A Pedestrian Navigation Method Based on Dual MIMU for Complex Gait Patterns

A pedestrian navigation method based on dual MIMU for complex gait patterns is proposed, to address the problem that the heading of pedestrian navigation system based on foot mounted MIMU drifts over time in satellite denied environments. A constraint model for dual foot, as well as a gait recognition method based on the characteristics of inertial measurement data are designed. The constraint threshold of the distance between two feet is adaptively adjusted according to the gait recognition result. The pedestrian position correction is completed by combining zero velocity update (ZUPT) and inequality constrained Kalman filtering algorithm. The experiments show that in a mixed motion pattern that includes six gaits, the positioning error of the pedestrian navigation system based on dual MIMU using adaptive threshold constraint is 0.5
^%
D, which is reduced by 11.03% and 2.53% compared to unconstrained and fixed threshold constraint methods, respectively.