基于规模压缩的混合蚁群算法

Hybrid ant colony algorithm based on scale compression

为了提高蚁群算法处理大规模问题的性能,提出一种基于规模压缩的混合蚁群算法.根据TSP问题的最优解与次优解共享部分路径片断的原理,设计城市压缩算法,减少了TSP问题的城市处理量.在求解过程中,引入最优解的区域特征的概念,采用优化状态转移规则,压缩了解空间.仿真实验结果证明,采用所提出算法得到解的质量和收敛速度都有显著提高.

To improve the performance of ant colony algorithm in solving large-scale TSP problem, a hybrid ant colony algorithm based on scale compression is proposed. Genetic algorithm is used to generate a suboptimal solution set and calculate their intersection. By eliminating all cities mapped by the elements among the intersection in the primal TSP problem, the original problem is converted into a new one with smalier scale. In addition, an optimal state transition rule is designed based on regional characteristics of optimal solutions to accelerate convergence speed. Simulation results show the approach possesses high searching ability and excellent convergence performance.