Research On Ant Colony Optimization With Immunity Ability

Ant Algorithm (AA) is a hotspot in the research of the meta-heuristic algorithms, and from it was introduced, it has been used to solve all kinds of complex optimization problem. As a global searching approach, AA has some characteristic, such as positive feedback, distributing, paralleling, self- organizing, etc. But AA also has many shortcomings, such as slow convergence and being premature. In order to improve its capability, this paper does a lot research of genetic algorithm (GA) and immune algorithm (IA) besides AA, and proposes two new algorithms. Making use of GA and IA's advantages separately, the new proposed algorithms' performance is meliorated.Firstly, aiming at solving AA's scarcity of the capability of recognizing the characteristic information of the problem, this paper introduces the idea of vaccine in IA then proposes a new ant colony algorithm with immune ability. The vaccine is picked up from the transcendental knowledge and then injected into the ant. So the ant has the "immune ability" and could find a better solution.Secondly, in order to improve the ability of exchanging information between ants, this paper introduces GA into ACS' every iteration. The cross operator in GA improves the communicating ability and the mutation operator can help the algorithm find a better solution. So the new algorithm's convergence speed is quickened and its performance is improved.At last, this paper applied two algorithms into traveling salesman problem to test their performances. The simulation results show that the new algorithms could find optimum solutions more effectively in time and quantity.