基于配送地点变化的物流路径优化研究

发布时间：2018-02-06 05:11

本文关键词： 车辆路径问题配送地点变化干扰管理蚁群算法　出处：《杭州电子科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：当今电子商务正在飞速发展和普及,物流配送成为了影响电子商务发展至关重要的一个环节,车辆路径问题便是研究如何规划配送路线的问题。然而在实际配送过程中,配送车辆会遇到一系列未知的干扰事件,如客户需求量变化、配送地点变化、交通事故、天气变化等,使得车辆无法继续按照原有配送方案执行任务,因此需要快速的生成一个调整方案,来指导车辆应对这些未知问题。目前,动态车辆路径问题便是针对这一问题所开展的研究,但是调整策略往往只是以配送费用最低为目标来重新规划路线,却忽略了客户与配送员的利益。因此,本文结合上述问题和目前车辆路径问题的研究现状,运用干扰管理思想,来制定新的调整方案,以满足客户、配送员、物流公司三方面的需求。首先,本文结合车辆路径问题和实际研究需要,设计了初始配送问题,并根据问题假设,建立了初始配送模型。选取客户配送地点发生变化这一现象作为干扰事件,分别从客户、物配送员,物流公司三个角度分析客户配送地点发生变化给配送计划带来的扰动,进行相应度量方法设计。以此构建了基于客户不满意度低、物流公司配送成本少、配送路线偏离程度最小的多目标配送干扰管理模型。其次,为了达到快速模型求解,满足实际应用需求的目的,本文选取蚁群算法进行问题求解。由于蚁群算法存在一些缺陷和不足,本文主要从转移概率函数、信息素更新策略和局部优化三个方面进行改进设计。考虑到算法参数影响,本文以初始配送问题为研究对象,通过仿真实验确定蚁群算法关键性参数的取值。最后,为了验证模型和求解算法的有效性和实用性,本文在MATLAB软件上,通过算例求解分别与遗传算法、退火算法、混合粒子群算法进行比较分析,验证算法性能的高效性。通过与重调度法进行对比,表明本文路线调整策略的可行性和有效性。本文提出的基于配送地点变化的物流路径优化策略,可以有效的兼顾多方利益,生成扰动较小配送方案,对动态车辆路径问题、干扰管理、启发式算法方面的研究有一定的参考价值。
[Abstract]:Nowadays, electronic commerce is developing rapidly and popularizing. Logistics distribution has become a crucial link that affects the development of electronic commerce. The vehicle routing problem is to study how to plan the distribution route. However, in the actual distribution process, the distribution vehicle will encounter a series of unknown interference events, such as the change of customer demand and the change of distribution location. Traffic accidents, weather changes and so on, make the vehicle can not continue to carry out the tasks according to the original distribution plan, so it is necessary to quickly generate an adjustment scheme to guide the vehicle to deal with these unknown problems. The dynamic vehicle routing problem is a research on this problem, but the adjustment strategy is usually aimed at the lowest distribution cost to re-plan the route, but ignore the interests of the customer and the distributor. This paper combines the above problems and the current research status of vehicle routing problem, using the idea of interference management, to formulate a new adjustment scheme to meet the customer, distribution staff, logistics company three aspects of the needs. First of all. Combined with the vehicle routing problem and the actual research needs, this paper designs the initial distribution problem, and establishes the initial distribution model according to the assumption of the problem. The phenomenon of customer distribution location change is selected as the interference event. From three angles of customer, material distributor and logistics company, this paper analyzes the disturbance caused by the change of customer distribution location to the distribution plan, and designs the corresponding measurement method to construct the low customer dissatisfactory degree. The multi-objective distribution interference management model with less cost and minimum deviation of distribution route. Secondly, in order to solve the model quickly, to meet the needs of practical applications. In this paper, ant colony algorithm is selected to solve the problem. Due to some shortcomings and shortcomings of ant colony algorithm, this paper mainly from the transfer probability function. The pheromone updating strategy and local optimization are improved. Considering the influence of algorithm parameters, this paper takes the initial distribution problem as the research object. Finally, in order to verify the effectiveness and practicability of the model and the algorithm, this paper is on the MATLAB software. The algorithm is compared with genetic algorithm, annealing algorithm and hybrid particle swarm optimization algorithm to verify the efficiency of the algorithm, and compared with the rescheduling method. The results show that the route adjustment strategy is feasible and effective. The logistics route optimization strategy based on the change of distribution location can effectively take into account the interests of many parties and generate a less disturbance distribution scheme. The research on dynamic vehicle routing problem, disturbance management and heuristic algorithm has some reference value.
【学位授予单位】：杭州电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F252

【参考文献】