面向配送路径优化问题的传输学习和多目标自适应模因计算方法研究

With the development of online shopping market, the last mile Logistics distribution problem is highlighted. Thus, this study intends to establish a multi-objective optimization model to balance between economic efficiency and customer satisfaction for vehicle routing problem. Moreover, vehicle routing problem is also NP- hard problem. To deal with this issue, this project not only intends to design efficient Memetic algorithms, history information based search direction inference and transfer learning strategies to improve the efficiency of vehicle routing problem and the speed of solving. The main research task include: (1) a multi-objective optimization model and frame based on adaptive Memetic algorithm; (2) a search direction reasoning algorithm and frame based on all evolutionary history; (3) construction the related problem of vehicle distribution problem, and transfer learning mechanism and theory for multi-objective vehicle distribution problem. We hope that this project taking into account the needs from between the logistics business and the customers. What’s more, this project provides a quick and efficient solution for the Logistics distribution problem in the Internet environment. Finally, we hope that our research is benefit for the logistics companies to make the scientific decision in distribution optimization on last mile distribution problem.

随着网购市场的兴起，快递企业“最后一公里”物流配送路径优化问题日益凸显。本课题拟建立兼顾企业经济效益和客户满意度的车辆路径规划的多目标物流配送路径优化模型。同时物流配送优化问题还是NP-难问题，针对这一难题，本课题拟设计高效的自适应模因算法(Memetic Algorithm)、历史进化信息驱动的进化算法、传输学习策略来提高车辆路径规划问题的求解效率和求解速度。主要研究的内容包括：（1）基于自适应模因算法(Memetic Algorithm)的多目标优化算法与框架；（2）基于全部进化历史信息的搜索方向推理算法与框架；（3）车辆路径优化相关问题的构造及多目标模型下这些相关问题间传输学习机制设计及理论。本课题期望能够兼顾物流企业与客户两方面的需求，为快速有效地解决互联网环境下的物流配送路径优化问题提供借鉴，为第三方物流企业的“最后一公里”配送优化决策提供科学依据。

随着网购市场的兴起，快递企业“最后一公里”物流配送路径优化问题日益凸显。本课题拟建立兼顾企业经济效益和客户满意度的车辆路径规划的多目标物流配送路径优化模型。同时物流配送优化问题还是NP-难问题，针对这一难题，本课题拟设计高效的自适应模因算法(Memetic Algorithm)、历史进化信息驱动的进化算法、传输学习策略来提高车辆路径规划问题的求解效率和求解速度。主要研究的内容包括：（1）基于自适应模因算法(Memetic Algorithm)的多目标优化算法与框架；（2）基于全部进化历史信息的搜索方向推理算法与框架；（3）车辆路径优化相关问题的构造及多目标模型下这些相关问题间传输学习机制设计及理论。本课题期望能够兼顾物流企业与客户两方面的需求，为快速有效地解决互联网环境下的物流配送路径优化问题提供借鉴，为第三方物流企业的“最后一公里”配送优化决策提供科学依据。