MEMS Accelerometer Noises Analysis Based on Triple Estimation Fractional Order Algorithm

Area of Science:

• Signal Processing
• Control Systems Engineering
• Sensor Technology

Background:

• Micro-Electro-Mechanical Systems (MEMS) accelerometers are vital for inertial navigation.
• Understanding noise characteristics in MEMS sensors is critical for accurate data.
• Fractional order dynamics offer a more nuanced model for complex noise phenomena.

Purpose of the Study:

• To identify parameters of fractional order noises.
• To apply these methods to noise data from MEMS accelerometers.
• To evaluate the effectiveness of the Triple Estimation algorithm for fractional noise analysis.

Main Methods:

• Utilized the Triple Estimation algorithm for simultaneous state, fractional order, and parameter estimation.
• Performed numerical analyses on fractional constant and variable order systems with Gaussian noise.
• Analyzed experimental data from a SparkFun MPU9250 Inertial Measurement Unit (IMU) accelerometer.

Main Results:

• Confirmed the capability of the Triple Estimation algorithm in estimating fractional noise parameters.
• Demonstrated the existence of fractional noise in MEMS accelerometer data across x, y, and z axes.
• Numerical simulations validated the algorithm's performance on fractional systems.

Conclusions:

• Fractional noise is present in MEMS accelerometer measurements.
• The Triple Estimation algorithm is effective for identifying fractional noise parameters.
• This characterization is essential for developing improved filtering algorithms for inertial navigation.