基于差分策略的多目标电力系统无功优化.pdfVIP

第 36 卷 第 12 期 电 网 技 术 Vol. 36 No. 12 2012 年 12 月 Power System Technology Dec. 2012 文章编号：1000-3673 （2012 ）12-0170-06 中图分类号：TM 72 文献标志码：A 学科代码：470·40 基于差分策略的多目标电力系统无功优化 李如琦，李芝荣，王维志，凌武能，唐林权 （广西大学电气工程学院，广西壮族自治区 南宁市 530004 ） Multi-Objective Reactive Power Optimization Based on Difference Strategy LI Ruqi, LI Zhirong, WANG Weizhi, LING Wuneng, TANG Linquan (School of Electrical Engineering, Guangxi University, Nanning 530004, Guangxi Zhuang Autonomous Region, China) ABSTRACT: In allusion to multi-objective reactive power 混合整数规划问题。目前，许多数学优化方法已经 optimization related to active network loss, voltage deviation 应用到求解无功优化问题上，传统的优化方法如： and voltage stability margin, a difference strategy-based 线性规划、二次规划、梯度法、牛顿法等[1-4]，常规 particle swarm optimization (PSO) algorithm is proposed to 算法求解无功优化问题时遇到的主要困难就是离 solve the multi-objective reactive power optimization problem. 散变量的归整问题和多峰多极值问题[5] 。为了解决 The difference strategy is introduced into updating operation of particle speed to avoid the algorithm falls into local optima; the 这些问题，研究人员逐渐把人工智能方法运用于无 objective series-based sorting matrix is utilized to evaluate the 功优化这一领域，其中包括免疫算法[6-8] 、混沌算 fitness of individuals, and according to the value of the fitness 法[8-9]、遗传算法[10-14]、粒子群算法[15]等。 the optimal guiding particle is chosen. In the proposed 粒子群优化算法(particle swarm optimization， algorithm the non-dominated sorting and crowding distance PSO)是美国 Kennedy 和 Eberhart 博士于 1995 年提 sorting are embedded in the search mechanism to implement the o