A Multi-strategy Improved Ant Colony Algorithm for Solving Traveling Salesman Problem

2018 5th International Conference on Advanced Composite Materials and Manufacturing Engineering

Xiao, Yanqiu^1 ; Jiao, Jianqiang^1 ; Pei, Jie^1 ; Zhou, Kun^1 ; Yang, Xianchao^1

Department of Electromechanical Science and Engineering, Zhengzhou University of Light Industry, Zhengzhou

450002, China^1

关键词: Ant colony algorithms;    Computing performance;    Elitist strategies;    Evolution strategies;    Improved ant colony algorithm;    Nearest neighbor method;    Solution accuracy;    Solution algorithms;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/394/4/042101/pdf
DOI  :  10.1088/1757-899X/394/4/042101

来源: IOP

【 摘 要 】

A multi-strategy improved ant colony algorithm is proposed. In order to solve the problem of solving the TSP problem in the ant colony algorithm, it has the problems of low solution accuracy, easy fall into local optimum, and low solution efficiency. The nearest neighbor method is used to influence the distribution of the initial pheromone to reduce the pheromone concentration on the short path in the initial stage of the algorithm. Based on the mutation adjustment of the transfer rule, a mean cross-evolution strategy is combined with the mean of the path to enhance the global solution space of the algorithm. Ability and ability to avoid falling into a local optimum. Then, the iterative and elitist strategies are combined to improve the pheromone update mechanism to further improve the solution algorithm's solution performance and solution efficiency. Finally, the eight instances selected from the TSPLIB database are solved and compared with other algorithms. The experimental results shows that the improved algorithm is efficient when solving the traveling salesman problem and has high computing performance.