Application And Research Of Expansibility Of Ant Algorithm

Ant algorithm is a new meta-heuristic algorithm. There is currently a lot of ongoing activity in the scientific community to extend or apply ant-based algorithms to many different discrete optimization problems. It has been used successfully in many fields, such as TSP, assignment problem, job-shop scheduling and graph coloring. The theory and application of ant algorithm are not mature as genetic algorithms and simulated annealing, so we must study it more.In this paper, we discuss the characteristic of ants when they search for food, give the theory and model of ant algorithm, and analyze the relation and difference between artificial ants and real ants. We use based and improved ant algorithm to solve problems, such as TSP, QoS multicast routing, VC routing, phone net routing, convex integer programming and degree-constrained minimum spanning tree problem, then compare the results to show that the improved ant algorithm is better. When solving the QoS multicast routing problem, we compare this algorithm with the other meta-heuristic algorithm梘enetic algorithm, and find this algorithm has good effect. Otherwise, we first put forward a viewpoint梕xpansibility of ant algorithm. We use this new viewpoint in some fields and find it is predominant and effective.Base on solving the problem of degree-constrained minimum spanning tree, we put forward a new method to solving clustering problem. The efficiency of this method is not high, but it can conquer the blindly of clustering.