求解复杂旅行商问题的混合粒子群算法

Hybrid Particle Swarm Optimization Algorithm for Solving Complex TSP

针对粒子群算法在解决组合优化时存在早熟和易陷入局部最优的问题,提出一种求解旅行商问题(TSP)的混合粒子群算法。将粒子群算法与遗传算法结合,引入遗传算法中的交叉和变异操作,通过个体极值和群体极值的交叉以及粒子自身变异的方式增加种群的多样性,避免粒子陷入局部最优,提高算法的局部搜索能力。仿真结果表明,新的混合粒子群算法在解决TSP问题时具有较好的收敛性及优化效果。

A hybrid particle swarm optimization algorithm for solving TSP was proposed in this paper. The particle swarm optimization was combined with genetic algorithm because it was premature convergence and easily fell into local optimum solution for solving combinatorial optimization. The crossover and mutation operation in genetic algorithm was introduced into the particle swarm optimization. Increased the diversity of swarm by crossover and mutation between individual extremum and global extremum,avoided particles falling into local optimum and improved the local search ability of algorithm. The experiments show that the hybrid particle swarm optimization is effective to solve the TSP.