基于改进微分进化算法的电力系统无功优化.pdfVIP

31 卷 18 期 电 网 技 术 Vol. 31No. 18 2007 年9 月 Power System Technology Sep. 2007 章编号：1000-3673 （2007 ）18-0068-05 中图分类号：TM734 文献标识码：A 学科代码：4704047 基于改进微分进化算法的电力系统无功优化 自发，闫景信，张建华，黄 伟，杨京燕 （电力系统保护与动态安全监控教育部重点实验室(华北电力大学) ，北京市 昌平区 102206 ） Power System Reactive Power Optimization Based on Improved Differential Evolution Algorithm LIU Zi-fa，YAN Jing-xin，ZHANG Jian-hua ，HUANG Wei，YANG Jing-yan （Key Laboratory of Power System Protection and Dynamic Security Monitoring and Control (North China Electric Power University) ，Ministry of Education，Changping District，Beijing 102206，China ） ABSTRACT: An improved differential evolution (IDE) 上，它是一个复杂的多目标、多约束、非线性、非 algorithm is proposed to solve power system reactive power 连续、混合整数优化问题。传统的数学优化方法， optimization. During the evolution, according to evolution 如线性规划、非线性规划、二次规划、混合整数法 condition in the proposed algorithm the dynamic parameter 等，各自都有一定的优越性和适应性，但是这些方 adjustment mechanism is adopted to enhance its search 法需要假设各控制变量是连续的，而且要求目标函 efficiency; the population overlap is monitored in real-time [1-3] 数可微，求解时间很长，易产生“维数灾”等问题 。 mode; for overlapped individual, by use of chaotic search 近年来随着计算机技术及人工智能技术的发展，产生 strategy the global optimal searching ability of the proposed algorithm is further improved. By means of reactive power 了遗传算法(genetic algorithm ，GA) 、粒子群算法 optimization of IEEE 6-bus, IEEE 30-bus and IEEE 118-bus (particle swarm optimization，PSO)、人工神经网络优 standard testing systems as well as that of a practical power 化算法(artificial neural networks，ANN) 、模拟退火算 system in a certain region, the results of calculation and 法(simulated ann