【作者】 李贻斌；

【导师】 刘鲁源；

【作者基本信息】 天津大学 ， 电力电子与电气传动， 2008， 博士

【摘要】 智能交通系统(Intelligent Transport System,ITS)可以显著提高运输效率,降低运输成本,减少能耗与排放,增强车辆运行安全性能,是未来交通运输方式的发展方向。智能车辆(Intelligent Vehicle,IV)是ITS核心和主要组成部分,更是未来汽车的发展方向。本文针对高速公路环境下IV的关键技术进行研究,包括状态检测与环境感知、建模,运动控制方法和算法,底层控制子系统和全局路径导航等。针对轮式车辆具有的典型非线性、强耦合、参数时变特性而导致车辆运行动态范围大、不易进行控制的问题,提出以空间压缩维代替时间维建立IV运动模式空间的思想。利用聚类技术构造IV的运动模式空间和子空间,确定车辆动态范围,建立了IV运动模式空间模型,为车辆动力学研究开辟了新思路,具有十分重要的理论意义和实践价值。在噪声加入、关键参数选择、参考轨迹设置、随机初始状态设置等算法的支持下,发展了“基于模式的智能控制器设计方法”,提供了一种新的复杂对象控制器设计手段PBICD。对利用PBICD设计的IV运动控制器进行了仿真研究,并对设计的ATRV移动机器人运动控制器进行了实验研究。通过仿真找出了对IV运动控制性能具有重要影响的车辆和执行机构参数,尤其是明确了刹车器特性对运动控制性能的主要影响因素,为相关IV集成系统的开发提供了依据,具有十分重要的意义。针对IV运行环境中有较多节点和道路的大网络,使用Mapinfo开发了其数字电子地图,提出了一种局部最优方向的A*启发式搜索算法,进行全局最短路径规划。结合济南城市交通,利用Mapinfo和Visual Basic 6.0作为仿真工具,证实这种算法可以快速地找到最短路径,减少了计算时间并实现可视化操作。分别采用MATLAB语言或VB语言开发了基于计算机视觉的车道线识别、车辆运动模式分析、车辆运动控制、车辆导航与路径规划等应用软件,为相关技术的研究提供了研究手段。利用软件集成技术,采用黑盒复用的方法,将MATLAB和VB编写的不同构件集成,实现了“智能车辆运动模式分析和运动控制器自动设计”的运行和测试软件平台,为IV的研究提供了一个较理想的开发环境。更多还原

【Abstract】 Intelligent transport system (ITS) can increase transportation efficiency, decrease transportation cost and energy consumption, and improve vehicle’s safety. ITS is the research direction of transportation. Intelligent vehicle (IV) is the core of ITS and the research direction of future automobile. Aiming at some key techniques of intelligent vehicle on the highway, the paper has some researches on state detection and environment perception, modeling, motion control algorithms, bottom control subsystems and global path planning and navigation.Due to some characteristics of vehicle’s motion, such as nonlinear, strong coupling, time-variant parameters, it is very difficult to control vehicle’s motion. The paper presents a new idea that with spatial compression dimension replacing temporal dimension, motion mode space of vehicle is established. Clustering technique is applied to establish the motion model space and sub-space, confirm the scope of vehicle motion, and model intelligent vehicle’s motion mode space. The paper pioneers the research of vehicle’s dynamics and has very important theoretical significance and practical worthiness.With noise addition, key parameters selection, reference path setting, and random initial state setting, we design the mode-based intelligent controller and provide a new design method PBICD for complex object’s controller. We do some simulation for intelligent vehicle’s motion controller designed by PBICD and do some experiments for ATRV mobile robot’s motion controller. With simulation, we find the key parameters of vehicle and execution system which are important for IV’s motion control performance. Especially, we make sure that the characteristics of brake is the key impact factor of motion control performance. The result provides some basis for the development of IV integration system. It has very important significance.Aiming at the IV’s environment is a large network with many nodes and roads, Mapinfo is applied to develop the digital e-map. A local optimal A* heuristic searching algorithm is presented for global shortest path planning. Combining with Jinan city transportation, Mapinfo and Visual Basic 6.0 are used as simulation tools. The simulation results prove that the algorithm can search the shortest path quickly and decrease the searching time. We develop some application software, such as computer vision based roadway line identification, vehicle motion mode analysis, vehicle motion control, vehicle navigation and path planning with MATLAB or VB. This software provides tool for research on some related key techniques. With software integrating technique, black-box duplication, various software developed by MATALB or VB are integrated to realize“intelligent vehicle motion mode analysis and motion controller auto designing”software. The software provides an ideal development environment for research of intelligent vehicles.更多还原