An Improved Ant Conlony Algorithm For The Traveling Salesman Problem

At the end of 1990s ant colony algorithm is proposed by Italian scholars M .Dorigo, who is inspired by the ant finding the shortest road from the nest to the destination. It is a new kind of heuristic searching algorithm after the tabu search, simulated annealing algorithm, genetic algorithm and artificial neural net and soon. Ant colony algorithm not only has the characteristics of intelligent search, global optimization but also has the characteristics of stability positive feedback, distributed computing and combination with certain heuristic algorithm. Stability is good, modify the algorithm on the base of the ant colony algorithm model, then it can be used into other problem. Positive feedback makes it quickly to find better solutions. Distributed computing makes it easily to come true parallel, the unit communicates and pass the information, this is good for finding the better solutions, and not easily plunging into the partial optimization. Combination with certain heuristic algorithm can improve the capability of the algorithm. It is used in combinatorial optimization problem successfully. Some research and application prove the superiority of the ant colony solving the complex optimized problem (especially the discrete optimized problem) .Now the research is mostly in Belgium, Italy, Germany and so on .In China, some schools and graduate schools in Shanghai, Beijing and the northeast made this work ,they mostly used the ant colony algorithm to solve travel man problem.The work of this task include follow three aspects: Firstly, researching on ant colony algorithm, summarize the studying headway .Sum up the applying area of ant colony algorithm and the disadvantage, and deeply analyzing the disadvantage to improve total capability of ant colony algorithm. but it has some shortcomings such as its slow computing speed and it easy to fall in local peak in large scale problem. To overcome these deficiencies, an improved ant colony algorithm is designed through abstracting the advantages of pheromone, The results of the experiment suggest that the improved algorithm is effective .Last the traveling salesman problem based on ant colony algorithm which published in nation and abroad are very few, so ant colony algorithm is applied in the traveling salesman problem, which is effective attempt.