考虑需求密度和客户满意度的城市末端服务网点选址问题

Location Problem of End-of-City Logistics Facilities with Capacity Constraints Considering Distribution Efficiency and Customer Satisfaction

为解决“最后一公里”的配送问题,结合城市末端配送需求的服务网点特点,考虑需求密度和客户满意度,建立了以时间满意度最大化、利润最大化和需求密度最大化为目标的多目标选址模型。设计了基于线性加权法的改进自适应遗传算法,采用自适应算子处理变异概率和交叉概率,并设计三种不同的交叉算子,扩大搜索邻域。在改进算子的基础上设计了多目标遗传算法,使用快速非支配排序法去除权重系数取值对选址结果的影响。利用北京市海淀区的快递服务网络数据进行实例分析,使用设计的两种遗传算法分别求解选址模型的全局最优解和帕累托前沿解;并将改进自适应遗传算法与传统遗传算法、自适应遗传算法的求解结果进行比较。结果显示:改进自适应遗传算法比传统遗传算法和自适应遗传算法收敛速度更快、搜索能力更优;若无法确定改进自适应遗传算法的权重,使用改进后的多目标遗传算法同样可以求到最优解。

In order to solve the“last mile”distribution problem,combined with the characteristics of the logistics facilities in the end of the city,considering demand density and customer satisfaction,we established a multi-objective location selection model with the goal of maximizing time satisfaction,profit and demand density.We designed an improved adaptive genetic algorithm based on linear weighting method.We used adaptive operator to deal with mutation probability and crossover probability,and design three different crossover operators to expand the search neighborhood.Based on the improved operator,we designed a multi-objective genetic algorithm,and use the fast non dominated sorting method to remove the influence of the weight coefficient on the location results.We used the express service network data of Haidian District in Beijing as an example.We used two kinds of genetic algorithms to solve the model,respectively solve the global optimal solution and Pareto frontier solution,and compare with the solution of traditional genetic algorithm to analyze the efficiency of the algorithm.The results show that the improved adaptive genetic algorithm has faster convergence speed and better search ability;if the weighted weight cannot be determined,it is more suitable to use multi-objective genetic algorithm to solve the model.