This research addresses a vehicle routing problem with simultaneous pickup and delivery with time windows from multiple depots (MVRPSPDTW) over a time horizon in the B2C e-commerce logistics system. We consider an e-commerce logistics system with a multi-period, which consists of customers, logistics service providers (LSPs), suppliers, and a decision-making platform. A mixed-integer non-linear programming (MINLP) model is developed and tested on small- and large-scale instances. To handle more realistic large-scale problems, we have used two approaches (i) exact optimisation approach using (i.e. CPLEX tool) and metaheuristic algorithms (i.e. Differential Evolutionary Algorithm (DE), Parallel Differential Evolutionary Algorithm (Par-DE), Genetic Algorithm (GA), and Block-based Genetic Algorithm (BBGA)) to minimise the total transportation cost and penalty due to the delay by logistics service providers. The computation experiment is conducted on the real practical scenario data and the comparative result is demonstrated.

本研究解决了在 B2C 电子商务物流系统中在一个时间范围内从多个仓库 (MVRPSPDTW) 的时间窗口同时取货和交货的车辆路线问题。我们考虑一个多周期的电子商务物流系统，它由客户、物流服务提供商 (LSP)、供应商和决策平台组成。在小型和大型实例上开发和测试了混合整数非线性规划 (MINLP) 模型。为了处理更现实的大规模问题，我们使用了两种方法 (i) 使用精确优化方法（即 CPLEX 工具）和元启发式算法（即差分进化算法 (DE)、并行差分进化算法 (Par-DE)、遗传算法） (GA), 和基于块的遗传算法（BBGA）），以最大限度地减少由于物流服务提供商的延误而导致的总运输成本和罚款。计算实验是在真实的实际场景数据上进行的，并给出了比较结果。