Calibration of inertial measurement unit with simultaneous estimation of the temperature time-derivative variations

Abstract

The paper presents an extension to the previously developed inertial measurement unit (IMU) calibration technique that uses a low-cost turntable with a horizontal axis of rotation. We consider a conventional error model for gyroscopes and accelerometers that include null biases, scale factor errors, misalignment errors and temperature variations. Temperature variations include parameter variations over temperature itself and over-temperature time-derivative. The aim of the study is to come up with conditions the experiment should satisfy for the temperature time-derivative coefficients to be well observed in calibration with changing temperature. We propose a modification to the existing algorithm that allows estimating the coefficients for temperature time-derivative variations simultaneously with the rest of conventional error parameters and applied it to partially simulated data.