一种基于动态局部搜索的蚁群算法及其对tsp问题的求解-计算机应用技三术专业毕业论文.docx

幅度加快收敛速度。各方面都显示，它的性能要优于传统蚁群算法。关键词：蚁群算法动态局部搜索TSP问题求解质量与稳定性 幅度加快收敛速度。各方面都显示，它的性能要优于传统蚁群算法。 关键词：蚁群算法动态局部搜索TSP问题求解质量与稳定性 Ⅱ 万方数据 A A NEW ANT COLONY ALGORITHM BASED ON DYNAMIC LOCAL SEARCH FOR TSP AB STRACT As the traditional ant colony algorithm(ACA)for solving TSP(traveling salesman problem)is easy to fall into a local optimal solution and slow convergence，also the quality and stability of solution is not ideal．To solve these problems，an core improved ant colony algorithm based on dynamic Iocal(DLACA)search which achieved by scout ant，the three auxiliary strategies are city classification，enlarge the gap of pheromone and clear pheromone after inherit is proposed． City classification means that at the beginning of the algorithm，we should classify all cities into several classes．cities of different types represent different levels of the city,and more ants to the higher ranks，while the lower ranks are opposite；And through the introduction of bionics scout ants，is that during ant routing process，if there is the difficulty of choosing cause that it is difficult to choose what way to go，dispatched as appropriate scout ants help to explore，at the end contrast the length of scout ants and ant path to decide which path a is better．In each routing is completed，the new pheromone update strategy appear, which is obtain when the time iteration of the shortest path，and ratio with the shortest path before when the time of seeking，if the ratio is greater,pheromone update amplitude is bigger also；Wrhen the algorithm falls into local optimal，in Ⅲ 万方数据 order order not to let the pheromone completely cleared，use a inherited strategy,SO that not only can jump out of local optimum，but also carl improve the efficiency of the algorithm． The principle of traditional ant colony algorithm to solve TSP were in detail， the analysis of ant colony algorithm in solving traveling salesman problem(TSP) produces the root cause of the three defects，starting from the root causes，design improved ant colo