Stabilizing multiple equilibria and cycles with noisy prediction-based control

From MaRDI portal



Abstract: Pulse stabilization of cycles with Prediction-Based Control including noise and stochastic stabilization of maps with multiple equilibrium points is analyzed for continuous but, generally, non-smooth maps. Sufficient conditions of global stabilization are obtained. Introduction of noise can relax restrictions on the control intensity. We estimate how the control can be decreased with noise and verify it numerically.


Prediction-based control (PBC) was introduced by \textit{T. Ushio} and \textit{S. Yamamoto} [Phys. Lett., A 264, No. 1, 30--35 (1999; Zbl 0941.93023)] as a technique for controlling unstable or chaotic dynamics in iterative systems. The control itself may be applied at every iteration or periodically, the latter case is referred to as pulsed PBC. In any case, the application of the control at the \(n^{\mathrm{th}}\) step is modulated by a control parameter \(\alpha\) that may be stochastic. PBC is often studied in the context of population dynamics, and its application to systems governed by a Ricker map, a logistic map, or a Maynard Smyth map, was studied in [\textit{E. Braverman} et al., Chaos 30, No. 9, 093116, 15 p. (2020; Zbl 1454.93286)]. In this article, the authors try to demonstrate how, for a scalar iterative equation governed by a map with several fixed points, PBC can be applied to simultaneously stabilise all odd-indexed fixed points. Their approach relies on the natural decomposition of the state space of the set of fixed points into intervals, where they show that the controlled system solutions can only converge to one of the equilibrium points or circulate infinitely between these intervals. They then identify how large the control parameter should be in order to suppress the possibility of infinite circulation. The map itself is assumed to obey one-sided Lipschitz-type bounds in each of these intervals. Finally, the authors introduce a stochastic variation around the mean value of the control parameter that is bounded in magnitude and stepwise independent. They are able to identify the range of noise intensities for which a PBC that is stabilising in the absence of noise, will continue to be stabilising after the introduction of noise, and they also show that the presence of noise can increase the range of possible mean control parameters for which stabilisation is possible. These results are illustrated with numerical examples that include compound forms of the Ricker map with four or more fixed points.











This page was built for publication: Stabilizing multiple equilibria and cycles with noisy prediction-based control

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2169011)