An empirical investigation of the achievable performance of sampled-data controllers with exponentially-weighted sampling and reconstruction

We investigate the effect of an exponential hold paired with an exponentially-weighted sampler on the achievable H/sub 2/ performance of a sampled-data system. In doing so, this paper represents, to our knowledge, the first use of a generalised sampled-data sampling function (GSDSF) which is the dual to the generalized sampled-data hold function. By using a GSDSF, we can provide an exact treatment of the white measurement noise required by LQG theory without introducing artificial discrete time measurement noise.