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.
出处
《电网技术》
EI
CSCD
北大核心
2006年第15期85-89,共5页
Power System Technology
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
distribution network optimization
improved ant colony algorithm
transition probability
updating mechanism
sociohormone
