Ant System And Its Application In TSP

Since creating bionics in middle period of 1950's, people are being inspired from the mechanism of the biological evolution constantly. Many new methods had been applied to solve the complicated optimization problems are proposed. Such as neural network,genetic algorithm,simulated annealing,and evolution computation. These new methods had been successfully applied to solve the practice problems. Ant colony optimization (ACO) was first proposed by Italy scholar M.Dorigo, V.Maniezzo, A.Colorni in1992 as a novel bionic evolutionary algorithm for solving complicated combinatorial optimization problems, like the TSP, the quadratic assignment problem(QAP) and job-shop problem, etc. At present, the research work about AS already arouse the attention from more scholars and expert gradually. Though, this research approach lies at primary stage, but some research results have already demonstrated the superiority of ACS. The main contents are composed of the following parts:â‘ The dissertation focuses on the principles, theory, and applications of ACO, especially, an in-deep and systemic study on how to improve the basic ACO algorithm in TSP. The principle, the model, the characteristics and the management about the basic algorithms of any colony (ant system) are also presented. The reasonable selection about the parameters is discussed in emulated test.â‘¡Because the traditional ant calculate way easily appears precocious and stagnation behavior phenomenon, so this text put forward one kind lately from orientation ant calculate way, it is to information pheromone of the traditional ants calculate within calculate way to carry on dynamic state of adjust from the orientation. Test result enunciation that calculate way to have better search the ability of superior solution of over all situation, and have better stability and astringency.Finally, the work of this dissertation is summarized and the prospective of future research is discussed.