Research On Indoor Fusion Positioning Technology Based On UWB/INS

Ultra Wide Band(UWB)technology stands out among many indoor positioning technologies due to its strong penetration,low transmission loss,strong anti-interference capability and wide coverage range.However,UWB signals are seriously affected by the NonLine-of-Sight(NLOS)environment,so it is important to suppress NLOS errors to improve positioning accuracy.Therefore,this paper investigates the cumulative error suppression of Inertial Navigation System(INS),the system NLOS error identification and correction of UWB,and the UWB/INS fusion positioning algorithm.First,a correction framework based on the generalized likelihood ratio detection algorithm is proposed to address the cumulative error caused by accelerometer and gyroscope drift in INS systems.The stationary and kinematic states of the pedestrian are first detected by the generalised likelihood ratio detection algorithm,then the velocity is corrected for the stationary state by the zero velocity update algorithm,the heading angle is corrected by the zero integrated heading rate algorithm,and finally the corrected data are fused by the extended Kalman filter.It is then verified that the proposed INS correction framework can effectively suppress the accumulated errors through comparison experiments.Secondly,to address the problem that the traditional NLOS signal recognition algorithm does not take into account the correlation between signal waveform features,this paper proposes a NLOS recognition algorithm based on fuzzy comprehensive evaluation,which can effectively improve the NLOS signal recognition accuracy by taking into account the correlation between signal waveform features.This algorithm can effectively improve the NLOS signal recognition accuracy by considering the correlation between signal waveform features.After that,the NLOS signal recognition is corrected by the NLOS correction algorithm based on semi-definite planning.The experimental results show that the proposed algorithm outperforms the comparison algorithm in terms of positioning performance when there are fewer base stations and redundant base stations.The NLOS suppression effect is better,and it has higher positioning accuracy and stability.Finally,to address the problem that the standard cubature Kalman filtering algorithm has a large error and poor robustness in the worst case estimation,this paper proposes the H∞cubature Kalman filtering algorithm,which is based on the cubature Kalman filtering algorithm by adding H∞ estimation as the cost function to minimize the impact of the worst-case perturbation on the estimation error.Secondly,to address the problem that the standard cubature Kalman filtering algorithm cannot handle unknown and time-varying noise,resulting in reduced filtering accuracy or even divergence,this paper proposes the adaptive cubature Kalman filtering algorithm,which dynamically processes the noise information by adding adaptive filtering,effectively reducing the impact of noise on the evaluation results.Finally,a framework of tightly coupled adaptive H∞ cubature Kalman filtering algorithm is designed.Afterwards,the experimental comparison shows that the tightly coupled adaptive H∞ cubature Kalman filtering algorithm reduced the average localisation error by 2.21cm(19.25%)compared to the tightly coupled cubature Kalman filtering algorithm,and the experimental results show that the tightly coupled adaptive H∞ cubature Kalman filtering algorithm has higher localisation accuracy and better localisation stability.