A new continuous self-calibration scheme for a gimbaled inertial measurement unit

A typical calibration scheme for a gimbaled inertial measurement unit (IMU) involves an estimation of error parameters of an IMU mounted on an inertial platform and the platform's misalignment angles. However, traditional calibration methods for the gimbaled IMU have some serious defects. The excitation for a gyro's scale factors and misalignment angles is only the Earth rate in multi-position calibration methods and dynamic errors (unneeded motion of gyro floaters) involved in a continuous calibration process. This paper presents a new continuous self-calibration scheme for the gimbaled IMU. By processing the multi-position and continuous rotation steps alternately, the dynamic errors are suppressed and the excitation is augmented. This is more effective than traditional methods. Additionally, the platform rotation trajectory is designed to provide adequate observability for all parameters through a new methodology. The Lie derivative is used to compute the observability, and the genetic algorithm is utilized to obtain the inertial platform's optimal rotation trajectory based on the measurement of observability for all parameters. Simulation results show that the error coefficients can be effectively calibrated within an hour by the proposed scheme, and it is of high significance for fast launching of missiles and rockets