多中心半开放式送取需求可拆分的车辆路径优化

针对多中心半开放式送取需求可拆分的车辆路径问题，构建了以车辆配送距离最短为目标的多中心半开放式送取需求可拆分的数学模型。设计大变异邻域遗传算法进行求解，采用二维染色体编码及顺序交叉策略，同时运用大变异策略和邻域搜索策略提高算法全局和局部的寻优能力，通过算例对比验证了所提模型与算法的有效性。算例实验表明，大变异邻域遗传算法在求解多中心物流配送车辆路径问题上求解质量较优、求解效率较高、求解结果较为稳定，同时验证了联合配送下多中心半开放式送取需求可拆分的配送模式优于独立配送下单中心送取需求可拆分的配送模式。研究成果不仅拓展了车辆路径问题，还可为相关快递物流企业配送优化提供决策参考。

Multi-depot half open vehicle routing problem with split deliveries and pickups

To solve the problem of multi-depot half-open vehicle routing problem with split deliveries and pickups, this article constructed a multi-depot half-open vehicle routing problem with split deliveries and pickups mathematical model with the shortest vehicle delivery distance as the goal. The solving process mainly used genetic algorithm based on big mutation and variable neighborhood search, which adopt two-dimensional chromosome coding and sequential crossover strategy, the big mutation strategy and neighborhood search strategy improve the global and local optimization capabilities of algorithm. The numerical examples verify the effectiveness of the model and the algorithm. Numerical examples show that the algorithm have benefits in solution quality, efficiency for solving the multi-center logistics distribution vehicle routing problem, and the result is relatively stable. Meanwhile, the multi-depot half-open vehicle routing problem with split deliveries and pickups is superior to the vehicle routing problem with split deliveries and pickups. The research results not only expand the vehicle routing problem, but also provide decision-making references for the distribution optimization of related logistics enterprises.