Correlation-Averaging Methods and Kalman Filter Based Parameter Identification for a Rotational Inertial Navigation System

Abstract

The attitude accuracy of the existing rotational inertial navigation system (RINS) is affected by oscillatory attitude errors caused by the installation errors of rotation axes or inertial sensors. Additional equipment is required to estimate installation errors under dynamic conditions. Methods that use the output of a single RINS to estimate installation errors under dynamic conditions are currently lacking. To address this challenge, this study proposes an installation error estimation method that combines a correlation method, an averaging method, and the Kalman filter. The proposed method adopts a correlation method to increase the signal-to-noise ratio, an averaging method to block certain sine signals, and the Kalman filter to identify installation errors in real time. Simulation, turntable, and sea tests were conducted to verify the proposed algorithm. Results show that the estimation accuracy of installation errors is at 10 arcsec levels, which indicates that said errors are estimated accurately using the RINS output initially obtained under dynamic conditions.