The Research And Application Of The Ant Colony Algorithm

Ant colony algorithm is a novel category of bionic evolutionary algorithm, firstly proposed in 1991. It is used in a variety of complex combinatorial optimization problems because it has positive feedback, parallel computing and robustness, etc. The algorithm has significant advantages, but it can not be ignored such as computing for a long time, being easy to fall into stagnation and other shortcomings. Now, with only twenty years of development of the ant colony algorithm, there is still a lot of room for improvement. Therefore, further study of the ant colony algorithm for both theory and practical application means a lot.In this paper, we propose two improved algorithm and verified with examples to solve the problems of long searching time and being easy to fall into local optimum on the basis of improved ant colony algorithm and the classical algorithm, and get a relatively good result. The concrete content and innovative points can be summarized as follows:(1) The basic principles of basic ant colony algorithm are summarized to establish the theoretical basis for the further study of the ant colony algorithm.(2) On the basis of ant colony system, the improved part is using a strategy of neighboring nodes priority in order to improve the efficiency of the search path and combining the pheromone update local and global dynamic to avoid stagnation. We use this improved algorithm to solve two typical problems, and the results are compared with the ant colony system, in order to prove the effectiveness of the improved algorithm.(3) The modified 2-opt algorithm is used based on the ant colony system to increase global search ability and improve search efficiency. Set different pheromone evaporation factor in the algorithm early, mid and late period, so that it can increase search algorithm range in the initial and speed up the convergence in the latter. The improved algorithm is used to solve the problem of the Chinese traveling salesman, and a brief analysis of the experimental results will be showed.