An Extended Adaptive Kalman Filtering in Tight Coupled GPS/INS Integration

In tight coupled GPS/INS integration by using Kalman filtering, the main error includes not only the dynamic model error (INS error), but also the errors caused by the poor geometry of GPS satellites or short of GPS satellites. In this paper, an extended adaptive Kalman filtering algorithm is presented based on the adaptive filter. The new algorithm can not only resist the influence of the dynamic model errors but also control the influence of the errors caused by the poor geometry of GPS satellites by adjusting the coefficient matrix of the predicted states. An actual computation example shows that the new algorithm can degrade the influence of the two kinds of errors and improve the precision of navigation.