基于UWB单基站系统/INS紧组合的智能车定位方法

Intelligent vehicle positioning method based on UWB single base station system/INS tight combination

针对狭长空间环境下测距组网超宽带(UWB)的定位几何构型较差导致定位精度低的问题,提出了一种基于UWB单基站系统/惯性导航系统(INS)紧组合的智能车定位方法。首先,通过标定UWB单基站系统的距离和角度观测量,提高其数据精度;然后,根据UWB距离和角度观测量建立量测方程,构建UWB单基站系统/INS紧组合的扩展卡尔曼滤波(EKF)模型;最后,利用EKF的位置预测结果计算智能车相对于UWB基站的距离和角度,并与UWB实测数据比较,根据残差大小分辨并剔除包含粗大误差的UWB距离或角度数据。实验结果表明,在狭长实验环境下,所提方法定位精度可达0.343 m,定位均方根误差比测距组网UWB的最小二乘法和泰勒级数展开法分别减小了68.62%和62.80%,比测距测角UWB直接定位算法和测距测角UWB/INS松组合算法分别减小了37.29%和16.34%,能够在狭长空间环境中实现较高精度定位。

Aiming at the problem of low positioning accuracy caused by poor positioning geometry of ultra-wideband(UWB)ranging network in narrow space environment,an intelligent vehicle positioning method based on tight combination of UWB single base station system and inertial navigation system(INS)is proposed.Firstly,the data accuracy of the UWB single base station system is improved by calibrating the measurements of distance and angle.Then,the measurement equation is established according to the UWB distance and angle measurements,and the extended Kalman filter(EKF)model of UWB single base station system/INS tight combination is constructed.Finally,the distance and angle of the intelligent vehicle relative to the UWB base station are calculated by using the position prediction results of EKF,and compared with the measured data of UWB,and the UWB distance or angle data containing gross errors are distinguished and eliminated according to the residuals.The experimental results show that the positioning accuracy of the proposed method can reach 0.343 m,and the root mean square error of positioning is reduced by 68.62%and 62.80%respectively compared with the least square method and Taylor series expansion method of ranging network UWB,and by 37.29%and 16.34%respectively compared with the direct positioning algorithm of ranging and angle measurement UWB and loose combination algorithm of ranging and angle measurement UWB/INS,which can achieve high precision positioning in the narrow space environment.