Control of a linear sampled-data system using round robin output feedback

A linear control problem is defined; the main feature of the problem is a constraint that the control be piecewise constant, as in a sampled-data system, and that updating of the control values depends on prior sampled values of the output variables obtained via a round-robin measurement scheme. Such constraints may arise as a consequence of the use of a microcomputer as part of the control and signal processing logic. The control problem is shown to be equivalent to a classical sampled-data control problem if a certain observability criterion is satisfied. In particular, the state is shown to be reconstructible from the output measurements, so that any classical state-feedback control can also be expressed in terms of the output measurements.