基于RFID、视觉和UWB的车辆定位系统

摘要

在智能交通系统中获取车辆的精确位置对提升主动安全和实现无人驾驶具有重要意义.针对目前全球定位系统(GPS)和车路通信(V2I)定位方法的不足,结合射频识别(RFID)、视觉和超宽带(UWB)通信提出了一种新的车辆定位系统,目的是在GPS效果不佳的城市环境下实现车道级的定位.提出了基于RFID和视觉的车道判别算法和基于UWB的单锚点V2I定位算法.分析了单锚点V2I定位算法在典型工况下的定位误差.结果表明,在合理布置路侧单元的前提下采用该定位方法可实现直道和弯道定位误差分别在0.3和0.5m以内的定位精度.%To obtain the accurate location of vehicles in intelligent transportation systems has important sig-nificance to enhancing active safety and implementing autonomous driving. In view of the defects of present global positioning system ( GPS) and vehicle-to-infrastructures ( V2I) positioning technology, a novel positioning system is proposed by combining radio-frequency identification ( RFID) , vision and ultra-wideband ( UWB) communications, aiming at realizing lane-level positioning in urban environment with poor GPS effects. Both algorithms of lane judg-ment based on RFID and vision and single RSU V2I positioning based on UWB are put forward, with the error of single RSU V2I positioning in typical conditions analyzed. The results indicate that the proposed system with rational arrangement of RSUs can achieve a positioning accuracy with an error less than 0. 3m for straight road and 0. 5m for curved road respectively.