带时间窗的多车场低碳车辆路径问题研究

发布时间：2018-12-14 13:14

【摘要】：传统的车辆路径问题(Vehicle Routing Problem, VRP)大多从经济效益视角考虑,研究目标为最小化旅行距离或物流成本,而从环境效益的角度考虑运输对环境影响的相关研究还比较少。物流和交通运输已成为碳排放的主要来源之一,是落实节能减排的重要领域。实际运输企业大多拥有多个车场,同时考虑到车场车辆数、车辆载重、车辆工作时间以及客户时间窗等实际约束,本文研究了一类新的车辆路径问题——带时间窗的多车场低碳车辆路径问题(Multi-Depot Low-Carbon Vehicle Routing Problem with Time Windows, MDLCVRPTW).该问题将研究目标转移到环境效益中去,是对传统的带时间窗的多车场车辆路径问题(Multi-Depot Vehicle Routing Problem with Time Windows, MDVRPTW)的扩展,通过合理安排车辆运输路线,以期减少能耗和碳排放。基于历史统计数据和已有研究成果,本文首先建立了车辆能耗和碳排放量的宏观计量模型,模型综合考虑了距离和重量这两个影响碳排放量的关键因素；在此基础上建立了MDLCVRPTW问题的数学优化模型,目标函数为最小化碳排放量。由于MDLCVRPTW属于NP-hard问题,本文设计了带适应性变异的混合遗传算法来求解该问题：设计了基于轮盘赌的初始解构造方法,采用了父代参与竞争的锦标赛选择机制,然后改进了Ox交叉算子,设计了基于适应性邻域搜索的变异操作,并通过种群管理保持种群多样性。为了评估算法的有效性,首先采用Cordeau的MDVRPTW标准测试算例数据,求解了MDVRPTW问题,从解的质量、算法稳定性以及收敛时间三方面表明了本文算法的有效性。然后将MDLCVRPTW模型计算结果与MDVRPTW模型计算结果相比较,结果表明,前者较后者旅行距离平均增加1.65%,但碳排放量平均降低0.90%,由此表明在小批量多批次货物配送中,适当增加旅行距离能减少碳排放量,同时也表明仅考虑旅行距离不利于节能减排。最后将本文的模型与算法应用于苏果社区店的实际配送作业中,给出了合理的运输路线安排。本文拓展了已有车辆路径问题的研究,能够为企业实施低碳运输提供重要的决策支持。
[Abstract]:The traditional vehicle routing problem (Vehicle Routing Problem, VRP) is mostly considered from the perspective of economic benefits. The research goal is to minimize travel distance or logistics cost, but there are few researches on the environmental impact of transport from the perspective of environmental benefits. Logistics and transportation have become one of the main sources of carbon emissions, which is an important field of energy saving and emission reduction. Most of the actual transportation enterprises have more than one vehicle yard, considering the actual constraints such as the number of vehicles in the yard, vehicle load, vehicle working time and customer time window, etc. In this paper, we study a new class of vehicle routing problem The low carbon vehicle routing problem with time windows (Multi-Depot Low-Carbon Vehicle Routing Problem with Time Windows, MDLCVRPTW).) This problem transfers the research target to the environmental benefit. It extends the traditional vehicle routing problem with time window (Multi-Depot Vehicle Routing Problem with Time Windows, MDVRPTW) and reasonably arranges the vehicle transportation route. To reduce energy consumption and carbon emissions. Based on the historical statistical data and the existing research results, this paper first established the vehicle energy consumption and carbon emissions macro measurement model, the model considered the distance and weight of the two key factors affecting carbon emissions; On this basis, the mathematical optimization model of MDLCVRPTW problem is established. The objective function is to minimize carbon emissions. As MDLCVRPTW belongs to the NP-hard problem, a hybrid genetic algorithm with adaptive mutation is designed to solve the problem. An initial solution construction method based on roulette is designed. Then the Ox crossover operator is improved and the mutation operation based on adaptive neighborhood search is designed and population diversity is maintained through population management. In order to evaluate the validity of the algorithm, the MDVRPTW problem is solved by using the MDVRPTW standard test data of Cordeau. The effectiveness of the proposed algorithm is demonstrated from three aspects: the quality of the solution, the stability of the algorithm and the convergence time. The results of MDLCVRPTW model are compared with those of MDVRPTW model. The results show that the former increases the travel distance by 1.65 on average, but the carbon emission decreases by 0.90 on average. A proper increase in travel distance can reduce carbon emissions, but it also shows that only considering travel distance is not conducive to energy saving and emission reduction. Finally, the model and algorithm of this paper are applied to the practical distribution of the community store of Suguo, and the reasonable transportation route is given. In this paper, the existing research on vehicle routing problem is extended, which can provide important decision support for enterprises to implement low-carbon transportation.
【学位授予单位】：南京农业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F259.22;U492.22

【相似文献】