基于Memetic算法的两级车辆路径优化

针对传统两级车辆路径优化过程,存在的精度不高,收敛时间过长的问题,提出一种基于Q学习理论和差分进化的Memetic算法。首先,对两级车辆路径优化问题进行研究,利用最优分割法获得第一级配送方案,以此确定中转站配送数量,然后求解第二级多配送中心车辆路径问题配送方案,获得两级优化问题的总里程及总配送车辆数量;其次,针对第二级MDVRP配送方案求解,利用Q学习理论和差分进化算法,设计新的Memetic算法,来实现对多配送中心车辆路径问题配送方案的全局优化;最后,通过仿真验证了所提算法的有效性。

Two-echelon vehicle path optimization based on Memetic algorithm

Aiming at the problem of low accuracy and long convergence time of traditional method in solving the two-echelon vehicle routing problem, we proposed a kind of Memetic algorithm based on Q learning theory and differential evolution. Firstly, the two-echelon vehicle routing problem was studied, and the optimum partition method was used to obtain the reasonable distribution plan for SDVRP(split delivery vehidle fouting problem) in first stage, and then the total mileage and delivery vehicles were determined for both the two stages. Secondly, according to the distribution scheme of the second level of MDVRP(multi-depot vehivle fouting problem), the Memetic algorithm was designed with Q learning theory and differential evolution algorithm, which was used to achieve the global optimization of MDVRP distribution scheme. Finally, through simulation verified the effectiveness of the proposed algorithm.