基于BQPSO的电网多目标优化规划-计算机工程与应用.PDF

266 2015 ，51（18） Computer Engineering and Applications 计算机工程与应用 基于BQPSO 的电网多目标优化规划 张明光，王世亮 ZHANG Mingguang, WANG Shiliang 兰州理工大学 电气工程与信息工程学院，兰州 730050 College of Electrical and Information Engineering, Lanzhou University of Technology, Lanzhou 730050, China ZHANG Mingguang, WANG Shiliang. Multi-objective optimization of network planning based on BQPSO. Com- puter Engineering and Applications, 2015, 51 （18）：266-270. Abstract ：For the multi-objective optimization problem of power network planning, the model is established based on economy and reliability as the goal, through optimizing the quantum particle swarm algorithm of binary code. In order to improve the diversity and distribution of the optimal solution, the optimal sorting method of crowding distance in external memory solutions for the renewal and the maintenance, makes it find as many Pareto optimal solutions as possible. The simulation results show that the optimal binary quantum particle based on crowding distance sorting algorithm for group has a better distribution and convergence than other intelligent algorithm. Key words ：multi-objective optimization; Quantum Particle Swarm Optimization （QPSO ）; binary code; crowding distance; Pareto optimal solution 摘 要：对于多目标电网优化规划问题，建立以经济性和可靠性为目标的电网规划模型，通过二进制编码的量子粒 子群算法进行优化。为了提高最优解的多样性和分布性，采用拥挤距离排序的方法对外部存储器中的最优解进行 更新和维护，使得算法找到尽可能多的Pareto 最优解。仿真结果显示，基于拥挤距离排序的二进制量子粒子群算法 比其他智能算法寻得的最优解有更好的分布性和收敛性。 关键词：多目标优化；量子粒子群算法；二进制编码；拥挤距离；Pareto 最优解 文献标志码：A 中图分类号：TP39 doi ：10.3778/j.issn. 1002-8331.1310-0074 [3] 1 引言 Algorithm ，MOGA ）、非支配遗传算法（Non-dominated [4] 电网规划是指在满足各项约束的前提下确定在何 Sorting Genetic Algorithm ，NSGA ）等。这些智能算法 时、何地投建何种类型的线路及其回路数，以达到规划 与传统的