Research on the Improved Data Processing Method for Foot-Mounted Inertial Pedestrian Positioning System

Abstract

Foot-mounted Inertial Pedestrian Position System (FIPPS) plays an important role for indoor position application. It can be used in the environment without GNSS such as firefighting and military. However, the accumulating position error produced by the MIMU measurement noise, makes the system for long time position impossible. Zero Velocity Update (ZUPT) is a proposed algorithm to reduce position error for FIPPS. For ZUPT, Kalman filter is used to estimate and compensate the position error. However, because the heading misalignment angle cannot be observed by ZUPT, part of the position error caused by heading misalignment angle cannot be compensated. According to the problem above, this paper shows the improving FIPPS position algorithms we proposed in recent years, including: a) Adaptive Gradient Descent Fast-Initial Alignment Algorithm for solving the problem of the inaccuracy initial attitude resulting in the position error; b) the FR-data algorithm for solving the stable problem when the pedestrian walking fast; c) Adaptive Inertial/Magnetometer Positioning Algorithm and Improved Attitude Algorithm for improving the heading misalignment angle observability; d) Dual-foot positioning algorithm based on Adaptive Inequality Constraints Kalman filter for correcting the position error. Finally, performance of the FIPPS improve algorithm is tested using MTi-G710 MIMU.