ADAPTIVE INCREMENTAL HIGH-DEGREE CUBATURE KALMAN FILTER UNDER POOR OBSERVATION CONDITION

This paper is concerned the state estimation problem for nonlinear systems with uncertain process noise covariance and poor observation condition. Firstly, incremental measurement equation is reconstructed by incremental modeling technology. Then, we estimate the uncertain process noise covariance by maximum a posterior (MAP) estimator. Finally, combining with high-degree cubature Kalman filter (HCKF), an adaptive high-degree cubature incremental Kalman filter (AHCIF) is proposed under poor observation condition. The simulations show that the developed algorithm can effectively eliminate the unknown system error. Furthermore, it can also improve estimation accuracy and have a great prospect in the application.