改进蚁群优化算法在配电网网架规划中的应用  

Application of Improved Ant Colony Optimization Algorithm in Distribution Network Planning

作　　者：孙薇[1] 商伟[1] 牛东晓[1] 

机构地区：[1]华北电力大学工商管理学院,河北省保定市071003

出　　处：《电网技术》2006年第15期85-89,共5页Power System Technology

摘　　要：从转移概率和信息素的动态更新机制两方面对传统蚁群算法进行了改进。转移概率设计中采用2个参数的联动控制,实现了Dorigo和Maniezzo的设想。信息素动态更新机制使信息素权重随迭代次数增加而改变,实现了对整个可行域的搜索全面而迅速,弥补了传统蚁群算法训练效率低下和易陷入局部极小的不足。配电网优化的结果表明该算法的有效性。Traditional ant colony algorithm is modified in the aspects of transition probability and dynamic updating mechanism sociohormone. In the design of transition probability the gang control of two parameters is adopted to implement the Dorigo and Maniezzo＇s conception proposed in the year of 1992. The dynamic updating mechanism of sociohormone makes the weight of sociohormone varied with the increment of iteration times, so the all-round and speedy search in whole feasible region is realized and the defects of low training efficiency and easy to fall into local minimum in traditional ant colony algorithm are remedied. The result of distribution network optimization shows that the modified algorithm is effective.

关 键 词：配电网优化 改进蚁群算法 转移概率 更新机制 信息素 

分 类 号：TM714[电气工程—电力系统及自动化]