Experimental study of the derivative-free Kalman filtering for chaos

Abstract

An experimental study related to the derivative-free Kalman filtering scheme for chaotic signals is presented in this paper. Some previously published papers had proposed some effective quasioptimum nonlinear filtering algorithms for chaotic signals. However, digital implementation of these approaches has certain limitations such as loose of stability, cumulative errors, high computational complexity, etc. In order to avoid these shortcomings, in this paper we propose a new robust and rather efficient alternative approach for the nonlinear filtering based on differential flatness property of some chaotic non-linear dynamic systems. Moreover, experimental results presented in this paper allow comparing non-linear filtering algorithms for chaos with derivative-free technique under different scenarios.