Petri Net Library Research Articles

An object-oriented Petri net simulation tool for hybrid biological processes

Abstract A Petri net library, called PNlib, is presented to enable graphical hierarchical modeling, hybrid simulation and animation of processes in life sciences among others. In order to model these processes, a powerful and universally usable mathematical modeling concept – xHPN (extended Hybrid Petri Nets) – has been established. This formalism is used as specification for the PNlib realized by the object-oriented modeling language Modelica. The PNlib is freely available and can be downloaded on the Modelica homepage www.modelica.org/libraries.

Hybrid Modelling and Process Optimization of Biological Systems

AbstractAn environment has been developed which comprises mathematical methods, concepts, and tools to enable the processing of experimental data to usable new insights about biological systems. Therefore, a mathematical modelling concept - Hybrid Petri-Nets for biological applications (HPNbio) – has been defined which is properly adapted to the demands of biological process. This concept has been implemented with the object-oriented modelling language Modelica. The developed Modelica library, called PNlib (Petri Net library), in combination with an appropriate Modelica-tool enables graphical hierarchical modelling, hybrid simulation, and animation of HPNbio-models. Thereby, an additional Modelica library, called PNproBio (Petri Nets for process modelling of Biological systems), provides wrapped HPNbio which offers on the one hand an easy-use-model at the top level with an intuitive and familiar adapted biological view and on the other hand the flexibility and generality of the HPNbio concept at a lower level. Based on an established model, the underlying processes can be optimized. This process optimization procedure is performed by means of hybrid optimization methods, i.e. a global method is combined with a local method by use of a specific switching strategy, in order to realize an open-loop control for biological processes.

An Advanced Environment for Hybrid Modeling of Biological Systems Based on Modelica

Summary Biological systems are often very complex so that an appropriate formalism is needed for modeling their behavior. Hybrid Petri Nets, consisting of time-discrete Petri Net elements as well as continuous ones, have proven to be ideal for this task. Therefore, a new Petri Net library was implemented based on the object-oriented modeling language Modelica which allows the modeling of discrete, stochastic and continuous Petri Net elements by differential, algebraic and discrete equations. An appropriate Modelica-tool performs the hybrid simulation with discrete events and the solution of continuous differential equations. A special sub-library contains so-called wrappers for specific reactions to simplify the modeling process.The Modelica-models can be connected to Simulink-models for parameter optimization, sensitivity analysis and stochastic simulation in Matlab.The present paper illustrates the implementation of the Petri Net component models, their usage within the modeling process and the coupling between the Modelica-tool Dymola and Matlab/Simulink. The application is demonstrated by modeling the metabolism of Chinese Hamster Ovary Cells.

An advanced environment for hybrid modeling of biological systems based on modelica.

Biological systems are often very complex so that an appropriate formalism is needed for modeling their behavior. Hybrid Petri Nets, consisting of time-discrete Petri Net elements as well as continuous ones, have proven to be ideal for this task. Therefore, a new Petri Net library was implemented based on the object-oriented modeling language Modelica which allows the modeling of discrete, stochastic and continuous Petri Net elements by differential, algebraic and discrete equations. An appropriate Modelica-tool performs the hybrid simulation with discrete events and the solution of continuous differential equations. A special sub-library contains so-called wrappers for specific reactions to simplify the modeling process. The Modelica-models can be connected to Simulink-models for parameter optimization, sensitivity analysis and stochastic simulation in Matlab. The present paper illustrates the implementation of the Petri Net component models, their usage within the modeling process and the coupling between the Modelica-tool Dymola and Matlab/Simulink. The application is demonstrated by modeling the metabolism of Chinese Hamster Ovary Cells.