On discrete stochastic bilinear systems stability (Q1088968): Difference between revisions

The author investigates the mean square stability of the discrete-time bilinear system \[ x(i+1)=(A_ 0+\sum^{p}_{k=1}A_ k\omega_ k(i))x(i)+Bu(i),\quad i=0,1,2,... \] in \({\mathbb{R}}^ n\) with constant matrices \(A_ 0,A_ 1,...,A_ p\) and B of appropriate order, which is disturbed by internal noise \(w(i)=(\omega_ 1(i),...,\omega_ p(i))\) and random input u(i), both being second order wide sense stationary, independently distributed random sequences. There is no ergodicity assumption. For w(i), u(i) being uncorrelated, sufficient conditions are given in terms of the second moment of \(W(i)=\sum A_ kw_ k(i)\) and the mean of \(A_ 0+W(i)\), in order that Ex(i) and \(E(x(i+k)x(i)^*)\) converge as i tends to \(\infty\).