Carla Martin-Villalba

Development of virtual-labs for education in chemical process control using Modelica

Abstract Virtual-labs are useful tools for education in chemical process control. A novel approach to virtual-lab implementation using the Modelica language is presented. This approach has the following advantages. Firstly, existing Modelica models can be employed to develop virtual-labs. If new models need to be programmed, the use of Modelica reduces considerably the modeling effort. Secondly, as the complete virtual-lab is described in Modelica, virtual-lab developers don’t need to use programming languages. Finally, virtual-labs can be easily distributed to the users, who don’t need to install any additional software. The proposed approach is based on the application of a methodology for interactive model development, and two software tools: a free Modelica library, named Interactive ( http://www.euclides.dia.uned.es/Interactive ), and the Dymola modeling environment. Several virtual-labs have been implemented and successfully applied to education in chemical process control. The development of a virtual-lab based on a complex model of a double-pipe heat exchanger is discussed.

An approach to virtual-lab implementation using Modelica

A novel approach to the implementation of interactive virtual-labs is proposed. The virtual-lab is completely described in Modelica language and translated using Dymola. To achieve this goal, a systematic methodology to transform any Modelica model into a formulation suitable for interactive simulation has been developed. In addition, VirtualLabBuilder Modelica library has been programmed. This library contains a set of Modelica models of visual interactive elements (i.e. containers, animated geometric shapes and interactive controls) that allows easy creation of the virtual-lab view (i.e. the model-to-user interface). This approach has two strong points. Firstly, it allows taking advantage of the Modelica capabilities for multi-domain modelling and model reuse. In particular, existing Modelica libraries for modelling of physical systems can be reused to build the virtual-lab models. Secondly, VirtualLabBuilder library allows describing the virtual-lab view with Modelica, which facilitates its development, maintenance and reuse. The proposed approach is discussed in this manuscript and it is illustrated by means of the following case study: a virtual-lab describing the thermodynamic behaviour of a solar house. VirtualLabBuilder library can be freely downloaded from http://www.euclides.dia.uned.es/

Development of an industrial boiler virtual‐lab for control education using Modelica

The use of the VirtualLabBuilder Modelica library and the Dymola modeling environment facilitates the implementation of virtual‐labs with elaborated user interfaces, and based on large and complex Modelica models. This implementation methodology is applied to develop an industrial boiler virtual‐lab for process control education. VirtualLabBuilder is freely available at www.euclides.dia.uned.es.

Visualization and interactive simulation of Modelica models for control education

An approach to the implementation of virtual-labs well-suited for control education is proposed. The virtual-lab model is described in the Modelica language and it is adapted for interactive simulation by applying a systematic methodology. The virtual-lab view (i.e., the user-to-model interactive interface) is composed by connecting predefined components of Interactive, a free Modelica library presented in this manuscript. Interactive contains a set of ready-to-use Modelica models of visual interactive elements (i.e., containers, animated 3D geometric shapes and interactive controls), intended to facilitate easy definition of the virtual-lab view and the model-to-view communication. Finally, the Modelica description of the virtual-lab is translated into executable code using the Dymola modeling environment. The interactive component models provided in the Interactive library are C++ code generators: when the virtual-lab is executed, the C++ code describing the virtual-lab view and the bidirectional model-to-view communication is automatically generated. The generated C++ code uses the VTK and Qt libraries. The proposed approach is illustrated by means of an example: the development of the virtual-lab of a double-pipe heat exchanger, which is intended for chemical-process control education.

Development of virtual-labs based on complex Modelica models using VirtualLabBuilder

The feasibility of implementing virtual-labs based on complex Modelica models is demonstrated. To this end, the design and implementation of the following three virtual-labs is discussed: 1) a virtual-lab of a double-pipe heat exchanger, which is a useful tool for control education; 2) a virtual-lab describing the thermodynamic behaviour of the solar house, which is based on a complex Modelica model developed by other authors; 3) the virtual-lab of a drum-type washing machine, which is an industrial application useful as design aid. The graphical user interface has been implemented by using the VirtualLabBuilder library, which can be freely downloaded from http:/ /www.euclides.dia.uned.es.

