Thermal Calibration of Triaxial Accelerometer for Tilt Measurement

Area of Science:

• * Instrumentation and Measurement
• * Sensor Technology
• * Micro-Electro-Mechanical Systems (MEMS)

Background:

• * MEMS accelerometers for inclination measurement are limited by temperature sensitivity and require individual calibration.
• * Existing calibration methods can be time-consuming and costly, impacting practical applicability.

Purpose of the Study:

• * To develop an optimized, low-cost, and time-efficient calibration and thermal compensation method for triaxial accelerometers.
• * To enhance the stability and accuracy of MEMS accelerometers in environments with significant temperature variations.

Main Methods:

• * Optimization of calibration positions using the Levenberg-Marquardt algorithm.
• * Thermal compensation employing the least squares method based on optimized calibration data.
• * Experimental validation on marketed triaxial accelerometers.

Main Results:

• * Achieved nearly 100 times improvement in accuracy post-calibration.
• * Significantly reduced temperature drift across a -40 °C to 60 °C range.
• * Reduced x-axis drift from -13.2 mg to -0.9 mg, y-axis from 11.8 mg to -1.1 mg, and z-axis from -17.9 mg to 1.8 mg.
• * Successfully calibrated Sensor Frame Error Model (SFEM) parameters.

Conclusions:

• * The proposed method offers a low-cost and efficient strategy for improving triaxial accelerometer performance.
• * The technique effectively mitigates temperature-induced errors, enhancing practical applicability in diverse environments.
• * The optimized calibration and thermal compensation approach ensures high accuracy and stability.