考虑路况、配送员与顾客满意度的冷链物流车辆配送路径优化方法

在城市交通拥堵日益严重的背景下，为解决冷链货物配送时效性较差、顾客满意度低以及配送员积极性不高等问题，研究了以总成本与满意度为目标的冷链物流车辆路径优化方法。采集并拟合了工作日、节假日、双休日交通拥堵指数趋势图，得到道路交通拥堵时间分布规律，提出了不同时期、时段内道路路段节点之间的实际通行时间计算方法；设计了配送员薪资与工作强度相结合的评价指标，构建了基于灰色白化权函数的配送员满意度评价模型。考虑随机需求、满意度和时间窗等约束，构建了多目标车辆路径优化模型；针对多目标模型求解的复杂性，设计了改进的自适应大规模邻域搜索算法，算法搜索过程中充分利用其算子的自适应性，有效平衡了NSGA-II算法大规模寻优与耗时之间的关系。以经典的Sioux-Falls交通网络为例进行算例分析，结果表明:①考虑配送员满意度模型后，在工作日、节假日、双休日总成本分别增加了2.05%、1.93%和1.16%，但配送员满意度分别提高了39.43%、46.26%和57.37%，顾客满意度平均提高了1.16%、4.76%和9.75%，运输时间缩短了2.42%、7.34%与8.41%。②以配送总成本最小为主要目标时，当需求变动（即随机需求的标准差σ＝1，2，3，4，5）时，得到的缺货成本比未考虑随机需求模型的结果分别增加了0.79%、0.89%、0.93%、0.94%和0.95%，印证了顾客随机需求对企业配送成本产生的影响。所提模型和算法为冷链物流管理中提升配送员和顾客满意度提供了一定的理论基础。 

A Route Optimization Method for Cold Chain Logistics Vehicles Considering Road Conditions,Satisfaction of Deliverymen and Customers

Considering increasing traffic congestion in urban areas, routing problem of vehicles serving the cold chain logistics is modeled as an optimization problem with multiple objectives (namely delivery cost and satisfaction of customers and deliverymen), which is expected to provide solutions to the punctuality and satisfaction issues within the cold chain logistics industry. Traffic data on weekdays, holidays, and weekends are collected, the temporal distribution of the congestion is studied, and a model for estimating travel time among different nodes within road network is also developed. A gray-whitening weight function is used to evaluate the satisfaction of deliverymen by considering workload and salary. A multi-objective vehicle routing problem is modeled with the following constraints, including stochastic demand, customer satisfaction, time windows, and others. An improved adaptive large neighborhood search (IALNS) algorithm is proposed, which balances the searching scale and computational time of the non-dominated sorted genetic algorithm-II (NSGA-II). The results of the proposed method in the classical Sioux-Falls transportation network are shown as follows: ①When the deliveryman's satisfaction is considered, the total cost of delivery on working days, holidays, and weekends increases by 2.05%, 1.93%, and 1.16%; the deliveryman's satisfaction increases by 39.43%, 46.26%, and 57.37%; the average customer satisfaction increases by 1.16%, 4.76%, and 9.75%; and the transportation time decreases by 2.42%, 7.34%, and 8.41%, respectively. ② When the total cost is set as the main objective andthe standard deviations of customer demand are set as 1, 2, 3, 4, and 5, the shortage cost increases by 0.79%, 0.89%, 0.93%, 0.94%, and 0.95%, respectively, compared with the case without considering the stochastic demand. Study results show that the stochastic demand from the customers has an impact on distribution costs. In conclusion, the proposed model and algorithm can provide a new methods for improving satisfaction of deliverymen and customers of cold chain logistics industry.