关键词: 蚁群算法（ACO）, 猫群算法（CSO）, 旅行商问题（TSP）, 动态分组, 自适应信息素扩散

Abstract:

Aiming at the problems that the ant colony algorithm is easy to fall into local optimum and the convergence speed is slow in traveling salesman problem (TSP), dynamic grouping ant colony algorithm combined with cat swarm optimization is proposed. Firstly, the ants are distributed manually in different cities evenly when the population is initialized. Secondly, based on the division of labor in the cat swarm optimization, a dynamic grouping mechanism is introduced into the ant colony system to classify the ants into two types: search ants and tracking ants. Search ants increase the algorithm??s diversity in the early stage by improving the path construction rules, tracking ants use pheromone diffusion mechanism to adaptively update local pheromone to highlight the role of superior sub-paths, and avoid the algorithm falling into local optimum. Finally, the convergence speed is accelerated by the pheromone global update mechanism. Simulation experiments are carried out on multiple sets of instances in TSPLIB through Matlab. The simulation results show that the improved algorithm balances the diversity and conver-gence speed, and effectively imporves the quality of the solutions.

Key words: ant colony optimization (ACO), cat swarm optimization (CSO), traveling salesman problem (TSP), dynamic grouping, adaptive pheromone diffusion