Controllers for minimizing sample time uncertainty in digital LQG regulation

Consideration is given to the problem of minimizing the effect of uncertainty in the sampling times of analog-to-digital and digital-to-analog conversion in the linear-quadratic Gaussian (LQG) digital regulation of a linear continuous-time system. The overall controller scheme consists of a nominal LQG controller in parallel (as in the class of all stabilizing controllers) with a finite-impulse-response sensitivity-reduction controller. The parameters of the sensitivity controller are selected so as to minimize a performance index which is quadratic in the state, the control, the state sensitivity and the control sensitivity.<>