The Research And Application Of Improved ACO Algorithm For DTSP

Ant Colony optimization algorithms inherent parallelism, robustness and good characteristics make it the solution of complex combinatorial optimization problems in a very efficient calculation model,but it also has any defects, as slow initial speed and premature or stagnation phenomenon. Aiming at improved this defects, this paper give three improved algorithms: Immune-based vaccine Ant Colony Optimization is the introduction of the Ant Colony system, improve the immunity vaccine algorithm of initial solution speeds; A pheromone dynamic update of Ant Colony optimization algorithm is based on immunization vaccines ACO add information in dynamic update mechanism and path smoothing mechanism, this algorithm improving the algorithms speed and avoiding stagnation phenomenon of algorithm; An approach based on Delaunay Triangulation of large-scale DTSP problem solving from the candidate set policy on the algorithm improvements, reduced by Delaunay triangulation point search space, and then uses the dynamic update mechanism and path smoothing mechanism to avoid stagnation of the algorithm. Experiments show that the problem of mass DTSP, the algorithm of solving speed increased by 20%-40%.