节点文献
无人驾驶智能车的导航系统研究
The Research of Unmanned Intelligent Car Navigation System
【作者】 王培;
【导师】 谭宝成;
【作者基本信息】 西安工业大学 , 控制理论与控制工程, 2012, 硕士
【摘要】 随着我国汽车行业的不断发展和进步,城市交通也遭遇了前所未有的拥挤。而车载导航系统的广泛应用则改善了这一问题。路径规划作为导航系统的重要组成部分,指导着驾驶员按照一条最安全、最快速的路径行驶,节约了汽车在公路上的逗留时间,提高了行车效率。无人驾驶智能车通过传感器采集的自身及环境的各种积极信息数据,然后传入车载中央计算机进行数据分析,并产生相应的控制策略,操纵无人驾驶智能车按照中央计算机计算所产生的最优路径行驶。论文首先说明了无人驾驶智能车研究的目的和意义,并对国内外无人驾驶智能车技术的研究和概况进行了介绍。通过研究车载导航系统的基本组成和常用的导航方式,对该车载导航系统提出了相关的技术要求。然后讨论了路网的表达方式及图论的相关知识,进而建立了交通道路网络的数学模型。同时在该交通路网模型的基础上,通过比较各种算法的优缺点,建立合理的估价函数及临界条件等,选择合适的搜索策略,寻找该路网上从起始点到目标点的最优路径。最后在Matlab环境下进行仿真,对该策略进行了最优性的验证。通过Matlab仿真实验,对改进后的双向A*算法和A*算法进行了比较。实验结果表明,改进后的算法在搜索时间和搜索得到的路径距离上都明显优于A*算法,说明改进后的算法实时性更好,能够搜索到最优路径的可能性更大。该搜索策略应用于车载导航系统中的路径规划,减少了搜索时间和存储空间,提高了效率。
【Abstract】 With development and progress of China’s auto industry, urban traffic suffered an unprecedented crowded. And the car’s navigation system which is widely used improves the problem. Path planning as an important part of navigation system is guiding the safest, most quick path running, it saves the car on the road linger time, and improves the efficiency of driving. Unmanned intelligent car collects all kinds of positive information data from itself and environment through their own sensors, passes the data into the central computer to analyze, and produces the corresponding control strategy, manipulates unmanned intelligence car running though the optimal path which is produced by the central computer.Firstly it explains the purpose and significance of research unmanned intelligent car, and introduces the technology situation of the unmanned intelligent car at home and abroad. And it puts forward relevant technical requirements to the navigation system through the study of the basic composition and the commonly used navigate mode of the navigation system. Then it expounds the basic composition and the commonly used navigate mode of the navigation system, and puts forward to the relevant technology requirements of the navigation system. Then it discusses the road network of the expression and the knowledge of graph theory, and sets up the mathematical model of road network traffic. And in based on the mathematical model of traffic network, the paper chooses appropriate search strategy by comparing the advantages and disadvantages of various algorithm, and builds a reasonable appraisal function and critical condition, looking for the optimal path which online the starting point to target, and simulates in the Matlab environment, validates the optimum of the strategies.Through the Matlab simulation experiments, it compares the improvement of the A* after bidirectional search algorithm and the A* algorithms. The experimental results show that the improved algorithm in search of time and search the path of the distance to get significantly better than A* algorithm, explain the improved real-time better, more likely can search for the optimal path. This search strategy is applied to the car’s navigation system of path planning, reduce the search time and storage space and improve efficiency.
【Key words】 navigation system; path planning; A* algorithm; appraisal function; bidirectional search algorithm;