The core innovation areas of OpenCPS

FMI run-time and master simulation framework including UML/Modelica interoperability: Support of distributed co-simulation that is more scalable than existing systems and optimized for handling large numbers of discrete-time events. This will allow a more scalable and reliable co-simulation of discrete-time software parts with continuous-time, physical processes. Furthermore, in contrast to available closed-source software package the OPENCPS FMI co-simulation will be based on an extensible open-source framework whose plug-in architecture will facilitate customization and uptake for industrial use-cases. The participation of the two leading open-source platforms for UML and Modelica will ensure that detailed know-how of software modelling (UML) and physical modelling (Modelica) is available for providing top-class support for the demanding cyber-physical applications of the industrial partners. New technology will be developed, partly based on the FMI standard, to allow development and run-time interoperability of models consisting of both UML and Modelica parts. To ensure the usability of this framework and beside its ability of integration, it is relevant to improve its simulation and validation capabilities.

State Machine and Real-Time Debugging & Validation: Industry-strength support for advanced state-machine modelling including smooth integration and conjoined debugging of state-machines with other modelling paradigms for real-time systems. Basic state-machines are a well-known concept in computer science, but using them as visual formalism for complex discrete-event systems requires more advanced extensions as provided in UML state diagrams or in the Modelica state machine extension. The project aims to provide tight integration between debugging and validation support for such advanced state-machines and complementing real-time systems modelling paradigms (e.g., clocked synchronous data-flow models) which extends the abilities of currently available tools. Two levels of debugging support are planned: limited debugging supporting (connected) black-box FMU components, and full debugging capabilities for components for which the model source code is available.

Efficient Multicore simulation: Improved compilation and simulation capabilities for large models. There is an increasing demand in industry to move away from special purpose simulation codes for large-scale simulations to more generic and easier maintainable and extensible modelling solutions, e.g., based on Modelica. However, realizing this vision requires that generic equation-based object-oriented (EOO) tools, like OpenModelica, are improved to handle such models almost as efficient as the special purpose code. The aim is to improve the performance of EOO tools and the developed FMI-based master simulation framework by requirements from benchmarks provided by industrial partners (e.g., simulation of national/European electrical power and grid systems). Efficient multi-core simulation techniques will be developed and applied at several levels, from coarse-grained, running whole simulations and/or FMUs in parallel, to more fine-grained by parallelization of equation models and algorithmic code inside model components.