Master-slave synchronization via dynamic control (Q2205395)

The classic master-slave synchronisation scheme is \[ \dot{x}_m=F(x_m),\ \dot{y}_m=F(y_m)+Bu_s \] where \(u_s=kC(x_m-x_s)\). Here \(x_m,x_s\in\mathbb{R}^n\), \(B\in\mathbb{R}^n\) is a constant column vector, \(C\in\mathbb{R}^{1\times n}\) is a constant row vector and \(k\in\mathbb{R}\) is the coupling strength. \(F\) describes the uncoupled dynamics. For some systems there is no value of \(k\) such that the two systems synchronise (i.e. \(\lim_{t\to\infty} x_m(t)-x_s(t)=0\)). The authors propose a modified scheme: \[ \dot{x}_m=F(x_m),\ \dot{y}_m=F(y_m)-Bh \] where \(\dot{h}=-\alpha h-kC(x_m-x_s)\) where \(\alpha>0\). For both systems the local stability of the synchronous solution is analysed. Several examples are shown for which the modified scheme synchronises while the classic one does not. The paper concludes with results from an electronic implementation of two Rössler systems.