OpenModelica 1.14.0 Final Release – Open Source Cyber-Physical Modeling and Simulation
OpenModelica is an open-source Modelica-based cyber-physical mathematical modeling, architectural description, simulation and optimization environment intended for industrial and academic usage. Its long-term development is supported by a non-profit organization – the Open Source Modelica Consortium (OSMC). For download and more information about the release, see www.openmodelica.org<http://www.openmodelica.org>. (If you do not wish to receive these messages, see the end of this email). See further below regarding the OpenModelica workshop February 3, 2020, and MODPROD workshop February 4-5. This OpenModelica 1.14.0 Final release December 6, 2019, includes more than 200 improvements and fixes compared to the previous 1.13.2 release. It is available here: (Download Linux<https://openmodelica.org/download/download-linux>, Mac<https://openmodelica.org/download/download-mac>, Windows<https://openmodelica.org/download/download-windows>. Report bugs in Trac<https://trac.openmodelica.org/OpenModelica/newticket>.) . The most recent documentation is here<https://www.openmodelica.org/doc/OpenModelicaUsersGuide/latest/>. The most dramatic enhancement of the OpenModelica Compiler is the New Frontend, which on the average gives a factor of 10-20 speed improvement in the flattening phase of compilation. The new frontend is default in this release, but a feature allows the user to switch to the old frontend if problems appear for a specific model. The speed of the OMEdit GUI has only slightly increased in this version, since it is still dependent mostly on the old frontend. Further GUI speed increases are available in the coming OpenModelica.1.15.0. The OMEdit graphical model editor has a number of improvements, including: Drag and drop for the text layer, Auto completion of class names, components and annotations, Improved duplication functionality for copying classes, Better handling of Compiler flags, Autocomplete annotations, Model copy paste functionality, Array plots update on re-simulation, Support for connectorSizing annotation, Drag and drop class names to text layer in OMEdit, OMPlot: Improved plotting e.g., top and bottom margins for better view, snap to curve etc, and many more. OpenModelica compiler backend and run-time system: A new more efficient and correct implementation of arrays and records, The FMI OMSimulator API calls are now also available in the OMC API functions, e.g. for use from OMNotebook, Jupyter notebooks, Added possibility to generate analytic Jacobians for linear strong components, Added output language options for linearization: Malab, Python, Julia, Added result check for linear solvers Lis, Klu, Total Pivot and Umfpack if a residual function is available, Improved debug dumping, Homotopy: Use simplified version only during initialization to avoid errors during matching and differentiation, and many more. Moreover, complete implementation of OMMatlab is now available. OMSimulator (FMI simulation) news: Improvements in error control and numeric stability including numerical adaptive step size control. Achieved by adding adaptive step-size control for hybrid CT simulation without rollback, method developed by Rebeka Farkas (presented at 13th Int. Modelica Conference, https://modelica.org/events/modelica2019/proceedings/html/papers/Modelica2019paper4D2.pdf). Not all tools exporting FMUs provide the optional FMI features, e.g. fmi2SetFMUstate and fmi2GetFMUstate. These are usually required for algorithms providing error control and adaptive step-sizes. OMSimulator now provides a method that doesn’t require these optional FMI features and uses instead additional knowledge about the simulation units provided by the modeler, Added adaptive step-size control based on roll-back mechanism, Improved handling of output/input derivatives, Provide first order derivatives for all continuous outputs of lookup tables, Fixed signal filter for result files, New functionality to add cpu time to the results for each simulation component, Improved FMI export from OpenModelica (further improvements ongoing now), and many more. Call for Participation – the OpenModelica and MODPROD workshops February 3-5, 2020 Linköping, Sweden Location: Linköping University, Linköping, Sweden · 14th MODPROD Workshop on Model-Based Cyber-Physical Product Development, February 4-5, 2020 · 12th OpenModelica Annual Workshop, February 3, 2020 · INCOSE workshop, February 3, 2020 Theme for this year: Digital Meets Reality You can find more information here: www.modprod.liu.se<http://www.modprod.liu.se> Preliminary program and registration information will soon appear. Keynotes for MODPROD2020 Workshop: · Peter Gorm Larsen, Professor at Aarhus University, Head of Digit Centre. “A Cloud-based Collaborative Platform for Model-Based development of Cyber-Physical Systems” · Richard Romano, Professor at University of Leeds, Chair in Driving Simulation. “Model based design of Automated Vehicles” · Gunnar Cedersund, Associate Professor at Linköping University, leader of Systems Biology group, “A digital twin for the human body: from data and validation to end-usage in healthcare and drug development”. · Juan Llorens, Professor at the Informatics Dept., Univ of Madrid, President of INCOSE Spain, CTO of REUSE company. "Knowledge Management and reuse in the context of Systems Engineering”. --------------------------------------------- The OpenModelica development work is organized by the Open Source Modelica Consortium, currently about 50 organizations. Thanks to all who contributed and worked hard for the OpenModelica 1.14.0 release! Best regards, Peter Fritzson Director of the Open Source Modelica Consortium Professor at Linkoping University ---------------------------------------------------------------------------------------------------------------- Note: This email list has a traffic of approximately 3-8 messages per year. If you ever want to remove yourself from this mailing list, you can send an email to: [email protected]<mailto:[email protected]> You will get a confirmation email that you have to reply to. If that does not work, send an email to [email protected]<mailto:[email protected]>
_______________________________________________ Openmodelicaannounce1.ida mailing list [email protected] https://lists.liu.se/mailman/listinfo/openmodelicaannounce1.ida
