Comparative analysis of calibration variants for inertial measurement unit based on microelectromechanical system

Abstract

Abstract This paper is about comparison of two methods of calibration – calibration method for inertial measurement unit with gyroscope and accelerometer direct information outputs and calibration method for strapdown inertial navigation system with coordinate, velocity and orientation outputs. Two mathematical methods - Least Squares and Kalman Filter methods are applied respectively. Both methods assume the same mathematical error model. This model includes zero drift, nonlinearity coefficients, nonorthogonal projections and also noises of these parameters The research compare these methods and make attempt to give approximate numerical results. The results reveal features of each case, e.g. optimal calibration time for Kalman Filter and opportunity to find unknown noise parameters in inertial measurement unit variant. These features can help a researcher to choose a suitable variant, all else being equal it is recommended to use the inertial measurement unit variant.