时变路网下基于混合调整策略的生鲜品配送路径优化研究

Route Optimization of Fresh Food Distribution Under Time-varying Network and Hybrid Adjustment Strategy

针对时变路网下城市车速变化与生鲜品配送过程中产生的各种成本问题,本文提出一种时变路网下基于混合调整策略的车辆路径优化方法。首先,构建包括时间窗惩罚成本、货损成本、车辆成本、碳排放成本,以及客户整体满意度成本等在内的成本最小化目标函数;其次,按区域划分城市道路,并采用深度学习方法对不同区域道路下的车辆速度进行预测;然后,为保证客户满意度和成本的平衡,采用延迟配送及跳过策略相结合的混合调整策略对配送过程中可能导致大规模延误的顾客点进行筛选;最后,通过对Solomon算例和广州市的某一生鲜品配送算例进行求解。结果显示,与传统遗传算法相比改进遗传算法能够加速最优解收敛过程,采用预测速度比采取单个调整策略能够使目标成本同比降低17.29%,准点率达到95.65%。可见,混合调整策略能够在降低目标成本的同时提升到达顾客的准点率,给生鲜品配送成本分析提供一定理论参考。

This paper proposes a vehicle routing optimization method based on hybrid adjustment strategy under timevarying road network.The cost minimization objective function include time window penalty cost,cargo damage cost,vehicle cost,carbon emission cost,and overall customer satisfaction cost.Then,the urban roads were divided by region,and the deep learning method was used to predict the vehicle speed under different regional roads.To ensure the balance between customer satisfaction and cost,the hybrid adjustment strategy was used to screen the customer points that may lead to large-scale delays in the distribution process.At last,a Solomon example and a lifetime fresh food distribution example in Guangzhou were performed.The results indicate that compared with the traditional genetic algorithm,the improved genetic algorithm can accelerate the convergence process of the optimal solution.Compared with the single adjustment strategy,the predicted speed can reduce the target cost by 17.29%,and the on-time rate reaches 95.65%.The hybrid adjustment strategy can reduce the target cost and improve the on-time rate of reaching customers,which provides theoretical reference for fresh food distribution cost analysis.