基于捕食搜索策略的粒子群算法在输电网络扩展规划中的应用  

Application of Particle Swarm Optimization Algorithm Based on Predatory Search Strategy in Transmission Power Grid Expansion Planning

作　　者：符杨[1] 孟令合[2] 罗萍萍[1] 曹家麟[1] 

机构地区：[1]上海电力学院,上海市200090 [2]上海大学机电工程与自动化学院,上海市200072

出　　处：《电力建设》2009年第3期1-4,共4页Electric Power Construction

基　　金：上海市重点科技攻关项目(071605123);上海市教委科研创新项目(08ZZ92);上海市教委重点学科建设项目(J51301)

摘　　要：针对标准粒子群算法粒子维数要求过高和收敛困难易陷入局部最优的缺点,提出一种基于捕食搜索策略的粒子群算法,并将其应用于输电网络扩展规划。捕食搜索在较差的区域进行全局搜索以找到较好的区域,然后在较好的区域进行集中地局域搜索以使解得到迅速改善。通过限制的调节,控制粒子群搜索空间的增大和减小,从而平衡探索能力和开发能力。局域搜索大大降低了粒子数的要求,全局搜索有效地避免陷入局部最优。算例分析证明了算法的有效性和正确性。Aiming at the disadvantages of standard particle swarm optimization algorithms （PSOA）, such as exceedingly high particle dimension, and hard to converge and prone to be trapped in local optimum, a PSOA based on predatory search strategy is presented and is applied to transmission power grid expansion planning. The predatory search strategy is to perform global search in poor regions to find better ones, and then perform concentrated local searches in these regions to improve solution quickly. Through adjusting limits, search spaces can be enlarged or narrowed in a controlled way, so that the exploration and development capabilities can be balanced. Local search greatly reduces the number of particles, and global search can effectively avoid being trapped in local optimum. Effectiveness and validity of the algorithm is validated by case study.

关 键 词：捕食搜索 粒子群算法 电网规划 

分 类 号：TM715[电气工程—电力系统及自动化]