Low-Cost and Efficient Thermal Calibration Scheme for MEMS Triaxial Accelerometer.

Tongxu Xu,Xiang Xu,Dacheng Xu,Zelan Zou,Heming Zhao

DOI: https://doi.org/10.1109/tim.2021.3096290

IF: 5.6

2021-01-01

IEEE Transactions on Instrumentation and Measurement

Abstract:In view of the large thermal drift of microelectromechanical system (MEMS) triaxial accelerometer and the high cost of traditional calibration schemes, this article proposes a low-cost and efficient thermal drift calibration scheme combining the parameter-correction method and the proposed least squares method based on the nine-parameter model. In this scheme, first, the parameter-correction method is applied to the data collected under the heating condition to obtain the thermal drift of calibrated parameters with constant error. The least squares method is then used to obtain the parameters at a temperature T c . Then, the two sets of parameters are combined, and the constant offset error is removed. Finally, higher precision thermal drift curves of the parameters are obtained after smoothing and filtering. One thousand simulations show that the previously proposed parameter-correction method has consistent accuracy and can obtain accurate drift trends of the triaxial accelerometer parameters. The actual triaxial accelerometer thermal drift calibration experiment shows that when the temperature rises by 22 °C, the sensor data converted by the parameters obtained by the proposed scheme can effectively reduce the drift. The drifts of x- and y-axes are reduced from -19.2 and 11.6 mg to -0.7 and -0.6 mg, respectively. The Z-axis drift is reduced from -23.9 to 3.5 mg. This proves the feasibility of the method and scheme proposed in this article.