scientific article; zbMATH DE number 1190610

From MaRDI portal
Publication:3841035





This paper deals with a class of chaotic systems modeled by the following first-order nonlinear difference state equation \[ x_{k+1}= F(x_k), \quad k=1,2,\dots, \tag{1} \] where \(x_k\in I(\subset R)\), the nonlinear mapping \(F(\cdot):I\to I\) is a piecewise \(C^2\) function satisfying \(| dF/dx|> 1\) for all points in \(I\) at which \(dF/dx\) exists. The system is treated as a special kind of stochastic system characterized by an ergodic first-order Markov process \(x(t)\) from the viewpoint of statistics and of the external behaviors of the system. Let \(p(A| x)\) be the transition probability from the point \(x\) to a Borel subset \(A\) and assume that the limit set \(\omega(x)\) of system (1) has an invariant probability measure \(\mu(A)= \int p(A| x)d\mu(x)\). Then, an equivalent probability model is defined for describing the system (1) in the sense of an invariant measure \(\mu(x)= \lim_{t\to\infty} \pi(x,t)\), \(x\in M\), where \(\pi(x,t)\) is the distribution function of \(x(t)\). The first-order probability \(P_k(x^i)\) and the transition probability \(P(x_{k+1}| x_k)\), \(x_k\in I\) may be easily obtained by using the ergodicity of \(x(t)\). Thus, the MAP estimation of \(x_{k+1}\) based on \(x_k\) or the one-step ahead prediction of \(x(t)\) is \[ \widehat{x}_{k+1}= \widehat{F}(x_k)= \{x_{k+1}\mid \max_{x_{k+1}} P(x_{k+1}| x_k)\}. \tag{2} \] Then, a simple and practical control strategy \(u_k\) is proposed to eliminate or suppress the chaos and to obtain an expected output \(x_r\) as follows: \[ x_{_k+1}= F(x_k)+ u_k \quad u_k= f(\widehat{x}_{k+1}), \quad \widehat{x}_{k+1}= \widehat{F}(x_k), \tag{3} \] where \[ u_k= x_r- \widehat{x}_{k+1}- \sum_{j=1}^k \frac{x_r-x_j}{T_j}. \tag{4} \] The effectiveness of the proposed approach has been illustrated by simulation examples on two chaotic systems: \[ x_{k+1}= 4x_k (1-x_k), \quad x_k\in I= [0,1] \] and \[ x_{k+1}= -1.4x_k+ 0.8x_k^2+ 0.8x_k^4, \quad x_k\in I\approx [-0.52,1]. \]











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3841035)