摘要：AbstractThe current paper proposes an approach to adequately select an optimal sampling period for linear time-invariant (LTI) regulators targeted for practical implementation on microprocessor-based systems. Although the classical Shannon-Nyquist sampling theorem proposes the solution in order to completely recover the original analog signal from its discrete samples, the scope of the sampling process, viewed from a control perspective, is to ensure asymptotic stability of the closed-loop system, with as small degradation of the initially-proposed performance indices as possible. As such, we propose a configurable optimization-based approach for selecting the sampling period by specifying the emphasis between closed-loop fidelity and ease of implementation. The proposed method is illustrated on a robust controller, having an anti-resonant peak, designed for a nonlinear and uncertain DC-to-DC SEPIC converter.
