Multiple objective control synthesis (Q1972255)

The first three chapters of this book are highly recommendable to those who have a basic knowledge of convex analysis and would like to have a short way to repeat (or teach) convex analysis in general spaces. Almost all theorems are given with full proofs. We find nets, compactness, Hahn-Banach, Alaoglu, duality, separation of sets, projections, Kuhn-Tucker and some remarks on sensitivity. In the second part of the book (chapters 4 to 9) using and extending the contents of the first part various special control design problems are formulated, solved, illustrated by examples and supplemented by explicitly given hints to numerical procedures and summaries at the end of each chapter. Even Pareto efficiency plays a certain role in some parts of the book, but only for convex cases, so that efficiency can be covered by scalarization. We add some remarks about the chapters 4 to 9. In chapter 4 a lot of definitions is given (causality, time invariance, stability, convolution maps, certain transforms which are analytic in the open unic disc, finite dimensional systems, controllability, observability, detectability, numerical realization, stability of closed loop maps, Youla parameter), so that one has an impression of the contents of this chapter (and the followers), which is of paradigmatic character. In chapter 5 for a certain standard feedback problem it is shown that an optimal solution exists. Such optimal solutions are characterized as finite impulse reponse sequences. Some sensitivity results are added. Also chapters 6 to 8 follow this pattern with problems having growing difficulty. Chapter 9, entitled ``Robust Performance'' (with respect to perturbations in a specifed class), contains results on robust stability and robust performance, lower and upper bounds and connections to quadratic programming. This is an interesting book on important problems of multiple objective control synthesis, which is very well readable and can be recommended to all who have to deal with this field theoretically but also practically. One feels that the authors have experience with this part of Operations Research (control) and didactic skill to present the results.