NONLINEAR FILTERING ALGORITHM FOR INSINITIALALIGNMENT

The initialalignm enterror equation of an INS(InertialNavigation System ) w ith large initialazim uth error has been derived and nonlinearcharacteristicsare included. When azim uth error is fairly sm all, the nonlinearequation can be reduced to a linearone. Extended Kalm an filter, iterat- ed filterand second order filter form ulas are derived for the nonlinear state equation w ith linearm ea- surem entequation. Sim ulations results show thatthe accuracy ofazim uth error estim ation using ex- tended Kalm an filter is better than that ofusing standard Kalm an filter w hile the iterated filter and second order filtercan give even better estim ation accuracy.