Zero-velocity interval detection algorithm with multi-condition constraint for pedestrian navigation

To deal with the large drift and low performance of MEMS inertial sensors, a zero-velocity interval detection algorithm with multi-condition constraint is designed based on the analysis of pedestrian gait characteristics by taking into account of pedestrian navigation system application requirements. This algorithm comprehensively utilizes the foot sensor's output parameters, such as amplitude, peak value and variance, and extracts the gait of the zero-speed points by setting the threshold. The multi-condition constraint effectively reduces the possibility of miscalculation. The detection tests of zero-velocity intervals with different walking speeds show that the zero velocity interval numbers detected by the proposed method are the same as those in the actual walking. The accurate extraction of zero-velocity intervals lays the basis for the speed error correction algorithm in future pedestrian navigation research.