Abstract
As UWB high-precision positioning in NLOS environment has become one of the hot topics in the research of indoor positioning, this paper firstly presents a method for the smoothing of original range data based on the Kalman filter by the analysis of the range error caused by UWB signals in LOS and NLOS environment. Then, it studies a UWB and foot-mounted IMU fusion positioning method with the integration of particle filter with extended Kalman filter. This method adopts EKF algorithm in the kinematic equation of particle filters algorithm to calculate the position of each particle, which is like the way of running N (number of particles) extended Kalman filters, and overcomes the disadvantages of the inconformity between kinematic equation and observation equation as well as the problem of sample degeneration under the nonlinear condition of the standard particle filters algorithm. The comparison with the foot-mounted IMU positioning algorithm, the optimization-based UWB positioning algorithm, the particle filter-based UWB positioning algorithm, and the particle filter-based IMU/UWB fusion positioning algorithm shows that our algorithm works very well in LOS and NLOS environment. Especially in an NLOS environment, our algorithm can better use the foot-mounted IMU positioning trajectory maintained by every particle to weaken the influence of range error caused by signal blockage. It outperforms the other four algorithms described as above in terms of the average and maximum positioning error.
Highlights
With the wide application of indoor positioning technologies in some areas such as supermarket shopping, fire emergency navigation, and hospital patient tracking, indoor positioning can be implemented through the following two approaches
As it might result in the problems of signal multipath effect or intensity attenuation, high-precision positioning can hardly be achieved in NLOS environment through the UWB positioning approach
E other approach is based on the inertial measurement unit (IMU), such as accelerometer, gyroscope, magnetometer, and so on [6], which can be used for positioning according to the integral or the PDR method
Summary
Received 1 August 2017; Revised 2 November 2017; Accepted 12 April 2018; Published 3 July 2018. En, it studies a UWB and foot-mounted IMU fusion positioning method with the integration of particle filter with extended Kalman filter. E comparison with the foot-mounted IMU positioning algorithm, the optimization-based UWB positioning algorithm, the particle filter-based UWB positioning algorithm, and the particle filter-based IMU/UWB fusion positioning algorithm shows that our algorithm works very well in LOS and NLOS environment. In an NLOS environment, our algorithm can better use the foot-mounted IMU positioning trajectory maintained by every particle to weaken the influence of range error caused by signal blockage. It outperforms the other four algorithms described as above in terms of the average and maximum positioning error
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
