基于启发式智能算法的多层网络路径规划研究  

Multilayer Network Pathway Planning Research Based on a Heuristic Intelligence Algorithm

作　　者：吴梦瑶 李思禹 熊宇帆 杨欣彤 李文卓 

机构地区：[1]长春理工大学数学与统计学院,吉林 长春 [2]长春理工大学计算机科学技术学院,吉林 长春 [3]长春理工大学物理学院,吉林 长春

出　　处：《应用数学进展》2024年第1期453-465,共13页Advances in Applied Mathematics

摘　　要：为了确保物流周转效率和质量,应当借助互联网以及大数据的发展实现物流管理的优化提升,针对物流配送过程中存在的仓储布局、物流路线优化等问题,构建面向物流配送的多品种物流配送的VRP模型,采用改进的遗传算法对其进行求解,并通过实例进行验证。本课题拟将物流配送系统中的智能调度问题建模,并将其转换为一类基于启发式智能算法的物流网络路径优化问题,以物流调度需求为输入变量,以物流系统中的物流调度为目标,设计物流调度系统中物流系统的物流调度优化方案。通过对全国公路、铁路、水路三大交通网络的优化设计,使车辆从生产基地先运至前置仓库,再运至4S门店。针对以上问题,本文将构建以仓储位置为对象的图聚类模型,在物流费用与时间最优相矛盾且不能兼顾的情况下,构建多目标优化模型,并利用GPU并行遗传算法求解。In order to ensure the efficiency and quality of logistics turnover, the optimization of logistics man-agement should be realized with the help of the Internet and the development of big data. Aiming at the problems existing in the process of logistics distribution, such as warehousing layout and lo-gistics route optimization, a multi-variety logistics distribution VRP model for logistics distribution is constructed, which is solved by improved genetic algorithm and verified by an example. This top-ic intends to model the intelligent scheduling problem in the logistics distribution system, and transform it into a kind of logistics network path optimization problem based on heuristic intelli-gent algorithm, taking the logistics scheduling demand as the input variable, taking the logistics scheduling in the logistics system as the goal, design the logistics scheduling optimization scheme of the logistics system in the logistics scheduling system. Through the optimization design of the na-tional highway, railway and waterway transportation networks, the vehicles are first transported from the production base to the front warehouse, and then to the 4S stores. In view of the above problems, this paper will build a graph clustering model based on warehousing location. Under the condition that logistics cost and time optimization are contradictory and can not be taken into ac-count, a multi-objective optimization model is constructed and solved by GPU parallel genetic algo-rithm.

关 键 词：遗传算法聚类 VRP 车辆配送 

分 类 号：F25[经济管理—国民经济]