【作者】 何智伟；

【导师】 吴青；

【作者基本信息】 武汉理工大学 ， 物流管理， 2009， 硕士

【摘要】 在智能车路协调环境下实现汽车列队行驶可以提高道路车辆密度,增加道路容量,同时能简化交通控制复杂度,增加交通的可控性,提高交通安全。汽车编队则是由汽车列队行驶派生出来的一种非常重要的车队驾驶行为,它与车辆离队、车队换道等车队驾驶行为结合起来能极大提高了汽车列队的柔性与灵活性,能使汽车列队行驶的优势得到充分发挥。本文将对智能车路环境下的车辆编队控制进行建模讨论。车辆编队控制与车队纵向行驶控制具有很大关联性,从某种意义上讲,车辆编队控制是建立在车队纵向行驶控制的基础上的,故本文在讨论车辆编队控制之前对车队纵向驾驶控制进行了讨论分析。将双层自适应模糊控制算法运用于车队纵向驾驶控制,能较好地抑制车队间距波动和速度波动,有效地提高了车队列队纵向行驶的稳定性,为车辆编队创造了良好的环境,极大减少了干扰。本文应用MAS的相关理论和方法来分析车辆编队控制问题,构建了基于MAS的车辆编队模型,并对其运行流程进行详细阐述。此外,在建模过程中还引入了Agent角色概念来降低模型复杂度和实现难度,提高模型效率。协调是MAS模型中非常重要组成部分。在比较分析多种多智能体协调机制优缺点后,根据编队模型的协商特点,将合同网协议(CNP)和通用部分全局规划(GPGP)有机结合起来构建了编队模型的协调框架。该框架能够充分利用CNP的任务分配动态性和灵活性来实现任务的合理分配,充分利用GPGP来考虑任务之间的关联和Agent之间的关联来减小对任务分解耦合度的依赖,从而增强了系统开放性和鲁棒性,优化了系统整体活动性能。对车辆编队实现过程中若干关键问题进行了讨论。将虚拟车思想和车辆纵向驾驶控制结合起来较好地实现了车辆相对位置定位控制;讨论并计算了几种典型车辆编队情况所需的换道空间长度;采用基于多项式曲线算法对车辆换道路径进行规划。此外,还对车辆编队仿真的程序实现进行探讨,并对几种典型的车辆编队情况进行了仿真,仿真结果表明了车辆编队模型的可行性。更多还原

【Abstract】 In intelligent vehicle-road system environment,vehicle platoon driving not only can increases the density of vehicles and augments road capacity,but also simplifies the complexity of traffic control,increases the controllability of traffic and enhances road safety.Vehicle merging is a very important platoon driving behavior,derived from vehicle platoon driving.Combined with vehicle splitting and vehicle changing lane,it will be able to greatly improve flexible and flexibility of the platoon driving, and will enable the advantages of vehicle platoon driving moving into full play.In this paper,the control model of vehicle merging is discussed under intelligent vehicle-infrastructure system environment.There is great relevance between the longitudinal control of the platoon and the control of vehicle merging,in a sense,the control of the vehicle merging is built on the basis of the longitudinal control of the platoon.So in this paper the longitudinal control of the platoon were discussed before the control of vehicle merging.The two-layer adaptive fuzzy control algorithm is applied to the longitudinal control of the platoon driving,which can better suppress fluctuation of headway distance and speed,and improve the stability of the platoon driving effectively.So Interference is reduced greatly;a good environment is created for the vehicle merging.In this paper,the theories and methods of MAS(multi-agent system) are applied to analyze the control of the vehicle merging.A model of vehicle merging based on the MAS is built,and the process of the model running is expounded.In addition,the concept of Agent role model is introduced to reduce the complexity of model and the difficulty to achieve,improving the efficiency model in the modeling process.The coordination mechanism is the most important part of the MAS model. After a comparative analysis the advantages and disadvantages of the coordination mechanism of MAS,in accordance with the consultation characteristics of vehicle merging model,the coordination framework model is built by combined the contract net protocol(CNP) and the generalized partial global planning(GPGP) in this paper. In the framework,tasks are allocated reasonably by given full play of the dynamic and flexibility of task distribution of CNP,and the Dependence of coupling degree of task decomposition is reduced by given full play of GPGP to thinking about the association between tasks and the association between agents,to enhance the open and robustness of system and optimize the overall system performance.Finally,some key issues about the implementation of vehicle merging are discussed.By combined the concept of virtual-car with vehicle longitudinal control, the control of vehicle relative position is realized better.Road space length needed for vehicle lane-changing is discussed and calculated in several typical Situations of vehicle merging.The curve algorithm based on polynomial is applied for Vehicle lane-changing path planning.In addition,the implementation of simulation program of vehicle merging is explored,and some simulation of several typical situations of vehicle-merging is carried out,simulation results show the feasibility of the vehicle merging model.更多还原