基于车路协同的车辆定位算法研究
Algorithm Based on Cooperative Vehicle Infrastructure Systems

为解决道路交叉口车辆由于定位信号缺失或者延迟引起的车辆定位偏差较大的问题，提出了基于车路协同的协同地图匹配算法（cooperative map-matching，CMM）.首先利用扩展Kalman滤波（extended Kalman filter，EKF）融合GPS与车载航位推算系统（vehicular dead reckoning，DR）信息作为协同地图匹配的预先定位；然后基于短程通讯技术实现车辆信息的交换与共享，在电子地图的基础上，利用道路约束实现车辆进一步定位.为了验证算法的有效性，搭建了模拟真实场景的仿真环境进行实验.研究结果表明：采用EKF融合GPS/DR数据的交叉口车辆定位平均偏差为9.09 m，相比GPS的14.31 m，定位偏差减小30.87%；采用CMM算法的交叉口车辆，当参与CMM车辆数为7时，平均位置偏差为4.5 m，参与CMM车辆数为10辆时，平均位置偏差为2.75 m，相比EKF定位偏差减小69.74%.
：In order to determine the problem of vehicle positioning deviation caused by a lack and delay of the positioning signal,a cooperative map-matching(CMM)algorithm,based on the cooperative vehicle infrastructure system,is proposed in this paper. First,the information obtained by GPS and that obtained by vehicular dead reckoning(DR)were fused to obtain the initial position of cooperative map-matching using an extended Kalman filter(EKF). Then,vehicle information was exchanged and shared based on dedicated short-range communication (DSRC). On the basis of an electronic map,the further positioning of vehicles was accomplished using road constraints. In order to verify the effectiveness of the proposed algorithm,an environment to simulate real scenes was set up to conduct the experiments. The experimental results demonstrate that the average positioning deviation of vehicles at intersections using EKF,which fuses data obtained from GPS and DR,is 9.90 m. The positioning deviation decreased by 30.87%,when compared with the average deviation of GPS,which is 14.31 m. The proposed CMM algorithm has an average position deviation of 4.5 m when the number of vehicles involved is 7,and 2.75 m when the number of vehicles involved is 10. The positioning deviation decreased by 69.74%