Research And Application Of Genetic Ant Colony Hybrid Algorithm

Intelligent optimization algorithm is develop from simulate or reveal a natural phenomenon or process, is a more macro sense bionic optimization algorithm. With the increase of the actual problem sizes, and the difficulty in modeling and other characteristics, the search space of combinatorial optimization problems is also increasing dramatically. The research directions of people are more and more tend to the optimization algorithm with many methods combined.This paper puts forward a kind of hybrid algorithm based on Genetic Algorithm and Ant Colony Algorithm—Genetic Ant Colony Hybrid Algorithm, the core of this algorithm is the design of the control function, through collected the evolution rate of offspring groups of Genetic Algorithm to control iterative times, so as to determined the fusion time of Genetic Algorithm and Ant Colony Algorithm. This algorithm firstly changed the results produced by Genetic Algorithm into initialization pheromone distribution of Ant Colony Algorithm, then used Ant Colony Algorithm to get the optimal solutions. This algorithm changed the initial value mechanism of Ant Colony Algorithm, improved the stability of the algorithm, and make full use of parallel, positive feedback mechanism and the high solution efficiency of Ant Colony Algorithm to get the optimal solutions.Through the algorithm testing of the discrete problems and continuous problems, and compared with the performance of the traditional Genetic Algorithm and Ant Colony Algorithm, the results showed that Genetic Ant Colony Hybrid Algorithm had stable, global optimization ability, and improved the convergence efficiency. And Genetic Ant Colony Hybrid Algorithm has applied to QoS multicast routing problems, the simulation experiment showed that the solutions of this algorithm were gradually approximation of the global optimal solutions in each generation groups, has better global convergence and higher convergence speed to achieve the optimal solutions. The algorithm has also achieved the optimization of network resources and had a great push forward for network theory and application development.