Interacting Multiple Model Algorithm for SINS Self-Alignment on Shipboard Aircraft

As Kalman filtering approach is adopted in initial self-alignment for SINS on shipboard aircraft, variances of the measurement noise usually can not be accurately obtained because of disturbances from wind waves and aircraft carrier flexure, which will probably lead to filtering divergence. To solve the uncertain problem, a so-called multiple model alignment approach is presented taking advantage of the Interacting Multiple Model (IMM) algorithm composed of a series of traditional Kalman filter models. Simulations are conducted compared with the single Kalman filter algorithm, and the results show that the IMM algorithm can not only overcome the uncertain problem but also shorten the convergence time, thus has an important reference value for practical applications.