Comparison on H∞ Filter and Kalman Filter for Initial Alignment of SINS on Static Base

摘要

This paper discusses two kinds of initial alignment for Strap-down Inertial Navigation System (SINS), one is based on Kalman filter, and the other is based on H∞ filter. Through modeling, simulating and comparing, it can be concluded that using the former, with given system model and the noise characteristics, the leveling misalignment angles converge relatively rapid, but the azimuth misalignment angle converges relatively slow. If there is too large disturbance on system model, Kalman filter would be prolonged, even be emanated. However, the standard Kalman filter performances can be improved by using H∞ filter. The latter performs better than the former in azimuth alignment and equivalent to the former in level alignment. And H∞ filter can also greatly enhances the robustness of the system. So H∞ filter is an available method for the initial alignment.