System Identification

System identification is the general term of extracting parameter and model information from gathered data. Such data-driven modeling is widespread in a lot of scientific domains including engineering, natural science and even econometrics. This chapter focuses on the basics of identification techniques for the instrumentation and measurement society, including the predictive capabilities of the models, the physical interpretation of the parameters, and a quality stamp through the computation of parameter uncertainties and model validation. All parameter estimation techniques, their uncertainty and the model validation will all be based on statistics.

The aim of this chapter is to give an educative introduction in system identification using a simple example which covers a lot of system identification aspects and positions these aspects in an instrumentation and measurement framework.

The examples are worked out in the same systematic way to stress the different choices in identification. Furthermore, the complexity will gradually increase, each time highlighting an important aspect of identification.

After completing the example, some initial directions are given to state some typical problems and to help the reader in finding some starting literature.

ISBN:978-1-118-17131-8

Keywords:measurement, system identfication, modelling