整合逆向物流协同配送动态路径优化问题研究

Dynamic vehicle routing problem with simultaneous pickup and delivery in collaborative distribution under demand concurrent

整合逆向物流的协同配送面临着客户需求随机并发、行驶路径开环多变等不确定性挑战,这给配送路径动态规划和方案柔性制定带来了更严苛的要求.为此,使用“时间片”划分方式,将动态配送问题转化为一系列静态优化问题,提出了整合逆向物流的协同配送动态路径优化两阶段求解策略.首先,通过考虑车栽量以及时间窗等软硬约束条件,构建以总物流成本最小和客户满意度最大的多配送中心协同配送初始-实时动态路後规划模型.其次,设计带有非支配排序与精英策略的多目标免疫遗传算法M0IGA对模型进行求解.最后,研究结果表明客户满意度与物流运营成本之间呈现悖反关系,高配送成本会带来高客户满意度,在相同成本支出情况下,M0IGA算法产生的规划方案客户满意度优于NSGA-Ⅱ、M0IA,这也验证了模型的有效性以及方法优越性.

The uncertainty in dynamic route planning and flexible scheme formulation will bring more severe challenges for collaborative distribution with simultaneous pickup and delivery,due to random concurrent of customer demand and variable open-loop of driving route.In this paper,“time slice”is applied to transforming dynamic vehicle routing problem(DVRP)into a series of static optimization problems,and a two-stage solution strategy for collaborative distribution with simultaneous pickup and delivery is proposed.Firstly,taking soft and hard constraints such as vehicle load and time window into account,a two-stage vehicle routing model composed by initial and actual time stage is constructed,which aims at minimizing the total logistics cost and maximizing customer satisfaction.Secondly,a multi-objective immune genetic algorithm(MOIGA)combined with non-dominated ranking and elite strategy is designed to solve the model.Finally,the results show that the relationship between customer satisfaction and logistics operation cost is contrary.High distribution cost will bring high customer satisfaction.Under the same cost expenditure,the customer satisfaction of schemes generated by MOIGA is higher than that by NSGA-Ⅱ or MOIA,which verifies the effectiveness and superiority of the model.