Parameter Identification of Systems with Uncertain Dynamics and Disturbances

A method is presented for parameter identification of systems with uncertain dynamics and input/output disturbances. The system model is in stable factor form and is assumed to contain both parametric and nonparametric uncertainty, but only parametric uncertainty, namely, unknown parameters, are identified from the available input/output data, which are contaminated by norm-bounded nonparametric uncertainty and disturbances. A finite-time energy bound of the perturbation due to nonparametric uncertainty and disturbances is constructed in real time and is used in the parameter identification algorithm to ensure monotone reduction of parameter errors.