Micromagnetometer calibration for accurate orientation estimation

Area of Science:

• Geophysics
• Robotics
• Sensor Technology

Background:

• Micromagnetometers and inertial sensors are crucial for attitude estimation.
• Accurate sensor calibration is essential for reliable attitude reconstruction.
• Existing methods address individual sensor errors but often neglect alignment differences.

Purpose of the Study:

• To propose a practical sensor calibration method for attitude estimation.
• To simultaneously correct for sensor errors, magnetic interference, and alignment differences.
• To improve the accuracy of attitude reconstruction in sensor nodes.

Main Methods:

• A two-step approach is presented to determine combined bias and transformation matrix separately.
• Combined bias is found by fitting an optimal ellipsoid to sensor readings.
• Transformation matrix is estimated using linear least-squares and singular value decomposition.

Main Results:

• The proposed method effectively calibrates sensor nodes by considering multiple error sources.
• Consistent calibration results were achieved across multiple trials.
• A root mean square error of less than 3° for orientation estimation was demonstrated.

Conclusions:

• The developed method offers a practical solution for comprehensive sensor calibration.
• Simultaneous correction of sensor, magnetic, and alignment errors enhances attitude estimation accuracy.
• The approach is effective for real-world applications requiring precise orientation determination.