考虑随机客户需求的两级车辆路径问题研究

Vehicle Routing Problem with Stochastic Demands in Two-echelon Logistics

以两级物流系统为研究对象,着重考虑了随机性质的客户需求,进行系统中每个层级的车辆路径优化求解。由于确定性的单级车辆路径问题即为NP-hard问题,为了对此复杂随机优化问题加以求解,设计了一种基于蒙特卡洛仿真的高效优化方法,将仿真过程嵌入启发式算法的邻域搜索过程。基于现有的两级物流车辆路径问题标准算例,生成具有随机客户需求的本问题算例,通过数值实验并与确定性方法比较,验证了所提出的基于仿真的启发式算法可行而有效。

The aim is to addresses the two-echelon logistics problem with a focus on the stochastic nature of customer demands,and derives the vehicle routing optimal solutions of two levels.Since the deterministic one-echelon vehicle routing problem is NP-hard,an effective Monte Carlo simulation based optimization approach was put forward to solve this complex stochastic optimization problem.The approach embeds the simulation process into the neighborhood search of the heuristic algorithm.Based on the existing two-echelon vehicle routing problem benchmarks,new benchmark instances considering stochastic customer demands are generated.Through the computational experiments as well as comparisons with the deterministic method,the final results indicate the feasibility and effectiveness of the proposed simulation-based heuristic algorithm.