配网规划中加入扰动策略的蚁群算法

作为一种求解组合优化问题的启发式方法,基本蚁群算法以其正反馈、并行计算和鲁棒性强等优点被许多领域的专家所关注;但是如果正反馈应用不得当,也会出现搜索时间过长或陷入局部最优解的情况。为此,提出采用加入扰动策略的蚁群算法,通过对各计算因子的动态调整,控制其正反馈过程和算法的搜索区域,以提高蚁群算法得到全局最优解的概率和收敛速度。配网规划的应用算例证明了这一改进的有效性。

Ant colony algorithm with little interference strategy in distribution networks planning

Basic ant colony algorithm is a new general-purpose heuristic algorithm for solving combination optimization problems. The main characteristics of this method are positive feedback, distributed computation and the use of constructive greedy heuristic. But if the positive feedback is not proper, it will take longer time on searching or easily running into the local optimization solution. To overcome those shortcomings, this paper puts forward an ant colony algorithms with little interference strategy, which timely adjusts the pheromone. It could control the positive feedback process and enhance the probability of full-scale optimization. The validity is proved by a power network optimization example.