L. von Wedel

CHEOPS: A tool-integration platform for chemical process modelling and simulation

A large number of modelling tools exist for the construction and solution of mathematical models of chemical processes. Each (chemical) process modelling tool provides its own model representation and model definition functions as well as its own solution algorithms, which are used for performing computer-aided studies for the process under consideration. However, in order to support reusability of existing models and to allow for the combined use of different modelling tools for the study of complex processes, model integration is needed. This paper presents a concept for an integration platform that allows for the integration of modelling tools, combining their models to build up a process model and performing computer-aided studies based on this integrated process model. In order to illustrate the concept without getting into complicated algorithmic issues, we focus on steady-state simulation using models comprising only algebraic equations. The concept is realized in the component-based integration platform CHEOPS, which focuses on integrating and solving existing models rather than providing its own modelling capabilities.

Rome: A repository to support the integration of models over the lifecycle of model-based engineering processes

Process modeling is currently performed by using a number of incompatible tools. The heterogeneous storage formats and systems of these modeling tools limit the reuse of models because semantic relationships among models originating from different or even the same tool cannot be maintained. A repository is proposed as a central storage system for models from various sources represented in various formats. The repository holds models representing chemical engineering and physics concepts by a conceptual data model rather than as mathematical equations or some programming language code. The conceptual representation of a model is accompanied by tool specific representations and documentation, enhancing the use of models over the lifecycle of model-based engineering processes such as chemical process design. The conversion among different tool formats is simplified by reducing the number of required converters if tools are connected to the repository instead of integrating them pair-wise.

A tool integration framework for dynamic simulation in process engineering

Abstract As a fundamental technology for the design and operation of process plants, simulation has been applied through different simulation tools. To perform complex simulation tasks in a technically and economically justified manner, it is sometimes necessary to integrate some of these tools to exploit their strengths. In this work, a “simulation framework” approach is proposed towards systematic support to tool integration. A number of key issues pertaining to simulation tool integration are identified and analyzed. A prototypical simulation framework –Cheops is developed to provide a solution to those issues. An application scenario connecting different dynamic simulation tools has been used to explain the integration issues and to guide the development work in Cheops.

Verwaltung und Wiederverwendung von Modellen im industriellen Entwicklungsprozess

The use of model-based methods for the support of the planning, designing, and operating of chemical processes and plants has recently gained increasing importance in the daily work of the process engineer. However, it is becoming more obvious that already the design of mathematical models by the use of standard flowsheet simulation tools requires an experienced designer. Therefore, the developed models not only represent a value considering the resources required for their development, but especially they contain a large amount of knowledge of an enterprise about its products and production processes. It is hence reasonable to lower the costs of model development by reuse of models and experiences. This report gives an overview about the systematic long-term management and use of mathematical models. Some motivating scenarios help to derive requirements for a model management system and to evaluate existing methods and tools. From these results, the concept of a model management system is derived. After presenting a prototypical implementation, this article concludes with an outlook.

A library for equation system processing based on the CAPE-OPEN ESO Interface

Abstract In this contribution we present the LptNumerics library for processing of equation systems that are provided via the CAPE-OPEN ESO interfaces. The library provides functionality for the pre-processing of an equation system and its solution. Here the term pre-processing refers to the steps required prior to solving a model in an equation-oriented solution approach. The library is implemented in C++ and can be used within any equation-oriented application. This contribution discusses the design and implementation of the library and presents two application examples.

Using design prototypes to build an ontology for automated process design

Recently there has been an increased interest in agent-based environments for automated design and simulation (Garcia-Flores et al. 2000). In such environments, responsibility for decision making is partly removed from the engineer to the underlying agent framework. Thus, it is vital that pertinent knowledge is embedded within this framework. This motivates the development of an ontology (Gruninger & Lee 2002) for the particular domain. An important first step is a suitable organisation of the knowledge in a given domain.

An integration of design data and mathematical models in chemical process design

Abstract Model-based methods, provided by process modeling environments, nowadays support a variety of activities throughout the lifecycle of designing and running chemical processes. On the other hand, computer-aided engineering (CAE) systems have become popular for maintaining the information relevant to describe the function of the process as well as equipment data. This contribution proposes an integration of these two specialized types of systems. First, such an integration allows to understand the design context in which models have been used. Further, an export functionality of the CAE system allows the modeler to generate a skeleton model that can be modified or enriched, instead of starting modeling from scratch for each model-based task.

An open thermodynamics server for integrated process engineering environments

Publisher Summary Computer-aided process engineering (CAPE) practice has shown that thermodynamics software tools play a critical role in the process engineering lifecycle, including process modeling, simulation, optimization, and process design. To provide CAPE applications unified and improved thermodynamics services, a component-based open thermodynamics server (OpenThermo) has been developed. OpenThermo is able to carry out open and closed-form physical property calculations and to supply and process symbolic thermodynamic model equations. These services are provided based on the integration of individual thermodynamics resources, such as physical property databases and physical property packages. Such functionality is achieved with a layered software architecture following a resource/service-based classification of the software components that comprise the thermodynamics sever. So far, the major functionality of the service components has been implemented, and several resource components have been connected to OpenThermo.