4. Parametric estimation methods for continuous-time models

Linear parametric continuous-time model identification techniques are mainly based on the minimization of a criterion based on either an output error or an equation error, requiring the use of a linear transformation coupled with a least-squares method. Numerous equation error methods have been proposed over the last thirty years. The main advantage of these methods, based on equation error, is the formulation of an explicit solution that leads to a unique optimum. However, these techniques require the estimation of successive derivatives of the input/output signals. Various methods have been developed to solve this inverse problem. A number of publications or books are available on the state-of-the-art in continuous-time model identification approaches, the most significant of which are