Fragility and robustness puzzles

Several important open problem areas, or puzzles, in robust, control and modelling are highlighted. The first puzzle deals with complications caused by the fragile nature of poles and zeros in standard parameterizations of systems due to unavoidable small parameter inaccuracies. This includes also a critique of the debate paper of Keel and Bhattacharyya (1997). The second and the third puzzle are concerned with difficulties in generalizing H/sub /spl infin// control theory to realistic persistent signal setups. It seems difficult to generalize the theory to such setups and simultaneously keep the terminology robust H/sub /spl infin// control intact. These puzzles have implications for control-oriented model validation.