蚁群算法优化策略及其仿真研究

Strategy of Optimization in Ant Colony Algorithm and Simulation Research

蚁群算法广泛应用于求解组合优化问题,但基本蚁群算法与其他模拟进化算法存在进化速度慢并易于陷入局部最小等缺陷。论文应用蚁群算法求解最短路径问题,从信息量的更新方式、局部搜索策略及参数选择等方面提出相应的改进策略。通过TSP问题的仿真表明,改进算法能够加快收敛速度,节省搜索时间,而且能够克服停滞行为的过早出现。

Ant colony algorithm has been widely applied to solving complicated combinatorial optimization problems. Much deficiency,such as low searching speed and easy falling to the local best,still exists in the basic ant colony algorithm available.In this paper,the algorithm is used for the shortest path problem.It is improved in three parts of information modification,local search strategy and parameters selection,The simulation for TSP problem shows that the improved algorithm can find better path at higher convergence speed,save the search time and overcome the precocity and stagnation.