Novel Drift Reduction Methods in Foot-Mounted PDR System.

Area of Science:

• Robotics and Autonomous Systems
• Navigation and Positioning

Background:

• Traditional zero-velocity update (ZUPT)-aided extended Kalman filter (EKF) in foot-mounted inertial navigation systems (INS) for pedestrian dead reckoning (PDR) struggles with drift.
• Drift arises from failed stationary phase detection and poorly observable heading in dynamic gaits.

Purpose of the Study:

• To improve the initial heading alignment accuracy of PDR systems.
• To enhance the detection of stationary phases during dynamic pedestrian gaits.
• To reduce heading drift over time in PDR systems.

Main Methods:

• A novel geometric method calibrates initial absolute heading using a calibration line.
• A new stationary phase detection method utilizes foot motion periodicity.
• A range constraint method limits estimated foot position based on maximum stride length.

Main Results:

• The geometric method provides more accurate initial heading calibration.
• The periodicity-based detector reliably identifies zero-speed intervals, reducing missed detections.
• The range constraint effectively mitigates heading drift during extended walking.

Conclusions:

• The proposed methods significantly improve the accuracy and robustness of foot-mounted PDR systems.
• These advancements address key limitations in traditional ZUPT-aided INS/PDR, enabling more reliable pedestrian navigation.