Two Approaches to Facilitate Virtual Lab Implementation

Here, two software tools are presented that facilitate virtual lab implementation: Interactive and Rand Model Designer. Interactive is a free Modelica library thats used in combination with the Dymola modeling environment. Rand Model Designer supports Model Vision Language, an object-oriented modeling language based on the Unified Modeling Language (UML).

Virtual-Lab of a Cement Clinker Cooler for Operator Training

Abstract Plant operator training plays a fundamental role in improving the energy efficiency of the the cement manufacturing process and reducing CO 2 emissions. A virtual-lab of a clinker grate cooler, intended for training of cement plant operators, has been developed. The grate cooler model has been derived from first principles, and has been validated consulting cement industry experts, and comparing the simulated results with published data and available information from cement industries. The model has been described in the Modelica language. The Interactive Modelica library has been used to develop the interactive user-to-model interface, and the communication between this interface and the model. The virtual-lab, which is completely described in Modelica, has been simulated using the Dymola modeling environment. The Interactive Modelica library can be freely downloaded from the website http://www.euclides.dia.uned.es/

DEVS Graph In Modelica For Real-Time Simulation.

Two new Modelica libraries are presented. The first library, named GGADLib, supports the DEVS graph notation in Modelica. The second Modelica library, named UDPLib, allows sending and receiving data using the User Datagram Protocol (UDP). GGADLib uses UDPLib. As a result, DEVS graph models composed using GGADLib can receive and send data (input and output events, according to DEVS terminology) using UDP. This feature allows Modelica DEVS graph models to communicate with hardware and with other models, e.g., with models developed using other languages and tools, and running in other computers. The implementation of UDPLib and GGADLib is discussed in this paper. Their use is illustrated by means of a case study: evaluating an obstacle avoidance controller for the e-puck mobile robot. The UDPLib and GGADLib Modelica libraries can be freely downloaded from http://www.euclides.dia.uned.es/

Virtual lab in Modelica of a cement clinker cooler for operator training

Plant operator training plays a fundamental role in improving the energy efficiency of the cement manufacturing process and in reducing the CO2 emission. A virtual lab of a clinker grate cooler, intended for the training of cement plant operators, has been developed. The grate cooler model has been derived from first principles, and has been validated by consulting cement industry experts, and comparing the simulated results with published data and available information from the cement industry. The model has been described in the Modelica language. The Interactive Modelica library has been used to develop the interactive user-to-model interface, and the communication between this interface and the model. The virtual lab, which is completely described in Modelica, has been simulated using the Dymola 6.1 modelling environment. The Interactive Modelica library can be freely downloaded from the website http://www.euclides.dia.uned.es/

Simulation of Variable Structure Models Using Rand Model Designer

Rand Model Designer (RMD) is a modeling environment that supports Model Vision Language, an object-oriented modeling language for hybrid-DAE systems. Model Vision Language allows describing the continuous-time part of the model combining the use of equations (i.e., a causal modeling) and sequences of assignment statements. The hybrid behavior is described using behavior charts (B-chart). A B-chart is a diagram consisting of modes, transitions internal to the modes and transitions between modes. The activity of the mode is specified by a local class (the so-called activity class of the mode) that describes the system structure and behavior while in this mode. The system of equations that describes the complete model at certain time is automatically built by RMD at run time. Every time a transition takes place (i.e., the model structure changes), RMD constructs the mathematical description of the actual model, eliminates the redundant variables and trivial equations resulting from component connections, analyzes the model solvability and structure, selects the best-suited numerical method and generates the input to this numerical method. This approach provides high flexibility in the description of variable structure models. This feature is demonstrated using variable structure models arisen in two different applications. The first application is the run-time change in the selection of the model state variables. Support for several selections of the state variables is typically required in interactive simulations. The second application is the description of systems with variable behavior. The model of an industrial boiler is used to illustrate the description of this type of variable structure models.