基于UWB/DGNSS的高速公路车辆协同定位方法研究

目前高速公路车辆碰撞事故发生率居高不下，但相关研究成果主要集中在基于前后距离测量的单车碰撞预警，无法避免车辆连环碰撞。车辆连环碰撞预警存在的主要技术难题是车辆的精确定位问题，这是因为高速公路存在缺少电源、车辆高速移动、不相邻车辆的非视距，以及海量设备安装维护的低成本要求等限制。为此，本课题提出了基于UWB无源RFID相对定位和基于自差分卫星的绝对定位方法，二者协同工作实现高速公路车辆的精确定位。在适应高速公路特殊性方面拟通过信号指纹、卡尔曼滤波等解决非视距问题，通过无源RFID解决电源问题，通过RF-Tag簇解决低成本问题，通过多普勒频移解决高速移动问题。在提高定位精度方面拟通过估计UWB信道品质因子、设定参考节点的权值来提高UWB定位精度；同时根据该定位结果可准确地推算出GNSS差分参考点的地理坐标，从而提高卫星定位的精度。研究成果可用于车辆连环碰撞预警、物联网及其他高精度位置服务领域。

Currently，highway vehicle collision accident rate remains high, but the related research is mainly focused on single-vehicle collision warning based-on ranging of adjacent vehicles, which is not a good way to avoid the chain collision. The main technical challenge of chain collision warning is precise positioning of vehicles. This challenge arises from special restrictions on highway such as lack of power supply, high vehicle speed, non-line-of-sight of the nonadjacent vehicles, low-cost requirement of massive equipment installation and maintenance, etc. In order to address this challenge, a hybrid positioning approach is proposed in this project, that is, both the relative positioning based on UWB passive RFID and the absolute positioning based on the self-differential satellite signal work collaboratively, and then the precise positioning of the vehicle on highway will be achieved.In the aspect of adapting to the highway particularity, the key elements of the proposed approach are using signal fingerprints and Kalman filters to solve the problem of non-line-of-sight, passive RFIDs to solve the energy problem, RF-Tag clusters to solve the low-cost problem and Doppler shift to solve the vehicle high speed problem. Further more, in the aspect of improving the positioning accuracy, estimating UWB Figure of Merit and setting weights of different references are employed to improve the accuracy of UWB positioning; At the same time, based on the estimated UWB positioning results, the geographical coordinates of the reference point for differential satellite positioning can be accurately deduced, and in this way the positioning precision of the DGNSS will be improved. The research results can be used for vehicle chain collision warning, Internet of Things and other areas of high-precision location-based services.