For mixture linear and nonlinear model in integrated navigation system, a new algorithm of mixture Gaussian
particle filtering(MGPF) is proposed. The stage of GPF state updating can be improved with the thought of Kalman filter
(KF). The updating stage is to update Gaussian distribution parameters of the particle rather than update all particles one by
one. Compared with the traditional GPF, the novel algorithm can improve filtering precision and reduce filtering time. The
MGPF algorithm is applied to SINS/GPS integrated navigation model. The simulation experiment on the established model
shows the effectiveness of the algorithm.