Belangrijkste concepten
Linear-matrix-inequality-based techniques are proposed for sampled-data controller synthesis to render closed-loop systems dissipative, with applications in passivity and H∞-norm minimization.
Samenvatting
The paper introduces techniques for designing sampled-data controllers to achieve dissipativity using linear periodic jump-flow system models. Three applications are presented: passivity-based controller synthesis, filter matching, and controller redesign. The results show improved closed-loop behavior compared to traditional discretization methods.
Statistieken
The vast majority of controller implementations have shifted from analogue to digital controllers over the last century.
SDCS design was successful in the 90s due to a technique called 'lifting'.
SDCS can be modeled as linear periodic jump-flow systems (LPJF).
LPJF systems enable designing controllers using Riccati differential and difference equations.
DLMIs can be turned into numerically tractable LMIs by approximating unknown variables.
LPJF systems can be made dissipative with respect to quadratic storage functions.