Cyclic feedback systems

From MaRDI portal





The author studies systems of differential equations of the type \[ dx_1/dt= f_1(x_1, x_n),\quad dx_2/dt= f_2(x_2, x_1),\dots, dx_n/dt= f_n(x_n, x_{n-1}),\tag{1} \] where constants \(\delta_i\in \{\pm 1\}\), \(i= 1,\dots, n\), exist such that \[ \delta_i f_i(0, v)v> 0 \text{ if }v\neq 0, \quad \delta_i\partial f_i(a,b)/\partial b(0,0)> 0 \text{ for }i= 1,\dots, n. \] Such systems are called cyclic feedback systems (CFSs). It is assumed that (1) has a global attractor. The basic result of the study is that, under these assumptions, the dynamics of (1) on the global attractor can be described even if the dimension of the phase space is arbitrary. The proof of this result is based on the fact that the global attractor of (1) admits a Morse decomposition, which implies restrictions on the possible dynamics and indicates the overall gradient-like structure of the flow on the attractor. Although the dynamics of monotone cyclic feedback systems (here, the functions \(f_i\) are monotone in the second variable) are essentially 2-dimensional, the author constructs a CFS with an invariant set in the lowest Morse set on which the flow is semi-conjugate to a suspension of a full shift on two symbols. That means that the dynamics inside individual Morse sets may be very complicated. Another tool used for studying CFSs is based upon discrete Lyapunov functions that is closely related to the existence of a Morse decomposition. The main results by the author show that a large class of CFSs exhibits a persistent periodic behavior. The paper is well written, the author exmphasizes geometric ideas.




Cited in
(31)








This page was built for publication: Cyclic feedback systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4214099)