A model-based methodology for tool supported design of automated systems

The model-based design of production processes and corresponding control systems yields a large amount of design data with complex interdependencies, which makes it difficult to detect inconsistencies that may lead to errors in an early stage of the design process. For different stages and parts of the design process, a variety of models in different formalisms is often used. The selected model formalisms restrict the choice of usable tools in each stage severely or require the new creation of a model in another formalism due to missing tool integration. Integrated model-based design promises to reduce the design effort significantly, and thus the cost, while improving the quality of the designed system by ensuring that errors are detected in an early stage, i. e. when it is cheap to correct them. This dissertation illustrates a recently developed software support system for model-based design of complex automated systems, using two challenging case studies, the model-based design of a pipeless plant and of the controller of a combustion engine carburetor. The inherent complexity of the case studies requires the features of the software support system, e. g. structured data and model management, algorithms for design consistency and parameter propagation, and the integration of model-based tools by model transformations, to ensure an efficient and high-quality design process. The model transformations are illustrated for the connection of the modeling language gPROMS and the interchange format CIF. A selection of transformation tool chains exemplifies their benefits. It is shown that tool supported model-based design, especially if aided by model transformation, has beneficial effects on the efficiency and quality of the workflow.