Optimal control of constrained piecewise affine systems. (Q2644065)

The monograph considers the class of constrained piecewise affine systems, and their optimal control and stability. The monograph is a revised and extended version of the authors PhD thesis written at the Automatic Control Laboratory of ETH Zurich in Switzerland. The monograph is intended for engineers and researchers in control, and due to its technical and abstract nature is not intended for lecture notes for students as such. One way to introduce a Piecewise Affine Systems (PWA) is to locally linearize a general nonlinear system. Thus, they can be considered a tool for approximating general nonlinear systems. Also, PWAs are under some mild assumptions equivalent to e.g. mixed logical dynamical systems, linear complementary systems, and max-min-plus-scaling systems. Although PWAs form a subclass of general nonlinear systems, most of the classical control theory developed for nonlinear controls does not apply for PWAs, since they lack the traditionally assumed continuity and smoothness properties. The monograph revolves around the efficient and systematic computation, analysis, and post-processing of closed-form, stabilizing, optimal, exact state-feedback controllers for PWAs. In Part I the author recalls some background of optimal control and related mathematical definitions, and finally defines PWAs. Optimal control is studied in Part II. The results include a solution to the constrained finite time optimal control for PWAs by using an algorithm that combines dynamic and linear programming and a solution to the constrained infinite time optimal control for PWAs by extending the ideas of the finite time case. Part III deals with stability of PWAs, and with efficient evaluation of piecewise control laws.