Unscented Kalman Filter Applied to noisy synchronization of Rossler chaotic system

Abstract

Extended Kalman Filter (EKF) has been widely used as an important tool in practical applications to estimate states of nonlinear systems. There are a number of deficiencies in EKF such as biased estimation, complexity in calculation and inefficacity in not being able to compute analytical derivatives affect its application in many fields. In this paper, Unscented Kalman Filter (UKF) is employed for estimation of the state variables of the chaotic dynamical system. The chaotic synchronization is implemented by the UKF in the presence of processing noise and measurement noise. The results of the simulation on the Rossler chaotic system by UKF and its comparison with EKF show that the UKF has more accuracy and efficiency than EKF.