Regularization and Stabilization of Randomly Switching Dynamic Systems.

G. Yin,Guangliang Zhao,Fuke Wu
DOI: https://doi.org/10.1137/110851171
IF: 1.9
2012-01-01
SIAM Journal on Applied Mathematics
Abstract:This work concernsregularization (or explosion suppression) and stabilizationof randomly switching dynamicsystems by selecting suitable input functions.Starting with a randomly switching system modulated by acontinuous-time Markov chain, our aim isto find a feedback control (input function) so that thesystem becomes stable. But before addressing the stabilization issue, anotherproblem needs to be resolved first.Because of the fast growth in the continuous state variable,the solution of the system has a finite explosion time with probability one.To ensure the existence of global solutions,a feedback control (a noise perturbation) is added tomake the system regular.Then another feedback control (asecond perturbing noise) is added to ensure the stability of theresulting systems. Owing to the nonlinearity, closed-form solutions are often virtuallyimpossible to obtain. Thus, a discrete-time approximation algorithm is constructed.Under simple conditions, a suitably interpolated sequence of the discrete-time algorithmis shown to converge to the limit switching diffusion. Finally, regularization and stabilization ofthe approximating sequence are dealt with.
What problem does this paper attempt to address?