Control of linear parameter-varying systems using dynamic parameter measurement

This paper extends existing results on a technique for gain-scheduled, L-2-gain control of linear parameter varying (LPV) systems with bounded rates of variation. Specifically, it introduces the use of both memoryless and dynamic (i.e., filtered) parameter measurements when the parameters' accelerations are bounded. Information from the filtered parameter signal can be used to reduce the conservatism of LPV controllers when the parameter variations are slow relative to their worst-case bounds; under such circumstances, the closed-loop performance is closer to that H-infinity-optimal point designs. The synthesis results are illustrated by a simulation example.