Research On Indoor Positioning Algorithm Based On UWB In NLOS Environment

With the rapid development of positioning technology,ultra wideband(UWB)based positioning technology has great research significance in the field of indoor positioning.There are a lot of interference in the actual indoor positioning environment,and UWB positioning system often works in non-line of sight(NLOS)environment.Therefore,this paper studies the indoor tracking algorithm and positioning algorithm based on UWB in NLOS environment.In order to improve the tracking accuracy of UWB indoor tracking algorithm for moving label nodes in NLOS environment and reduce the positioning error,a UWB indoor tracking algorithm based on Kalman filter is proposed.The algorithm first estimates the coordinates of moving label nodes using traditional Kalman filter algorithm,and then obtains the time difference of arrival(TDOA)value with errors based on the estimation results.The NLOS error is then discriminated based on the TDOA difference,and the measured values affected by the NLOS error are reconstructed according to the discriminant results.At the same time,the Kalman filter covariance matrix is updated using the historical values only affected by the LOS error.Finally,the new Kalman filter iteration process after updating the covariance matrix is used to process the data after the line-of-sight reconstruction,and the final tracking and positioning results are obtained.The experimental results show that the algorithm can achieve decimeter-level positioning accuracy in LOS environment.Under the influence of NLOS,the root mean square error(RMSE)value of the algorithm is 0.3261 m,which can improve the positioning accuracy and reduce the positioning error more effectively than the current related algorithms.To solve the problem of low accuracy and large positioning error of UWB indoor positioning algorithm in NLOS environment,a weighted algorithm for UWB indoor positioning based on TDOA is proposed to reduce NLOS error.The algorithm first estimates the coordinates of label nodes using Chan algorithm and calculates the TDOA values between anchor nodes.Then,based on the TDOA difference,the channel environment between anchor node and label node is discriminated by NLOS environment,and the TDOA value with NLOS error is reduced by NLOS error according to the result of environment discrimination.Then,the TDOA values that reduce the errors are brought into the Taylor series expansion method for iterative calculation.Finally,a weighting factor related to the distance difference is constructed to process the iteration results and obtain the final exact coordinates of the label nodes.The simulation results show that the algorithm can effectively reduce the positioning error under the influence of NLOS.The RMSE value is 0.9372 m,and the positioning performance is better than the comparison algorithm.The paper has 27 pictures,5 tables,and 58 references.