Error calibration of inertial measurement unit based on strapdown inertial navigation system

Abstract

Traditional error calibration methods for the Inertial Measurement Unit (IMU) of strapdown inertial navigation systems require high-precision turntables, north references, attitude leveling, and other challenges. Additionally, long-term temperature variations of sensors and circuit boards can lead to significant differences between the calibrated and actual IMU error values. This paper proposes a system-level calibration algorithm using velocity as the observation variable and employing least squares fitting. Through multi-position measurements on a three-axis turntable, this algorithm quickly calibrates 24 error parameters, including zero bias, installation angle, and scale factor, of three gyroscopes and three accelerometers, with a calibration time of approximately 2 hours. The experimental results demonstrate that the calibration process of IMU system is streamlined, the operation is simple, and it effectively improves the error time and accuracy of the strapdown inertial navigation system.