Development of a Modelling Procedure for the Description of Test Set-Ups with Measuring Instruments

Measurement techniques are a highly complex field in which small technical details decide over the accuracy and therefore the success or failure of a measurement. What sensors can be used? How can noise be handled properly? These and many more are questions one has to answer when wanting to capture the measurand of interest.

Modelling a test set-up requires experience and is therefore mostly done by experts in the field. The lack of uniform modelling approaches makes this an even bigger problem, because discussing models can lead to confusion when a common language is missing. Furthermore, the validation of a model on different abstraction levels has to be done manually in the absence of the necessary tooling, which is error-prone.

This thesis provides a common formal language for describing test set-ups. It does so by proposing a modelling approach for test set-ups which allows the definition of different abstraction levels on top of a general set of rules. These rules allow the validation of the set-up. For example, it can be validated if components are configured correctly. This was achieved by defining a set of validity constraints in first-order logic. These constraints are then applied to a graph representing the test set-up. A formal specification of the model is provided and backed by an implementation in the language Prolog.