考虑碳排放的多车场多车型VRP模型及算法研究

发布时间：2018-01-04 03:42

本文关键词：考虑碳排放的多车场多车型VRP模型及算法研究　出处：《深圳大学》2017年硕士论文　论文类型：学位论文

【摘要】：国际能源机构声明,交通运输是第二大二氧化碳排放来源,而交通运输中几乎3/4的CO2排放来自于公路运输,因此,降低在车辆运输中所产生的二氧化碳排量是非常必要的。如今物流配送网络日益发达,为扩大企业业务,更及时地满足客户需求,许多企业都建立多个配送中心,并提供多种车型来适用各种不同类型的货物。为使研究更趋于现实,具有更大的实用价值,论文研究了考虑碳排放的多车场多车型车辆路径问题。本文在国家自然科学基金(71571120,71271140,71471158)、广东省自然科学基金(2016A030310074)、广东省促进科技服务业发展计划项目(2013B040403005)的资助下开展了如下研究:首先,本文构建了一个新的数学模型——考虑碳排放的多车场多车型车辆路径问题模型。论文对车辆路径问题的几种不同模型进行了分析研究,并针对碳排放问题,考虑了碳排放量的计算和碳交易机制。碳排放量的计算主要通过计算能源消耗量来获得,论文对车辆路径问题中能源消耗的几种计算模式进行了研究,并选用综合燃油消耗计算模式来计算配送过程中车辆的油耗量。对于不同类型的车辆,固定费用也会有所不同。模型还考虑了时间窗的约束条件,早到或者晚到都会产生时间窗惩罚费用。因此,模型主要研究的是在多车场多车型VRP模型考虑了碳排放因素,加入碳交易机制,使总费用达到最小。其次,论文对求解模型的算法进行了分析和改进。细菌觅食优化算法(BFO)是一种较新颖的群体智能优化算法,具有并行搜索、善于局部搜索的优点,而单循环结构的细菌觅食优化算法(SRBFO)降低了运算复杂度,具有更好的收敛性和优化效果。改进的综合学习粒子群算法(ECLPSO)采用了综合学习的机制,能够很大地提高解决方案的准确性。基于已有的算法研究成果,本文结合了单循环结构的细菌觅食优化算法(SRBFO)和改进的综合学习粒子群算法(ECLPSO)中的综合学习机制,构造了新的算法——单循环结构综合学习细菌觅食优化算法(SRCLBFO)。针对SRCLBFO算法,论文还分别选取了三组单峰函数和三组多峰函数进行有效性验证。最后,论文将算法应用到考虑碳排放的多车场多车型车辆路径问题实例中进行优化求解。通过实验结果对比,验证了模型和改进算法的有效性。同时,结合实例,论文还对单车场、多车场进行对比实验,证明多车场相对于单车场的优势;对单车型、多车型进行对比实验,验证多车型较单车型的有效性。此外,论文还对碳交易机制中碳价格和碳配额上下波动对费用、碳排放量和距离等因素的影响进行了分析。本文拓展了考虑低碳的车辆路径问题的研究,为企业实施低碳运输提供了一定的借鉴意义。
[Abstract]:Transport is the second largest source of carbon dioxide emissions, while almost 3/4 of CO2 emissions in transport come from road transport, the International Energy Agency said. It is necessary to reduce the carbon dioxide emissions generated in the vehicle transportation. Nowadays, the logistics distribution network is increasingly developed, in order to expand the business, more timely to meet customer needs. Many enterprises have set up multiple distribution centers, and provide a variety of models to apply to different types of goods. In order to make the research more realistic, it has more practical value. This paper studies the problem of multi-vehicle vehicle routing considering carbon emissions. This paper is based on the National Natural Science Foundation of China 71571120 71271140 71471158). Natural Science Foundation of Guangdong Province, 2016A030310074. The following research has been carried out with the support of the Guangdong Provincial Project for the Development of Science and Technology Services (2013B040403005): first of all. In this paper, a new mathematical model, multi-vehicle vehicle routing problem with carbon emissions, is constructed. Several different models of vehicle routing problem are analyzed and studied in this paper, and the carbon emission problem is analyzed. The calculation of carbon emissions and carbon trading mechanism are considered. The calculation of carbon emissions is mainly through the calculation of energy consumption. In this paper, several calculation models of energy consumption in vehicle routing problem are studied. A comprehensive fuel consumption calculation model is used to calculate the fuel consumption of vehicles in the distribution process. For different types of vehicles, the fixed costs will be different. The model also takes into account the constraints of the time window. Therefore, the main research of the model is to consider the carbon emission factors in multi-vehicle VRP model, add carbon trading mechanism to make the total cost to the minimum. This paper analyzes and improves the algorithm for solving the model. The bacterial foraging optimization algorithm (BFOO) is a novel swarm intelligence optimization algorithm with the advantages of parallel search and local search. The single cycle structure of bacteria foraging optimization algorithm (SRBFOO) reduces the computational complexity. The improved synthetic learning particle swarm optimization algorithm (ECLPSO) adopts the mechanism of integrated learning. Can greatly improve the accuracy of the solution. Based on the existing algorithm research results. This paper combines the integrated learning mechanism of single cycle structure bacterial foraging optimization algorithm (SRBFOO) and improved integrated learning particle swarm optimization (ECLPSO). A new algorithm, single cycle structure integrated learning bacteria foraging optimization algorithm, is constructed. The SRCLBFO algorithm is aimed at it. Three groups of single-peak functions and three groups of multi-peak functions are selected to verify the validity. Finally. In this paper, the algorithm is applied to solve the multi-vehicle vehicle routing problem with carbon emissions. The experimental results show that the model and the improved algorithm are effective. At the same time, an example is given. The paper also carries on the contrast experiment to the bicycle yard and the multi-car yard, proves the advantage of the multi-car yard compared with the single car yard; To verify the effectiveness of multi-vehicle model compared with single-vehicle model, the paper also makes a comparative experiment on the carbon price and carbon quota in the carbon trading mechanism. The effects of carbon emissions and distance are analyzed. This paper extends the research on the vehicle routing problem considering low carbon and provides a reference for enterprises to implement low-carbon transportation.
【学位授予单位】：深圳大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U492.3

【相似文献】