作者: J. S. Thorp , B. R. Barmish
DOI: 10.1007/BF00934932
关键词: Stability (probability) 、 Lyapunov function 、 Dynamical systems theory 、 Linear system 、 Nonlinear control 、 Control theory 、 Applied mathematics 、 Matrix (mathematics) 、 Mathematics 、 Positive-definite matrix 、 Contrast (statistics)
摘要: This paper addresses the problem of stabilizing an uncertain linear system. The uncertaintyq(·) which enters dynamics is nonstistical in nature. That is, noa priori statistics forq(·) are assumed; only boundsQ on admissible variations ofq(·) taken as given. results given here applied to so-called matched systems differ from previous two ways. Firstly, control this linear; for same class problems, many existing would require a nonlinear control. Furthermore, those do fact yield controls valid when certain matrix Ω(q) (formed using data) negative definite allq ∈Q. In contrast, theory requires compactness bounding setQ. Secondly, we show that matching conditions (used earlier work) can be generalized so encompass larger dynamical systems.