Two algorithms for model quality estimation in state-space systems with time-varying parameter uncertainty

We present two methods to estimate bounds of parameter uncertainty in state-space systems. In the first method, we minimize the loo-norm of the perturbation and its derivative. In the second method, an estimate of the perturbation is produced based on a quantized approximation of the uncertainty and the sparse structure of its derivative. Less sensitivity to increased noise and changed model parameters is achieved by the second method. We use an overhead crane as an illustrative example.