Abstract: Aiming at the problem of "multiple warehouses in one location, terminal chiasmas, high cost and high emissions" faced by the front distribution center delivery mode, a multi-depot lateral collaborative distribution model for heterogeneous fleets was proposed. By introducing the heterogeneous fleet comprehensive carbon emission function which considered factors such as speed, time, distance, load and engine, the carbon emission cost, driver cost and dispatch cost were converted into an objective function by linear weighting. An improved genetic algorithm with special head structure and chromosome coding mechanism was designed, and a crossover operator and dynamic weight adjustment mechanism were introduced to achieve efficient search in solution space. The research results show that compared with the traditional front warehouse lateral replenishment scheduling, the heterogeneous fleet multi-depot lateral collaboralive replenishment scheduling reduces the total cost by 39.1%, reduces the CO2 emission by 12.5% and saves the number of vehicles by 27.2%. As the proportion of collaborative customers to total customers increases, the total cost further decreases. Compared to carbon emission costs and dispatch costs, driver costs account for a larger proportion of the total cost, which needs to be paid attention to by logistics companies.

摘要： 针对前置仓配送模式面临的“一地多仓、末端交叉、高成本、高排放”难题，提出异构车队多中心横向协同配送模型。通过引入考虑速度、时间、距离、载重和发动机等因素的异构车队综合碳排放函数，将碳排放成本、司机成本和派车成本通过线性加权方式转换为目标函数；设计含特殊头部结构和染色体编码机制的改进遗传算法，引入交叉算子和动态权重调整机制以实现解空间的高效搜索。研究结果表明：相比传统前置仓横向转移补货调度，异构车队多中心横向协同补货调度在总成本方面减少39.1%，CO2排放量减少12.5%，车辆使用数节约27.2%；随着协同客户占总客户的比例增加，总成本进一步减少；相较碳排放成本、派车成本，司机成本在总成本中占较大比例，需引起物流企业重视。