,改进的差分进化算法有着更广阔的应用价值。Aiming at the capacity configuration of the independent microgrid of wind power,an improved differential evolution algorithm was proposed.In order to enhance the global search speed in the early stage and the local search precision in the late stage,the improved algorithm adopts the reverse learning strategy to obtain a better initial population at beginning.In the mutation process,the elite strategy,dual-variation strategy and adaptive parameters were used to divide the population and apply different operations to different populations,respectively.The experimental results show that the proposed method can reduce the cost by 4.9%and the annual power shortage rate by 0.2% when applied to the capacity optimization of microgrid.In the larger capacity configuration,the improved differential evolution algorithm has wider application value.

关 键 词：新能源微网 差分进化算法 容量配置 反向学习策略 

分 类 号：TM61[电气工程—电力系统及自动化]