An Improved Maximum Stride Length Constraint Method for Pedestrian Localization

To reduce heading angle errors in pedestrian localization systems based on inertial measurement units (IMUs) and improve positioning accuracy, this paper proposes an enhanced foot-mounted inertial navigation method with an improved maximum stride length constraint. First, the principle of the pedestrian localization error constraint algorithm based on the dual-foot maximum stride length constraint is introduced. Then, considering the inherent upper limit of single-foot stride length during walking, a pedestrian localization method integrating both single-foot and dual-foot maximum stride length constraints is designed. This method assumes that the current moving foot must satisfy both constraints simultaneously, thereby determining a reasonable foot position. This position is treated as an observation and fused with the inertial navigation solution using a Kalman filter (KF) to mitigate heading angle errors and enhance localization accuracy. Indoor straight-line walking experiments demonstrate that, compared to the zero-velocity update method and the traditional maximum stride length constraint method, the proposed approach reduces pedestrian localization errors from 3.67% and 1.68% of the total travel distance to 0.59%, significantly decreasing the deviation between the estimated trajectory and the actual path.