Abstract: To overcome the slow convergence and local extrema of ant colony algorithm, the cloud model theory is adopted to regulate reasonably the degree of randomness of the ant colony algorithm. Several adaptive strategies are proposed for the parameters of the ant colony algorithm and the cloud model, and for the optimum path determination. Meanwhile, the evaluation algorithm of pheromone distribution is proposed. Simulation results for multiple TSP validate the efficiency and stability of the proposed algorithm.

Key words: pheromone, ant colony algorithm, cloud model, traveling salesman problem

摘要： 为了解决蚁群算法易早熟于局部最优及收敛速度慢的问题，采用云模型理论来合理调控蚁群算法的随机性程度，分别提出针对蚁群算法参数、云模型参数以及较优路径判定的自适应调整策略，同时提出信息素分布状态的评价算法。针对多个TSP问题进行仿真实验，结果验证了提出的算法的高效性与稳定性。

关键词: 信息素, 蚁群算法, 云模型, 旅行商问题