A framework for automation the generating of machine simulation models

During development, systems to be created are partitioned into several aspects, which are managed as separate documents. Their reuse for deriving other documents, like simulation models, currently involves mostly manual interpretation of the given information. This paper targets the process of creating simulation models, which are semantically derived of already existing information. Based on the steps taken by a simulation expert to create a model, a general approach for transforming data models is presented. A self developed software framework enables the definition of various transformations. Their results are, but not limited to, behavioural simulation models containing kinematic and kinetic associations. As an example, a model is automatically created and used as a machine emulation to perform control software tests.