A Novel 3D Pedestrian Navigation Method for a Multiple Sensors-Based Foot-Mounted Inertial System

Area of Science:

• Robotics and Navigation
• Sensor Fusion
• Human Motion Analysis

Background:

• Foot-mounted inertial systems are crucial for 3D pedestrian navigation.
• Traditional Zero-Velocity Update (ZUPT) methods struggle with accurate stance phase detection during running.
• Barometer-based height estimation is sensitive to environmental variations.

Purpose of the Study:

• To develop a robust 3D pedestrian navigation system for foot-mounted inertial sensors.
• To improve the accuracy of Zero-Velocity Update (ZUPT) algorithms, especially for running gaits.
• To enhance height estimation accuracy by integrating inertial measurement units (IMUs) and barometers.

Main Methods:

• Integration of a Micro-Electro-Mechanical Systems (MEMS) Inertial Measurement Unit (IMU), barometer, and permanent magnet.
• Introduction of a permanent magnet-based ZUPT detector to accurately identify stance phases.
• Development of a Height Difference Information Aided Barometer (HDIB) algorithm for improved altitude tracking.
• Proposal of a ZUPT-based Adaptive Average Window Length (ZUPT-AAWL) algorithm for precise total travelled distance calculation.

Main Results:

• The permanent magnet-based ZUPT detector effectively handles running gaits.
• The HDIB algorithm provides improved height estimation, with errors ranging from 0 m to 2.35 m.
• The ZUPT-AAWL algorithm achieved a True Total Travelled Distance Error (TTDE) between 0.32% and 1.04% for both walking and running.

Conclusions:

• The proposed navigation system demonstrates significant improvements in accuracy for 3D pedestrian positioning.
• The novel ZUPT detector and HDIB algorithm effectively address limitations of traditional methods.
• This research offers a more reliable solution for pedestrian navigation using foot-mounted inertial systems across various gaits.