Signal Filtering and Calibration of MEMS-Based Inertial Measurement Units

The paper is devoted to improving the technology of bench calibration of micro-electro-mechanical sensors (MEMS) for inertial measurement units (IMU). A combined calibration procedure is proposed, including the processing of MEMS signals in the frequency and time domain. Filtering in the frequency domain makes it possible to eliminate the influence of noise on the accuracy of estimating MEMS errors. The estimation of systematic MEMS errors in the time domain is based on reliable mathematical models of signal drifts relative to bench standards. Therefore, the calibration technology also includes procedures for parametric identification of MEMS error models. The estimation of systematic errors is realized by processing observation of MEMS signals relative to their bench standards in three IMU position using the extended Kalman filter. The results of semi-natural experiments based on the registered MEMS data are analyzed.