A novel method for localization and tracking in NLOS environments of coal mines

Abstract The performance of target localization and tracking in coal mines is seriously affected by non-line-of-sight (NLOS) propagation. This paper proposes a new method to solve this problem. First, we introduce the NLOS error vector and the correction factor vector, and determine whether there is NLOS propagation through the variance of the measured value. Then the improved Chan method is proposed to decrease the NLOS error of localization results and enhance the localization accuracy. To track moving objects, we proposed ChopThin resampling-based particle filter(CTPF) algorithm to decrease the tracking error whose observed value are the corrected distance difference during localization. It ensures the effective number of effective particles and their diversity by using the ChopThin resampling to mitigate the adverse effects of the particle degeneracy. Besides, a threshold is set to decide whether to resample to reduce the amount of computation. Simulation results show that the improved Chan localization method proposed in this paper can suppress most of the NLOS errors and improve positioning accuracy, and the proposed CTPF has smaller tracking error than the traditional resampling filter algorithm.