Stabilizing nonlinear dynamical systems by an adaptive adjustment mechanism

An adaptive adjustment mechanism is applied to stabilize multidimensional dynamical systems. Without utilizing any prior knowledge of the system itself, nor extra external control signals, the mechanism can ensure a large class of chaotic systems to converge to their "generic" stable periodic orbits.     

Invariant integral and the transition to steady states in separable dynamical systems

We show that the transition between fixed points in a separable dynamical system is fully described by an invariant integral. We discuss in detail the case of a system with two temporal variables with bilinear coupling, where the new stable state is attained asymptotically through spiraling into the fixed point. Through the invariance, it is possible to establish conditions for the control parameter that permit a (targeted) transition in finite time and without relaxation oscillations.     

Stabilizing unstable steady states using extended time-delay autosynchronization

We describe a method for stabilizing unstable steady states in nonlinear dynamical systems using a form of extended time-delay autosynchronization. Specifically, stabilization is achieved by applying a feedback signal generated by high-pass-filtering in real time the dynamical state of the system to an accessible system parameter or variables. Our technique is easy to implement, does not require knowledge of the unstable steady state coordinates in phase space, automatically tracks changes in the system parameters, and is more robust to broadband noise than previous schemes. We demonstrate the controller's efficacy by stabilizing unstable steady states in an electronic circuit exhibiting low-dimensional temporal chaos. The simplicity and robustness of the scheme suggests that it is ideally suited for stabilizing unstable steady states in ultra-high-speed systems. (c) 1998 American Institute of Physics.     

Robust adaptive fuzzy neural tracking control for a class of unknown chaotic systems

In this paper, an adaptive fuzzy neural controller (AFNC) for a class of unknown chaotic systems is proposed. The proposed AFNC is comprised of a fuzzy neural controller and a robust controller. The fuzzy neural controller including a fuzzy neural network identifier (FNNI) is the principal controller. The FNNI is used for online estimation of the controlled system dynamics by tuning the parameters of fuzzy neural network (FNN). The Gaussian function, a specific example of radial basis function, is adopted here as a membership function. So, the tuning parameters include the weighting factors in the consequent part and the means and variances of the Gaussian membership functions in the antecedent part of fuzzy implications. To tune the parameters online, the back-propagation (BP) algorithm is developed. The robust controller is used to guarantee the stability and to control the performance of the closed-loop adaptive system, which is achieved always. Finally, simulation results show that the AFNC can achieve favourable tracking performances.     

Iterated function systems and dynamical systems

We study the relationship between measures invariant for a piecewise expanding transformation tau of a compact metric space endowed with a underlying measure and measures invariant for an iterated function system T(tau), generated by inverse branches of tau. The main result says that the tau-invariant absolutely continuous measure &mgr; is also T(tau) invariant if and only if tau is absolutely continuously conjugated with a piecewise linear transformation. Measures of maximal entropy and general equilibrium states are also discussed. (c) 1995 American Institute of Physics.     

Punctures and dynamical systems

Letters in Mathematical Physics - With the aim of better understanding the class of 4D theories generated by compactifications of 6D superconformal field theories (SCFTs), we study the structure of...     

Non-Hamiltonian dynamical systems

The class of dynamical systems is considered, which are described by several mutually noncommuting Hamiltonian currents, in particular, relativistic bi-Hamiltonian systems, the evolution of which is described by a pair of 4-momenta p and p The examination is conducted in classical and quantum realizations. The evolution equations are derived of relativistic bi-Hamiltonian systems in the Heisenberg and Schrödinger pictures. It is shown that the quantum theory of relativistic bi-Hamiltonian systems is not compatible with the unitary condition and is nonunitary. A physical interpretation is given of nonunitary quantum theory.Translated from Izvestiya Vysshikh Uchebnykh Zavedenii, Fizika, No. 10, pp. 5–12, October, 1990.     

Adaptively locating unknown steady states: Formalism and basin of attraction

The adaptive technique, which includes both dynamical estimators and coupling gains, has been recently verified to be practical for locating the unknown steady states numerically. This Letter, in the light of the center manifold theory for dynamical systems and the matrix spectrum principle, establishes an analytical formalism of this adaptive technique and reveals a connection between this technique and the original adaptive controller which includes only the dynamical estimator. More interestingly, in study of the well-known Lorenz system, the selections of the estimator parameters and initial values are found to be crucial to the successful application of the adaptive technique. Some Milnor-like basins of attraction with fractal structures are found quantitatively. All the results obtained in the Letter can be further extended to more general dynamical systems of higher dimensions.     

Stabilizing near-nonhyperbolic chaotic systems with applications

Based on the invariance principle of differential equations a simple, systematic, and rigorous feedback scheme with the variable feedback strength is proposed to stabilize nonlinearly finite-dimensional chaotic systems without any prior analytical knowledge of the systems. Especially the method may be used to control near-nonhyperbolic chaotic systems, which, although arising naturally from models in astrophysics to those for neurobiology, all Ott-Grebogi-York type methods will fail to stabilize. The technique is successfully used for the famous Hindmarsh-Rose neuron model, the FitzHugh-Rinzel neuron model, and the R?ssler hyperchaos system, respectively.