电动汽车路径规划建模与优化研究

发布时间：2018-05-07 15:36

本文选题：电动汽车路径规划 + 改进粒子群算法　；参考：《东华大学》2017年硕士论文

【摘要】：随着互联网的发展,城市物流配送逐年增多,出现城市交通状况拥堵以及尾气污染等现象,城市物流的外部不经济逐渐引起了大家的重视。以石油为主要燃料的交通领域,正面临着全球范围内资源紧缺与环境污染的挑战,加快推进普及新能源交通工具,成为全世界实现交通领域可持续发展的共识。受到购买新能汽车给予多项补助以及要求物流运输企业采用的车辆30%必须为电动汽车等相关政策的推动。我国新能源汽车尤其是能实现零排放的电动汽车,因为能源利用率高,技术相对成熟,成为我国新能源汽车发展和汽车工业转型的主要战略方向。随着电动汽车越来越多地出现在车辆调度系统中,尤其是电动出租车调度系统。电动汽车的一些特性,比如:续航里程短,充电时间长,充电服务设施少以及制动时动能可转化为电能储存在电容里等,使得电动汽车调度过程中需要考虑电动汽车的充电时间,充电站位置,电动汽车的容量限制,对传统车辆调度理论有很大的冲击。传统车辆调度理论并不适合指导电动汽车路径规划问题,人们迫切地需要以电动汽车为研究对象的相关路径规划理论为指导依据。本文在已有研究的基础上,以电动出租车为研究背景,考虑电动出租车独占性特性,首先研究了基于客户满意度的独占性多目标带时间窗口的电动汽车路径规划问题(Multi-Objective Exclu-sive Electric Vehicle Routing Problem with Time Window,以下简称MOEE-VRPTW问题),建立了基于客户满意度的MOEE-VRPTW混合整数规划模型,然后提出了通过增加粒子共享信息类型来提高粒子全局搜索能力的改进粒子群算法,并从粒子更新方向的角度分析其原理。再次根据司机是否为理性人,提出了两种充电策略:(1)司机为理性人——任务导向型充电策略;(2)司机为非理性人——最低百分点充电策略。并分析了这两种充电策略对上述模型的影响,以及给出这两种充电策略的适用情况。通过仿真测试将改进粒子群算法应用到模型的求解中,仿真测试结果表明,改进粒子群算法能够克服标准粒子群算法易陷入局部最优的缺点。并进行了参数分析,以及不同的充电策略对模型求解的影响分析。最后在MOEE-VRPTW模型的基础上,给出客户信息随时间逐渐到来的动态MOEE-VRPTW模型,并提出了结合时间驱动和事件驱动的滚动时域算法来求解该模型,仿真测试结果表明该算法能快速求解该问题。
[Abstract]:With the development of the Internet, the urban logistics distribution is increasing year by year, and the phenomenon of urban traffic congestion and exhaust gas pollution appears, so the external diseconomy of urban logistics has been paid more and more attention. The transportation field with petroleum as the main fuel is facing the challenge of the shortage of resources and environmental pollution in the global scope. It has become the consensus of the world to realize the sustainable development of the transportation field to speed up the popularization of new energy vehicles. Driven by policies such as multiple subsidies for the purchase of NNCs and the requirement that 30 per cent of vehicles used by logistics transport companies must be electric vehicles. China's new energy vehicles, especially electric vehicles, which can achieve zero emissions, have become the main strategic direction for the development of new energy vehicles and the transformation of automobile industry because of their high energy efficiency and relatively mature technology. Along with the electric vehicle more and more appear in the vehicle dispatch system, especially the electric taxi dispatch system. Some of the features of electric vehicles, such as short range, long charging time, less charging facilities, and braking kinetic energy can be converted into electric energy stored in capacitors, etc. It is necessary to consider the charging time, charging station location and capacity limitation of electric vehicle in the process of electric vehicle scheduling, which has a great impact on the traditional vehicle scheduling theory. The traditional vehicle scheduling theory is not suitable to guide the path planning of electric vehicles, and people urgently need to take the related path planning theory of electric vehicles as the guidance basis. In this paper, based on the existing research, taking electric taxi as the research background, the exclusive characteristics of electric taxi are considered. In this paper, the path planning problem of electric vehicle with time window based on customer satisfaction is studied. The MOEE-VRPTW problem is referred to as Multi-Objective Exclu-sive Electric Vehicle Routing Problem with Time Window, and the MOEE-VRPTW mixed integer programming model based on customer satisfaction is established. Then, an improved particle swarm optimization algorithm is proposed to improve the global searching ability of particles by increasing the type of information shared by particles, and the principle of the algorithm is analyzed from the point of view of the direction of particle renewal. Thirdly, according to whether the driver is a rational person, two charging strategies are proposed: 1) driver is a rational human-task-oriented charging strategy & # 2) the driver is an irrational person-the lowest percentage point charging strategy. The influence of these two charging strategies on the above models is analyzed, and the application of these two charging strategies is given. The simulation results show that the improved particle swarm optimization algorithm can overcome the shortcoming that standard particle swarm optimization algorithm is easy to fall into local optimum. The parameter analysis and the influence of different charging strategies on the solution of the model are also carried out. Finally, on the basis of MOEE-VRPTW model, a dynamic MOEE-VRPTW model with customer information coming along with time is given, and a rolling time-domain algorithm combining time-driven and event-driven is proposed to solve the model. The simulation results show that the algorithm can solve the problem quickly.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U492.3

【相似文献】