交通路口基于时空网格的车辆碰撞预警方法

发布时间：2018-02-16 10:49

本文关键词： 交通路口车辆碰撞预警道路网格速度建议　出处：《大连理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：城市交通系统中,交通路口是一个重要节点。交通路口由于车流量大,交通状况复杂,成为了交通安全事故的高发地,严重地影响了车辆的通行安全和车辆通过效率的提高。因此,交通路口的安全成为了一个亟待解决的问题。随着无线通信技术和定位技术的发展,车-路之间进行安全可靠的通信在今天已成为可能。车路协同已成为解决交通路口安全通行问题的主要技术手段。通过车路协同技术构造的车辆安全预警系统能够保证交通路口中车辆的安全,同时还能提高交通路口的通行效率。本文基于车路协同技术,提出了交通路口基于道路时空网格的车辆碰撞预警方法。我们将交通路口划分为多个时空网格,路侧终端根据车辆的当前运动状态模拟车辆的行驶轨迹,确定车辆在不同时刻占用的网格数。若在未来某一时刻存在不同车辆占用相同的网格,那么碰撞就会发生。在判断出车辆的潜在碰撞后,系统还要根据车辆的当前运动状态,给出碰撞车辆的安全速度建议。为了更加精确地预测车辆的运行轨迹,本文还提出了基于扩展卡尔曼滤波的车辆位姿估计精确度提高方法。通过不断地进行预测和更新来提高车辆位姿估计的精度,更加准确的地估计车辆的运行轨迹,进而提高安全预警的准确性。为了对算法进行验证,我们在有信号和无信号交通路口环境下对算法进行了仿真实验。实验结果表明在GPS数据存在误差的情况下,提出的算法能够对大部分碰撞冲突进行预警,但存在遗漏检测的情况；在GPS数据经过平滑处理,车辆位姿估计较为精确的情况下,算法能够有效地检测出交通路口的碰撞,并提出精确的速度建议,具有较高的准确性。
[Abstract]:In the urban traffic system, traffic intersection is an important node. Because of the heavy traffic flow and complicated traffic situation, the intersection has become a high incidence of traffic safety accidents. It seriously affects the safety of vehicles and the improvement of the efficiency of traffic passage. Therefore, the safety of traffic junctions becomes a problem to be solved urgently. With the development of wireless communication technology and positioning technology, The safe and reliable communication between vehicles and roads has become possible today. Vehicle-road coordination has become the main technical means to solve the problem of safe passage at traffic junctions. The vehicle safety early warning system can be constructed by means of vehicle-road coordination technology. Enough to ensure the safety of traffic vehicles, At the same time, it can improve the efficiency of traffic intersection. In this paper, based on the vehicle-road cooperation technology, a vehicle collision early warning method based on the road space-time grid is proposed, and the intersection is divided into several space-time grids. According to the current moving state of the vehicle, the road side terminal simulates the vehicle's moving track and determines the number of grids occupied by the vehicle at different times. If there are different vehicles occupying the same grid at a certain time in the future, After judging the potential collision of the vehicle, the system should give the safety speed suggestion according to the current motion state of the vehicle. In order to predict vehicle trajectory more accurately, this paper also proposes a method to improve the accuracy of vehicle pose estimation based on extended Kalman filter, which can improve the accuracy of vehicle pose estimation by continuous prediction and updating. More accurate estimation of vehicle trajectory, and then improve the accuracy of safety warning. In order to verify the algorithm, the algorithm is simulated at signalized and unsignalized traffic junctions. The experimental results show that there are errors in the GPS data. The proposed algorithm can warn most collision conflicts, but there exists the situation of missing detection. When the GPS data is smoothed and the vehicle pose estimation is more accurate, the algorithm can effectively detect the collision at traffic junctions. And put forward the accurate speed suggestion, has the higher accuracy.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491

【参考文献】