Model-based computer aided product-process engineering has attained increased importance in a number of industries, including pharmaceuticals, petrochemicals, fine chemicals, polymers, biotechnology, food, energy and water. This trend is set to continue due to the substantial benefits computer-aided methods provide. The key prerequisite of computer-aided product-process engineering is however the availability of models of different types, forms and application modes. The development of the models required for the systems under investigation tends to be a challenging and time-consuming task involving numerous steps, expert skills and different modelling tools. This motivates the development of a computer-aided modelling framework that supports the user during model development, documentation, analysis, identification, application and re-use with the goal to increase the efficiency of the modelling process. The developed modelling framework is structured based on the work-flow and data-flow the modeller needs to follow to fulfil the desired modelling task. The framework consists of two main parts (work-flows) for model development. The first part is dedicated to single-scale model development while the second part supports the modeller during the multi-scale scenario development and comparison. The second part of the modelling framework manages different multi-scale scenarios, supports in the systematic derivation of a new scenario and links the modelsfor the different scales according to the linking scheme of the multi-scale scenario. This is where the single-scale model development part is connected to the multi-scale scenario development part because it assists the modeller in providing the single-scale models that are being linked to form the multi-scale scenario. The single-scale model development part supports the modeller in model documentation, construction and analysis.Different models for properties, phenomena, unit operations, processes can be developed and analysed here or retrieved from model libraries.The model equations are introduced in a simple text format and are translated by the modelling tool with thereverse polish notation (RPN). A model object is generated which can be applied in a stand-alone-mode, stored in libraries and/or linked to other models. The model analysis section contains different features like variable classification, degree of freedom analysis, incidence matrix generation, optimal equation ordering, eigenvalue analysis. Once the models have been constructed and analysed the modelling framework incorporates 3 application work-flows for: identification, simulation and design. For these application work-flows different solvers that can solve a large range of different problems are connected to the modelling tool. The tool is able to determine the solution strategy together with the required solvers based on the results of the model analysis during model development. For model identification features like sensitivity analysis, identifiability analysis, optimizer, confidence interval calculation and uncertainty analysis have been incorporated into the tool. The structure of the modelling tool as well as its features will be highlighted through a case study involving a problem from industry related to multi-scale model development and application.
Computer-aided Chemical Engineering, 2011, p. 16-20