Linear Systems Identification

Identification, presented in this document, is a tool for modeling systems when our knowledge is limited to the evolution of an output parameter as a function of an input parameter. We introduce two groups of methods: analog methods and numerical methods. After an introductory chapter, the second chapter is dedicated to the identification of continuous systems. Our knowledge of the characteristics of first and second-order systems, as well as the concept of dominant pole, will be utilized to develop identification models for systems governed by the evolution of an output as a function of an input. Other methods (Strèjc, Broida, etc.) will also be developed and applied.In the third chapter, as a numerical approach to identification, the least squares method proves to be very promising. The principle of this method, its formalism, and applications will be presented.At the end of this document (Chapter four), an implementation of the various methods will be provided to deepen the introduced concepts.