Research And Implementation Of Indoor Positioning Method Based On UWB And Inertial Technology

In order to solve the positioning problem in the indoor environment,there are positioning technologies based on Bluetooth,ultrasound and RFID,but the above several positioning technologies are not satisfactory in terms of positioning accuracy and system cost.In this context,this paper studies the positioning technology based on Ultra Wide Band(UWB)and inertial technology,and combines the two technologies.By integrating the advantages of both,they complement each other's shortcomings and achieve a more stable and reliable indoor positioning effect.The specific research content is as follows(1)The UWB-based positioning technology was studied.Aiming at the problem of random noise interference and non-line-of-sight path(NLOS)error in the arrival time(TOA)ranging in indoor environment,a linear Kalman filter is used to smooth the TOA ranging value and introduce new information in the filter.Measures to improve the anti-NLOS error performance of the filter.In terms of position estimation algorithm,this paper combines the results of fixed-point positioning experiments to analyze several positioning algorithms,and finally selects the least squares algorithm with high positioning accuracy and short calculation period.Finally,the UWB dynamic positioning experiment verifies that UWB positioning has higher accuracy in the line-of-sight environment,but it also has the disadvantage of being susceptible to environmental interference.(2)The indoor positioning technology based on pedestrian trajectory estimation(PDR)is studied.Aiming at the step detection problem,based on the law that the acceleration data will change periodically during pedestrian walking,a multi-parameter constrained step detection algorithm is designed.Aiming at the step size estimation problem,a step size estimation model based on motion state is designed by analyzing the advantages and disadvantages of the common step size and the applicable range.For the heading detection problem,the self-correcting measures of heading are adopted by studying the correspondence between the heading sample value and the walking attitude during straight walking.The experimental results show that the optimized PDR algorithm is accurate in step counting and is suitable for step estimation in different motion states.The positioning effect has the characteristics of high short-time accuracy,but there is still a problem of cumulative error inherent in PDR.(3)In order to combine the advantages of UWB and PDR and make up for each other's shortcomings,this paper designs a UWB/PDR combined positioning method.On the one hand,the PDR-assisted UWB ranging positioning method solves the problem that the UWB positioning method is susceptible to environmental interference.Because the new interest discriminant method is difficult to identify NLOS error in the sudden change state of motion,such as start and fast turn,this paper proposes a NLOS error identification method based on PDR region estimation.At the same time,in the stage,for the problem that the error between the filtering model and the actual motion model affects the convergence of the state quantity,a fading factor matrix is introduced to improve the tracking ability of the filter at the sudden change of the state variable;The problem is that the PDR recursive prediction value is used instead of the observation method to ensure the continuous output of the positioning result.On the other hand,since UWB positioning is an absolute position fix,this paper uses the positioning result of UWB to correct the cumulative error of PDR.The experimental results show that the UWB/PDR combined positioning method designed in this paper has higher positioning accuracy and stability than the UWB or PDR positioning method alone.(4)In order to cooperate with the implementation of the combined positioning algorithm based on UWB and inertial technology,this paper sets up a hardware development platform,and designs the positioning engine and Web server to realize the real-time position solution refresh and remote monitoring of pedestrians.