State estimation of nonlinear fractional order system with Lévy noises by using EKF

In this paper, the state estimation problem is discussed for discrete nonlinear fractional order system with non-Gaussian Lévy noises. By disposing the non-Gaussian Lévy noises, the system state vector and measurement vector can be approximated directly, which could enable to deduce the new system noise covariance matrix and measurement noise covariance matrix, respectively. Based on the proposed approximating method, a novel fractional order extended Kalman filter (EKF) designing strategy is developed for discrete nonlinear fractional order system with non-Gaussian Lévy noises, which is more general and effective than some existing results. Finally, some simulation results are provided to verify and illustrate the effectiveness of the obtain results.