Research on UWB Localization Method Based on EKF-LSTM in NLOS Environment
XIA Junhui
LI Jie
CHANG Honglei
ZHAO Lening
PENG Xiaofei
Abstract：Aiming at the problem of UWB signal rejection caused by NLOS environment in UWB/INS integrated navigation,this paper proposes an integrated navigation algorithm based on EKF-LSTM assisted UWB/INS.The extended Kalman filter is used to fusion the positioning information of UWB and INS.When the UWB signal is rejected,the combined positioning system enters the pure inertial navigation mode,in which the positioning accuracy gradually decreases with the accumulation of time.To solve this problem,LSTM short-duration memory neural network is used to assist localization.When UWB signal is not interrupted,the posi-tion,speed and acceleration information of INS settlement is used as the input of LSTM neural network,and the location informa-tion of fusion positioning is used as the output for training.When UWB signal is rejected,the error of INS is corrected by the trained neural network.The experimental results show that with the aid of LSTM neural network,the positioning accuracy is greatly im-proved and the output is stable.
Keywords：UWBINSLSTM short-duration memory neural networkextended Kalman filter
Publication Date：2025-10-20
Online Publishing Date：2026-01-16(First online date of this platform, not the publication date of the document)
Pages：5( 55-59 )