A Calibration Algorithm for Microelectromechanical Inertial Sensors

Abstract

Introduction. Systematic errors of microelectromechanical (MEMS) inertial sensors, such as those related to zero drift, scale factor, and nonorthogonality of sensitive axes, are the main sources of errors in strapdown inertial navigation systems (SINS). Uncompensated sensor errors accumulate over time as motion state errors, thus reducing the overall accuracy of SINS. Consequently, calibration of inertial sensors is a relevant research task. The disadvantage of existing sensor calibration methods consists in a strict requirement for the initial alignment of sensitive sensor axes relative to a reference coordinate system, which complicates the entire process of calibration. Therefore, alternative methods for MEMS sensor calibration should be developed.Aim. To develop a calibration algorithm for microelectromechanical (MEMS) sensors, which allows calibrating sensors regardless of the angular orientation of the sensor axes relative to a reference coordinate system at the initial installation, as well as to simplify the design of testing tools.Materials and methods. Publications in national and international journals on the theory of calibration of inertial sensors were reviewed. A calibration algorithm was developed based on the least squares method.Results. An algorithm for determining the calibration parameters of sensors regardless of the initial alignment of the sensor sensitive axes relative to a reference system was developed. A simple alternative design for testing MEMS sensors was proposed.Conclusion. The method of calibrating MEMS inertial sensors proposed in this work differs from conventional calibration methods by increased reliability of the results and a simplified design of testing tools. Importantly, the results of determining the calibration coefficients of a micromechanical accelerometer (MMA) do not depend on its angular position relative to a geographic coordinate system (GСS). This works contributes to improving the accuracy of SINS based on MEMS inertial sensors.