Model Verification and Validation in Engineering: Updating of Mathematical Models by the Use of Test Data

Mathematical models are uncertain. Therefore they have to be verified and validated in order to obtain usable models.

Models are maps of systems. Thus, one first has to define the expressions, and secondly, one has to give the relations between them.

The best available information about the system stems from the existing system: measurements. Consequently, system identification methods are used for model verification and validation.

The uncertainties of the prior model and of the measurements have to be modelled. This is done probabilistically. The resulting updating methods are discussed, including the problems arising, and procedures for overcoming them.