Nonlinear filtering for a kind of divisible systems

For the nonlinear system whose state equation is composed of linear part and non-linear part, a novel nonlinear filter design method combined with Kalman Filter and Unscented Kalman filtering (UKF) is proposed in this paper. The algorithm divides the system into the linear sub-system and the non-linear sub-system. The optimal Kalman filter is used for the linear sub-system, on this basis, an improved UKF is utilized for the non-linear sub-system, in the sequential fusion frame. Compared with the traditional unscented Kalman filtering method, the optimality of the proposed methodfor the linear sub-system can be ensured. Therefore, the accuracy of the proposed method is superior to the traditional UKF. And the simulation results show that the improved algorithm estimation value is closer to the true value, and the filtering performance of the nonlinear system is improved.