论文部分内容阅读
本文以大规模成品油二次配送路径规划为对象,研究了具有成品油物流特征的多车场带时间窗的车辆路径问题的数学模型,提出了新的基于子问题分解的两阶段优化算法.首先采用改进的系统聚类算法将配送需求合并分载,随后设计了改进的遗传算法生成最终的配送路径.针对客户划分问题,提出了‘期望节约里程’指标,用以描述客户的地理空间分布特征,并以此为特征信息设计了启发式的遗传算子,提升了大规模问题优化收敛的速度、质量和稳定性.仿真实验结果验证了模型和算法的可行性和有效性.
In this paper, the second-order route planning of large-scale refined oil products is taken as an object to study the mathematic model of vehicle routing problem with time windows in multi-depot with multi-depot logistics features and to propose a new two-stage optimization algorithm based on sub-problem decomposition. An improved system clustering algorithm is used to split the distribution requirements and then an improved genetic algorithm is designed to generate the final delivery route.According to the problem of customer division, the “expected mileage saving” index is proposed to describe the customer’s geospatial distribution characteristics The heuristic genetic operators are designed to improve the speed, quality and stability of optimal convergence of large-scale problems.The simulation results verify the feasibility and effectiveness of the proposed model and algorithm.