动态随机环境下电动汽车柔性配送路径优化问题研究

At present, our country is vigorously promoting the use of electric vehicles for logistics distribution. In comparison to traditional vehicles, electric vehicles have the characteristics of shorter driving mileage and longer charging time. According to the operation characteristics of electric vehicles and considering the time-dependence and randomness of the actual distribution network environment, this research will carry out the research on the routing optimization of urban logistics distribution of electric vehicles. Firstly, on the basis of analyzing the time-dependent and random characteristics of road network environment, the research describes the service time window, delivery, charging and other activities of electric vehicle logistics distribution in the space and time network effectively. Secondly, considering the path flexibility between any two customers, the optimal routing model of flexible distribution for electric vehicles with time window and pickup and delivery in time-dependent and stochastic environment is formulated and an efficient algorithm is designed. Thirdly, based on the idea of multi-stage stochastic programming, the distribution activity network of electric vehicles is divided into three levels, and a stochastic three-stage distribution routing optimization model based on electric vehicle is formulated and an efficient heuristic algorithm is also designed. Finally, in cooperation with postal and express enterprises, this research makes an empirical analysis of the proposed models and algorithms, and gradually modifies the models and algorithms in order to make it effective in the distribution of electric vehicles in postal and express enterprises. The research will provide a strong theoretical basis and effective method for the promotion of the use of electric vehicles and green transportation in the field of logistics in China.

目前，我国正在大力推广使用电动汽车进行物流配送。与传统汽车相比，电动汽车具有续驶里程短、充电时间长等特点，本项目将针对电动汽车的运营特性，考虑实际配送路网环境的动态性和随机性，开展电动汽车城市物流配送路径优化问题的研究。首先，在分析路网环境动态随机性的基础上，对电动汽车物流配送中的服务时间窗、取送货、充电等活动在时空网络中进行有效刻画；其次，考虑客户间的路径柔性，建立动态随机环境下带时间窗和取送货需求的电动汽车柔性配送路径优化模型并设计高效求解算法；再次，借鉴多阶段随机规划思想，将配送网络划分为三个层次，建立基于电动汽车的随机三阶段配送路径优化模型并设计高效启发式求解算法；最后，与邮政快递企业相互合作，对所提模型与算法进行实证分析，逐步修正模型与算法，力图使其在邮政快递企业电动汽车物流配送中行之有效。该项目的研究将为我国在物流领域推广使用电动汽车及绿色运输提供强有力的理论基础和有效方法。

在我国大力推广使用电动汽车进行城市物流配送的背景下，本项目以配送中心、客户及充电站为研究对象，探讨了如何在复杂路网环境下通过综合考虑客户点之间的路径柔性及电动汽车里程约束、充电需求等因素，合理规划包括配送路线、行驶路径、充电活动在内的高效配送方案。主要研究工作如下：. （1）研究了复杂路网环境下考虑速度时变性且带时间窗的电动汽车柔性配送路径优化模型及算法。针对路网复杂性，将电动汽车行驶的时间区间离散为多个间隔相同的时间段，在各离散时间段上采用与时间相关的变量表示行驶速度，考虑了任意两节点间的路径柔性，并对电动汽车进行了耗电分析，建立了以总行驶距离最小为目标的0-1混合整数规划模型。基于多线程技术，采用改进的变邻域搜索算法对所提出的模型进行有效求解，同时显著提高了求解效率。. （2）研究了考虑多车型及取送货一体化的电动汽车配送路径优化问题。针对电动汽车的特性，通过考虑电动汽车的里程约束和充电需求，以最小化行驶距离为目标，构建了考虑多车型和同时取送货的电动汽车配送路径优化模型。根据所构建模型的性质与特点，设计了基于变邻域搜索的启发式算法求解模型。进一步地引入了信号灯控制规则，即根据信号灯信息来动态调整行驶时间，以减少能源消耗，并且消除出行者的等待焦虑。. （3）开展了基于所提出模型和算法的实证研究，为快递物流企业使用电动汽车进行城市配送提供了强有力的理论基础和技术支持。通过对某物流企业的实地调研，深入了解了电动汽车城市物流配送的业务流程及资源约束条件，如随时插入任务、配送途中方案重规划及对求解方案的鲁棒性要求等。进一步地，通过仿真试验及实证分析，逐步修正提出的带时间窗和取送货的配送路径优化模型及变邻域搜索算法，为某物流企业开发了能够有效求解配送方案的VC++软件包。. 上述研究为电动汽车城市物流配送路径优化提供了强有力的理论基础和有效方法。