Shoe-Mounted Personal Navigation System Based on MEMS

Shoe-mounted personal navigation system(PNS) based on MEMS is presented aiming at locating a person. Large error remains a problem in a standalone navigation system that uses only low cost MEMS IMU. This paper therefore proposes a approach which integrated with ZUPT and compass heading information in Kalman-based framework. The filter propagates and estimates the errors during the stance phase, which are fed back to the INS for compensating the accumulated error during walking. Finally, the performance of the proposed algorithm are verified by outdoors actual walking experiments of closed-loop path, were undertaken on campus of Northwestern Polytechnical University. The results show that the accuracy of positioning were almost always below 1% of the total traveled distance.