Assessment of zero-velocity detectors for pedestrian navigation system using MIMU

Abstract

a design of pedestrian navigation system based on EKF+ZUPT+SINS is proposed in this study. Three different zero-velocity detectors are assessed by the position accuracy which is calculated through PNS with different choices of detectors and parameters. In order to show the algorithm performance and the influence of zero-velocity detectors on the whole system, the generated pedestrian trajectories fusing the information of MIMU are illustrated. The final results show that a position accuracy of 0.39% in 10 minutes can be achieved and the detectors using the information of gyroscope signals such as SHOE Detector prove to be more reliable to detect the time epochs when MIMU is stationary. Through the experiments conducted in outdoor and indoor environments, the proposed algorithm proves to be efficient.