Some Issues of Parameter Estimation for Linear Systems

Parameter estimation is the process of using system input-output observations to find a set of optimal parameters for a parametric model that allows the model to describe the behaviour of the system. In this thesis, two new parameter estimation methods are proposed: one contributes to the study of observer-based parameter estimation methods by introducing a new sliding mode observer, and the other focuses on the new method design of parameter estimation in the frequency domain. It is believed that the new parameter estimation methodologies developed in this research can be well applied to a wide range of control system designs.