Model reduction by identification. Linear case
Inverse techniques and parameter estimation Part 2

We propose here a practical application of the concepts seen previously, consisting in the identification of a reduced-order dynamic model, and then the use of such a model to estimate time-dependent input quantities. The linearity of the model is understood here in relation to its inputs, cf. [AF 4 515, equation (10)].

The first phase of model identification, requires us to give ourselves a model equation structure, defined by its parameters (matrices) which we then seek to identify through its output vector Y_mo;;. This is compared :

...