关键词: 多态蚁群算法, 捕食搜索, 旅行商问题（TSP）

Abstract: Aiming at the problem of iterative stagnation in solving the traveling salesman problem with polymorphic ant colony algorithm, the improved ant colony algorithm is proposed. Combining with the predatory search algorithm, a new fusion ant colony algorithm is proposed. This fusion algorithm introduces the following mechanisms. In the artificial ants path-selecting stage, the snoop path is set to be a priority, a penalty factor is set for a non-snoop path, the pheromone is applied to both the snoop and non-snoop paths, and the mechanism is used to avoid the polymorphic ant colony algorithm’s stagnation. The predator search strategy is applied in the neighborhood of the optimal result of each round of artificial ant, and the optimal solution is selected by the competitive mechanism to update the pheromone. The research shows that the above mechanism can overcome the shortcomings of the polymorphic ant colony algorithm. The predatory search strategy can be used to guide the distribution of pheromone effectively and speed up the convergence of the algorithm to the optimal solution and improve the quality of the optimal solution. Simulation results for the TSP problem prove the effectiveness of the proposed algorithm.

Key words: polymorphic ant colony algorithm, predatory search, Traveling Salesman Problem（TSP）