Stride and building heading aided algorithm for pedestrian navigation

In view of the low positioning accuracy achieved owing to poor observability of the cause of heading error during zero velocity updating (ZUPT), on the basis of ZUPT, we propose a heading aided algorithm based on pedestrian stride and the building direction to expand the dimensions of the observation. From the movement of a pedestrian, we obtained the stride direction and building heading. Then, we established a measurement model based on the velocity and heading errors, using a Kalman filter for the update. Furthermore, we investigated the Rauch Rung Streibel (RTS) smoothing algorithm, and smoothed the trajectory using the RTS smoother to solve the mutation problem. Finally, we conducted experiments to prove the algorithm's validity by using the MTi-G inertial measurement unit (IMU) of Xsens in the Netherlands, and verified the effectiveness of the algorithm by comparing its positioning error with those of different algorithms. The results show that the maximum positioning error does not exceed 3.4 m when the walking distance is 70 m.