Innovative Modeling of IMU Arrays Under the Generic Multi-Sensor Integration Strategy

Area of Science:

• Navigation Systems
• Sensor Fusion
• Kinematics

Background:

• Integrating Inertial Measurement Unit (IMU) arrays into multi-sensor systems presents challenges in characterizing individual sensor errors.
• Existing methods often assume homogeneous sensor behavior, which may not reflect reality for tactical-grade MEMS IMUs.
• Accurate modeling of sensor errors is crucial for robust kinematic positioning and navigation.

Purpose of the Study:

• To propose a novel modeling method for integrating IMU arrays into multi-sensor kinematic positioning/navigation systems.
• To enable time-varying estimation of individual sensor errors within an IMU array.
• To implement rigorous fault detection and exclusion for outlying measurements from IMU sensors.

Main Methods:

• Leveraging the Generic Multisensor Integration Strategy (GMIS) for IMU array integration.
• Utilizing a Discrete Kalman filtering framework for comprehensive error analysis.
• Applying Variance Component Estimation (VCE) to characterize individual IMU gyroscope and accelerometer performance.

Main Results:

• The multi-IMU solution demonstrated an average accuracy improvement of 14-16% in position, 30% in roll/pitch, and 40% in heading compared to single IMU solutions.
• The proposed method successfully estimated time-varying sensor errors for each IMU in the array.
• Variance Component Estimation (VCE) enabled performance comparison between individual IMU sensors, revealing non-homogeneous behavior.

Conclusions:

• The developed method effectively integrates IMU arrays into kinematic positioning/navigation systems, significantly enhancing accuracy.
• IMUs within an array, even of the same model, exhibit non-homogeneous behavior that requires individual characterization.
• This research provides a foundation for future evaluations of IMU array sensor configurations and advanced sensor fusion techniques.