摘要：AbstractThis paper studies observer-based state-feedback control of a linear system, where the observation signal is corrupted by the quantization error of an error feedback quantizer that consists of a uniform quantizer and a feedback filter. To mitigate the efect of the quantization error on the system output, we minimize theH2norm of the transfer function from the round-off error of the uniform quantizer to the system output with respect to the observer gain and the feedback filter alternatively. The minimization with respect to the observer gain is formulated into an optimization under a linear matrix inequality and a bilinear matrix inequality, whose minimum is numerically found by an XY-centering algorithm. The minimization with respect to the feedback filter is a convex optimization whose solution can be obtained numerically. The two minimization problems are iteratively solved, until theH2norm converges. A numerical example is provided to see the performance of our method.
关键词：Keywordsnetworked control systemobserverquantizationerror feedback quantizer
