An Improved Return Maps Method for Parameter Estimation of Chaotic Systems

Abstract

Recently, an effective method called return maps is proposed for the parameter estimation of chaotic systems. However, high time-consumption limits practical applications. In this paper, we focus on this problem, and an improved return maps method is proposed. It combines the differential evolution algorithm with the return maps method, and simplifies the calculation process of Euclidean distance. Numerical simulations are carried out on two fractional-order chaotic systems, and the other five methods are used as the comparison. Results show that the improved method can accurately estimate the parameters of chaotic systems, and it saves much time than does the classical return maps method. Furthermore, the proposed method also exhibits good anti-noise performance.