The integrated positioning precision of ultra-wideband(UWB)and inertial measurement unit(IMU)is susceptible to the dual influence of non-line-of-sight(NLOS)error and sensor lever arm.To attain reliable localization results,a UWB/IMU positioning algorithm considering lever arm compensation in NLOS environment is proposed.The NLOS identification and suppression algorithm based on the robust local weighted regression method is designed by exploiting the local linear characteristics of UWB localization pseudo-range in the localization process,which improves the accuracy of the UWB iterative localization algorithm.The improved robust singular value decomposition(SVD)unscented Kalman filter(UKF)algorithm considering lever arm(LA-IRUKF)is used to combine the optimized observation data and the state model to improve the positioning accuracy and effective lever arm compensation.Finally,the LA-IRUKF algorithm is validated by using simulated and empirical data.The results of the simulation experiments demonstrate the effectiveness of the NLOS error identification and suppression algorithm as well as the necessity of the lever arm compensation.The experimental results of the real scene show that the LA-IRUKF algorithm can not only accurately compensates for the lever arm,but also limits the detrimental effects of the NLOS error in NLOS environment.Compared with the robust SVD-UKF(RUKF)algorithm,the robust SVD-UKF considering lever arm(LA-RUKF)algorithm,and the improved robust SVD-UKF(IRUKF)algorithm,the LA-IRUKF algorithm improves the positioning accuracy by 42.4%,41.6%and 28.5%respectively,which has the advantages of high accuracy and good resistance to aberration.
robust local weighted regressionlever arm compensationUWB/IMUunscented Kalman filter
NETL
NSTL
万方数据
