Analysis of steady-state tracking errors in sampled-data systems with uncertainty (Q5935482)

The paper deals with robust steady-state \(\varepsilon\)-tracking of sampled data systems in the presence of structured norm bounded time-varying finite memory uncertainty. An appropriate steady-state norm has been defined, and a steady-state norm matrix has been constructed based on this norm. Exact conditions for robust \(\varepsilon\)-tracking of sampled data systems were derived in terms of the spectral radius of this matrix. The results show that even if zero steady-state tracking can be achieved for the nominal system, the steady-state tracking error may no longer be zero in the presence of time-varying uncertainty in the system, and that it can be arbitrarily large. An example for computing the steady-state tracking error for a sampled data system with uncertainty is provided.