车辆路径问题的并行粒子群算法研究

Research on parallel particle swarm optimization algorithm for vehicle routing problem

设计了一种引入了模拟退火机制的并行粒子群算法.该算法结合了基本粒子群优化算法的快速寻优能力和模拟退火算法的概率突跳性,避免了基本粒子群优化算法易于陷入局部最优的缺点,提高了进化后期算法的收敛精度.将该算法用于解决车辆路径问题,实验结果表明该算法具有较好的性能.

The proposed parallel particle swarm optimization （PSO） algorithm combines the fast optimum search ablity of original PSO with probability jump property of simulated annealing （SA）. It can avoid trapping to local minima as compared with original PSO and improve the accuracy in the later evolution period. The proposed algorithm was applied to the vehicle routing problem. The experiment results verify that the new algorithm is effective.