考虑分散协作及数量折扣的双目标电动车辆路径优化

Dual-objective Electric Vehicle Routing Optimization Considering Decentralized Collaboration and Quantity Discount

当前传统燃油车辆造成了极大的空气污染和资源浪费,电动车辆和协作物流是降低碳排放、提高运输效率的有效途径。本文基于协作物流的思想,建立以运输利润最大及配送任务完成量最大为双目标,考虑分散协作及数量折扣的带时间窗电动车辆路径优化模型。设计将贪婪随机自适应搜索—进化邻域搜索(GRASP-ELS)混合算法与ε-约束法相结合的ε-约束混合进化算法,并通过算例对模型和算法进行测试。实验结果表明:所提出的算法优于多目标优化算法NSGA-Ⅱ;通过灵敏度分析给出管理启示。本文为分散协作情境下电动车辆配送优化提供方法借鉴与决策参考。

The logistics industry is one of the main sources of global energy consumption and carbon emissions.The use of traditional fuel vehicles has caused significant carbon emissions and resource wastage.Electric vehicle(EV),in comparison,can reduce lifecycle carbon emissions by up to 47%,making their adoption for delivery purposes an effective strategy for lowering transportation carbon emissions.Moreover,collaborative logistics models enable different logistics stakeholders to share logistics resources and information,thereby enhancing the efficiency of EV transportation and reducing energy consumption.However,in reality,most enterprises are unwilling to share all information with competitors and prioritize their own interests over collective benefits,making it difficult to achieve optimal overall benefits.Thus,many companies prefer a decentralized collaboration model that involves sharing partial information.Based on this,this paper studies the optimization of electric vehicle routing within a decentralized collaborative framework.Drawing from real-world scenarios and existing research on decentralized collaboration,this paper adopts a platform-based order selection model.Transport companies participating in collaboration use a platform to select profitable delivery orders and complete delivery services for compensation.Orders that are unprofitable due to long distances or high marginal costs are submitted to the platform,where they are available for selection by other companies.In relevant studies,unit transportation costs are often set as fixed values,but due to the economies of scale in transportation,the actual unit costs tend to decrease as transportation volume increases.Consequently,this paper reflects a quantity discount strategy in the cost structure,making transportation fees a piecewise linear function of the total task volume.Additionally,in collaborative logistics,the number of customers served by a logistics company reflects its market share,which companies aim to maximize.Therefore,this paper optimizes the completion volume of delivery tasks.Taking into account the context and objectives of the problem,we establish a dual-objective optimization model for electric vehicle routing problem with time windows(EVRPTW),considering decentralized collaboration and quantity discount.To address the proposed problem,we design anε-constrained hybrid evolutionary algorithm.This algorithm converts the dual-objective problem into a single-objective problem through theε-constraint method and combines the greedy randomized adaptive search procedure(GRASP)and evolutionary local search(ELS)algorithms to solve the converted single-objective problem.To validate the effectiveness of our algorithm,computational results are compared with the classic multi-objective optimization algorithm NSGA-II,evaluating algorithm performance based on indicators such as hypervolume and computation time.Based on the Solomon standard test set,we randomly generated 4 sets of 12 test instances with scales of 50,75,100,and 125 customer points,each set containing three instances of the same scale.Numerical results indicate that our algorithm surpasses the NSGA-II algorithm in both solution quality and computational speed.This demonstrates that our algorithm enhances the search and optimization capabilities within the solution space and can effectively solve the dual-objective EVRPTW considering decentralized collaboration and quantity discount.Further sensitivity analysis tests the impact of different customer distribution scenarios on collaborative efficacy.The innovations of this paper include:(1)proposing a dual-objective optimization model for EVRPTW based on decentralized collaboration;(2)applying a quantity discount strategy to transportation costs and optimizing for both profits and delivery task completion volumes,thereby more accurately depicting the decentralized collaboration model;(3)designing anε-constrained hybrid evolutionary algorithm for this problem,embedding an improved GRASP-ELS hybrid algorithm to solve the researched problem.This paper provides decision-making references and managerial insights for logistics distribution companies:adopting decentralized collaboration strategies can effectively enhance transportation efficiency and economic benefits,especially in areas where customer geographical locations are more clustered.Additionally,this paper broadens the research perspective in the field of electric vehicle routing optimization and offers theoretical references and methodological guidance for further research on logistics scheduling problems under a decentralized collaboration condition.