Constrained control and estimation. An optimisation approach. (Q1780009): Difference between revisions

In practice, the states and inputs of a control system are bounded and a realistic control system design has to take this into account. There is no systematic theoretical background to handle such important problems. This book is a study on such problems for linear finite-dimensional dicrete-time systems. The methods used are receding horizon control and moving horizon estimation. There are three parts in this book: The first deals with ``fundamentals'', and two smaller ones are devoted to case studies (rudder roll stabilization of ships, cross-directional control, control over communication networks) and ``further developments'' (a study of the Hessian associated to the cost involving an eigenvalue analysis, output feedback in the context of this book, finite-alphabet controllers and estimators). In the ``fundamentals'', after a review on basics of (constrained) optimisation (quadratic programming, Karush-Kuhn-Tucker necessary and sufficient conditions, Lagrangean duality), the chapters successively study: fixed horizon optimal control with constraints, receding horizon optimal control with constraints, constrained linear quadratic optimal control, relation of the latter with dynamic programming, a study of the effect of increasing the moving horizon on the calculated values of the controls (sometimes this effect is limited), computational issues, constrained estimation and duality between constrained estimation and control. References and an index are included at the end. The authors state on page 4: ``Most of the existing literature on control theory deals with unconstrained problems.'' This is true. Let us mention, however, that the authors neither cite the recently published book whose topics are closely related to this one [\textit{A. H. Glattfelder} and \textit{W. Schaufelberger}, Control systems with input and output constraints. London: Springer (2003; Zbl 1030.93002)] nor \textit{Z. Lin}'s book [Low gain feedback. London: Springer (1998; Zbl 0927.93001)] that is also closely related to this one. Overall, this book is worth acquiring for those involved in control system design, and it produces a valuable contribution to the area of constrained control systems.