(456e) MOSAIC: A Web-Based Modeling Environment for Process Systems Engineering
AIChE Annual Meeting
2010
2010 Annual Meeting
Computing and Systems Technology Division
Chemical Engineering Applications of Web 2.0 and the Semantic Web
Wednesday, November 10, 2010 - 2:30pm to 3:00pm
In research, development, and the educational sector, the cooperation on internet becomes increasingly important. The personal workforce is locally decentralized and information technology is used to coordinate the collaboration and to merge the respective results into large research projects. One key issue in this process is the centralized access of cooperative results in knowledge databases. This allows for the use and the reuse of information created by collaborating researchers. The used set of mathematical equations describing process systems related phenomena and their assembly to models for different applications is an important part of this sharable knowledge. This kind of knowledge is traditionally shared via public or internal publications i.e. mathematical formulations are exchanged in the documentation level. In the documentation level, model equations and functions are laid down in a two-dimensional symbolic mathematic language. For custom models, a wide variety of numerical softwares are available to solve the resulting equation systems. However, it is necessary to translate the model given in the documentation level into the programming language of the respective tool, which results into a numerical model. This translation step takes engineer resources and it is prone to errors. Furthermore, the resulting numerical model cannot be exchanged unless the cooperating workers use the same numerical environment, which represents a considerable limitation for collaboration.
There are a lot of well established models and programs for the simulation of process systems engineering models. Depending on the application, the mathematic models differ in their level of detail, but they are always expressed as differential algebraic equation systems. However, in a lot of situations, the standard models are not applicable. In those cases, custom models must be provided, which is generally done by specifying the applicable system of model equations.
In this work, a new modeling environment for the creation, administration and translation of custom models is proposed that addresses the issues above. The proposed modeling environment is named MOSAIC (Modelling, Simulation, Application and Interaction for Chemical processes) and it is provided as Software as a Service. The tool consists of a model editor and evaluation environment that is implemented as a Java Applet and a modeling server that allows for collaborative storage and controlled access of the entered model equations, equation systems and case studies. The environment can be integrated into other Cloud Computing applications. The mathematical and structural content of the models is stored in XML and MathML. To further the reuse of models, the new modeling environment provides a very high extend of modularity. Equations are defined in separate modular structures that are independent from equation systems. An equation system is build from modular equations or from other equation systems.
The modeling environment MOSAIC is a versatile code generator that allows for the mathematical models to be translated into programming languages suitable for pertinent numerical environments. At present, translation is supported for the BzzMath Library (C++) [1], MATLAB, gPROMS [2], Microsoft Excel, and others. The proposed modeling environment MOSAIC tries to narrow the gap between the model description in the documentation level and the numerical model. This is done by using two-dimensional symbolic expressions for the mathematical content and by considering the notation as separate and mandatory model element belonging to equations and equation systems. The notation carries descriptions for all identifiers that are present in the equations. More importantly, the notation will carry information about engineering units. Applicable data structures allow the translation of one model in the variable naming and the engineering units setup of another. To introduce the proposed modeling environment, a process related to the oxidative coupling of methane is implemented in MOSAIC. This complete process is being investigated in a cooperation project within the framework of UNICAT (unifying concepts in catalysis) [3]. It will be demonstrated, how the new modeling concepts allow the cooperation of two groups that use different naming conventions and work without copy-pasting of equations or any agreements on interfaces. The internal XML/MathML model will be discussed briefly and the generated code is shown (exemplary translations to C++/BzzMath [1], MATLAB, and gPROMS [2]). Furthermore, the integration of the modeling environment into other Cloud Computing Applications is presented in the example of a collaboration software [4].
[1] D. Manca, G. Buzzi-Ferraris, Comp. Aided Chem. Eng., 25(2008) 587-592.
[2] M. Oh, C. C. Pantelides, Computers chem. Eng., 20(1996) 611-633.
[3] http://www.unicat.tu-berlin.de/
[4] G. Wozny, A. Klein & a. Grote, 2009, Chem. Eng. Transactions, 18(2009) 291-295.
Checkout
This paper has an Extended Abstract file available; you must purchase the conference proceedings to access it.
Do you already own this?
Log In for instructions on accessing this content.
Pricing
Individuals
AIChE Pro Members | $150.00 |
AIChE Graduate Student Members | Free |
AIChE Undergraduate Student Members | Free |
AIChE Explorer Members | $225.00 |
Non-Members | $225.00 |