Novel Multi-IMU Tight Coupling Pedestrian Localization Exploiting Biomechanical Motion Constraints.

Area of Science:

• Biomechanical engineering
• Robotics
• Sensor fusion

Background:

• Inertial localization systems rely on inertial measurement units (IMUs) for motion tracking.
• Existing systems often struggle with accuracy, especially in estimating position and height.
• Exploiting biomechanical constraints can potentially enhance the performance of these systems.

Purpose of the Study:

• To develop a novel tight coupling inertial localization system using dual IMUs on the leg.
• To integrate biomechanical constraints of the thigh and foot into the localization algorithm.
• To evaluate the system's performance against state-of-the-art methods.

Main Methods:

• Simultaneous processing of measurements from two IMUs (upper thigh and foot).
• Incorporation of leg link motion constraints (thigh and foot) derived from motion tracking experiments.
• Tight coupling approach combining IMU data and biomechanical models.

Main Results:

• Achieved at least 50% improvement in average 2D-position error compared to state-of-the-art systems.
• Achieved at least 75% improvement in average height error compared to state-of-the-art systems.
• Demonstrated the ability to observe heading errors using only inertial measurements, without maps or trajectory repetition.

Conclusions:

• The proposed tight coupling system significantly enhances localization accuracy by integrating biomechanical constraints.
• This approach offers a more robust and accurate method for inertial localization, particularly for leg-mounted systems.
• The study provides novel insights into heading error analysis in inertial localization systems.