Pedestrian Dead Reckoning with Low-Cost Foot-Mounted IMU Sensor

Area of Science:

• Robotics
• Sensor Fusion
• Inertial Navigation

Background:

• Pedestrian Dead Reckoning (PDR) systems often suffer from accumulated errors and magnetic noise.
• Noise in acceleration and gyro data from IMU sensors degrades PDR accuracy.
• Existing PDR methods struggle with precise gait-phase estimation and error correction.

Purpose of the Study:

• To propose novel methods for enhancing PDR accuracy using a single foot-mounted IMU sensor.
• To address the challenges of magnetic noise and accumulated errors in PDR.
• To improve the reliability of distance and direction estimation in PDR.

Main Methods:

• Developed a gait-phase-estimation method utilizing pitch angle for Zero Velocity Update (ZVU).
• Implemented an error-avoidance technique by updating roll and pitch angles using acceleration data.
• Conducted experiments to evaluate gait-phase and distance estimation errors.

Main Results:

• The proposed methods demonstrated reduced accumulated errors in PDR.
• Distance estimation error was approximately 7.40% for straight walking.
• Distance estimation error reached about 12.27% when travel direction shifted.

Conclusions:

• The novel PDR methods effectively mitigate noise and accumulated errors.
• The gait-phase estimation and acceleration-based angle updates improve PDR performance.
• Accurate pedestrian navigation is achievable with a single foot-mounted IMU sensor.