基于曲线聚类-覆盖法的变电站无功优化配置  

Reactive Power Optimal Configuration of Substation Based on Curve Clustering-covering Method

作　　者：宋新新 赵吉祥 潘志远 赵义术 崔西友 王婧 SONG Xinxin;ZHAO Jixiang;PAN Zhiyuan;ZHAO Yishu;CUI Xiyou;WANG Jing(State Grid of China Technology College,Jinan 250000,China)

机构地区：[1]国家电网有限公司技术学院分公司,济南250000

出　　处：《电力系统及其自动化学报》2023年第3期151-158,共8页Proceedings of the CSU-EPSA

基　　金：国网公司科技项目(5108202040036A0000)。

摘　　要：针对电容器不等容分组设备频繁投切问题,提出一种基于曲线聚类-覆盖法的不等容分组优化配置方法。根据变电站年负荷曲线求取年无功需求,通过K-means聚类算法将365条年需求曲线聚类成k条聚类曲线,补偿总容量即为曲线最大值。首先以无功失配面积最小为目标函数,在聚类曲线上进行电容器档位阶梯状曲线的最优覆盖;然后建立以投切次数最小、失配面积最小、投资成本最小的多目标无功优化模型,利用改进遗传算法进行求解,用以指导变电站无功补偿优化规划;最后,通过与最优覆盖法及曲线分段-聚类法进行对比,验证所提方法的有效性。In view of the frequent switching problem of capacitor unequal capacity grouping equipment,an unequal capacity grouping optimal configuration method based on the curve clustering-covering method is proposed.The annual reactive power demand is calculated according to the annual load curve of the substation,365 annual demand curves are clustered into k clustering curves through the K-means clustering algorithm,and the total compensation capacity is the maximum value of the curve.The minimum reactive power mismatch area is taken as the objective function,and the optimal coverage of the stepped curve of capacitor gears is performed on the clustering curve.A multi-objective reactive power optimization model is established,with the smallest switching times,smallest mismatch area and lowest investment cost as its objectives.The improved genetic algorithm is used to solve the problem,thus guiding the optimal planning of reactive power compensation in the substation.Finally,the effectiveness of the proposed method is verified by comparing with the optimal coverage method and the curve segmentation-clustering method.

关 键 词：无功优化配置 电容器不等容分组 曲线聚类-覆盖法 无功失配面积 无功调节能力 

分 类 号：TM715[电气工程—电力系统及自动化]