Application examples
Noisy signal processing methods

In the field of automation, identification is a body of doctrine whose vocation is the mathematical description of a given system, by determining a model belonging to a given class and compatible with a given pair of input and corresponding output.

Given an unknown system whose behavior is to be described, and a parametric model, linear or not, an identification method consists in modifying the model parameters according to the result of a criterion built on the error e between the system and model outputs, until the model's operation is satisfactory in the sense of the criterion.

As the system's output s (k ) is only available in the noisy form x (k ), the method amounts to searching (figure