Research And Implementation Of Ultra-Wideband Indoor Positioning Algorithm

Measurement noise and non-line-of-sight errors are inevitably introduced in the ultrawideband indoor positioning system due to the complex and varied indoor environment.The large NLOS error often leads to a sharp decline in the positioning performance of the UWB indoor positioning system,especially in the three-dimensional positioning system,the traditional positioning algorithm can not meet the actual positioning requirements.Compared with other positioning methods,TDOA,TOA and other time-based positioning methods have higher positioning accuracy.In this paper,the classical TDOA localization algorithm is simulated in the line-of-sight environment and NLOS environment.The results show that Chan and Taylor perform well in the line-of-sight environment,but the performance is poor in non-line-of-sight environment.After the RBF neural network algorithm corrects the positioning error,the positioning accuracy of the classical positioning algorithm is not far from the ideal environment.The Chan algorithm is always the best positioning algorithm.In the NLOS environment,the modified Chan algorithm based on TDOA needs to locate the strict synchronization of the clock between the base stations,and the time complexity of the algorithm is too high.This paper designs an improved UWB positioning algorithm based on TOF,The algorithm uses the least squares fitting and Kalman filtering algorithm to correct the TOF measurement with NLOS positioning error.Firstly,the classical Chan algorithm is used to obtain the initial solution,Reusing the advantage of genetic algorithm in nonlinear optimization,the difference between the true value of TOF and the measured value is taken as the fitness of the genetic iteration,and the positioning accuracy is used as the constraint condition to correct the initial solution of the positioning with low precision.The algorithm sets the search solution of the genetic algorithm to the uncertainty of the Chan algorithm,which effectively reduces the search space of the genetic algorithm and reduces the overall complexity of the algorithm and avoid synchronization problems between base stations.In the past,the research of localization algorithms is mostly based on the simulation platform.It is difficult to guarantee the practicability of the algorithm.In view of this,based on the BPM modulation method and microprocessor control technology,this paper completes the development of the experimental platform of UWB indoor positioning system,and carries out the 3-dimensional positioning verification on the platform by the classical TDOA positioning algorithm and the algorithm proposed in this paper.The results show that the performance of the proposed algorithm is higher than that of the traditional positioning algorithm in non-line-of-sight environment,and the average positioning accuracy can reach less than 5cm.