Improvement And Simulation Research On Ant Colony Algorithm

Ant Colony Optimization algorithm (ACO) is an algorithmic approach, inspired by the foraging behavior of the real animals, which had been applied to many problems. The dissertation focuses on the principles, theory, and applications of Ant Colony Optimization algorithm (ACO) ,especially, an in-deep and systemic study on how to improve the basic ACO algorithm, parallel implementation of ACO, solving the problems such as the TSP problem and Multiple objective optimization.First,as is known stagnation behavior is a disadvantage of the ant colony algorithm. A dynamic weight based on select strategy is proposed to enhance its global searching ability. The search range is initially enlarged by the pheromone diffusion. After certain iterations, the search range is mainly enlarged by path expectation. As a result, the performance of the ant colony algorithm is improved. The new algorithm and the classical ant colony algorithm are applied to the TSP problem. Numerical results show that the new algorithm is feasible and the stagnation behavior is avoided.Secondly,in order to preserve the diversity of Pareto optimal solutions in multi-objective optimization problems, A new ant colony algorithm is proposed. In the proposed algorithm, the selection strategy is multi-pheromone-weighted, and pheromone update uses the combination of the local and global pheromone update. Especially, the global pheromone update adopts the best solution and the second-best solution. In addition, an external set is set up outside to store the Pareto solution, and the improved algorithm is used to solve the bi-criterion TSP. After the simulation experiment, it is shown that the new algorithm is more efficient than SPEA2 and NSGA-II.