基于智能算法的车辆合乘研究与设计

发布时间：2018-01-13 12:56

本文关键词：基于智能算法的车辆合乘研究与设计　出处：《东华理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着城镇化建设和汽车产业的迅速发展,城市交通供给与交通需求之间的矛盾日益突出。为了缓解城市交通压力,车辆合乘的出行方式得到乘客青睐,路径优化和车辆合乘算法成为研究热点。目前大部分车辆合乘的网站和手机软件只提供信息发布和供求平台,不具备车辆和乘客的匹配功能,没有实现车辆合乘路径的优化。本文在研究车辆合乘问题的相关理论和信息技术的基础上,将智能处理算法与车辆合乘问题相结合,讨论了求解车辆合乘匹配问题和合乘路径优化问题的算法,从算法的原理、步骤和特点等方面探索了粒子群算法和遗传算法这两种智能算法在车辆合乘中的应用。针对带软时间窗的多车辆合乘问题,在保证车辆匹配率的基础上建立了以路径最短和费用最少为目标函数的数学模型,对超时情况采取相应的处理措施。求解多车辆合乘问题首先实现车辆和乘客的匹配,根据匹配的结果规划车辆合乘的路径。在解决车辆合乘匹配时以车辆的匹配率为目标函数,利用粒子群算法优化车辆所经过站点的搜索半径,使车辆与乘客的匹配率最高。车辆合乘匹配完成后,利用改进的遗传算法求解车辆最优合乘路径。针对本文的研究问题,从种群初始化、选择策略、交叉算子和变异算子这四个方面对遗传算法进行改进,加快了算法的收敛性,更易得到最优解。为了验证算法,本文设计了一个路网,路网中有多辆车和多名乘客,每个司机有一个初始路线和时间窗,乘客都有相应的时间窗和起终点。通过实验结果表明,粒子群算法能有效的解决车辆合乘匹配问题,建立的数学模型和改进的遗传算法能有效的解决车辆合乘路径优化问题。最后建立了一个基于android平台的车辆合乘系统,系统的客户端主要负责和用户交互、界面的显示,服务器端主要负责利用优化算法来实现合乘匹配和合乘路径优化。用户登录系统发布初始路线和时间窗,系统利用粒子群算法和改进的遗传算法进行司机和乘客的匹配和车辆合乘路径优化,并将匹配结果和优化路径显示在客户端上,验证了理论模型和算法设计的可行性。
[Abstract]:With the rapid development of urbanization and automobile industry, the contradiction between urban traffic supply and traffic demand is increasingly prominent. At present, most of the websites and mobile phone software only provide information publishing and supply and demand platform, and do not have the matching function between vehicles and passengers. In this paper, based on the research of the theory and information technology of the vehicle ride problem, the intelligent processing algorithm is combined with the vehicle ride problem. This paper discusses the algorithm to solve the vehicle matching problem and the optimization problem of the vehicle ride path, which is based on the principle of the algorithm. The application of particle swarm optimization (PSO) and genetic algorithm (GA) in vehicle combination is discussed in the aspects of steps and characteristics. On the basis of ensuring the matching rate of the vehicle, a mathematical model with the shortest path and the least cost as the objective function is established. Corresponding measures are taken to deal with the overtime situation. First, the matching of vehicles and passengers is realized by solving the multi-vehicle problem. According to the matching result, the vehicle path is planned. The vehicle matching rate is taken as the objective function, and the searching radius of the station through which the vehicle passes is optimized by using the particle swarm optimization (PSO) algorithm. After the vehicle matching is completed, the improved genetic algorithm is used to solve the optimal vehicle routing. In view of the problem of this paper, the strategy is selected from population initialization. The crossover operator and mutation operator improve the genetic algorithm, accelerate the convergence of the algorithm, and get the optimal solution more easily. In order to verify the algorithm, this paper designs a road network. There are many vehicles and many passengers in the road network, each driver has an initial route and time window, the passenger has a corresponding time window and terminal point. The experimental results show that. Particle swarm optimization algorithm can effectively solve the problem of vehicle matching. The established mathematical model and improved genetic algorithm can effectively solve the problem of vehicle ride path optimization. Finally, a vehicle riding system based on android platform is established. The client of the system is mainly responsible for interacting with the user, the interface is displayed, and the server is mainly responsible for using the optimization algorithm to realize the path optimization. The user login system issues the initial route and time window. Particle swarm optimization (PSO) and improved genetic algorithm (GA) are used to match the driver and passenger and optimize the vehicle ride path. The matching results and the optimized path are displayed on the client side. The feasibility of theoretical model and algorithm design is verified.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491

【相似文献】