Package Credibility is a free Modelica package to add traceability, uncertainty and calibration information to scalar and 1D-table parameters in a standardized way. For details of this library, see also the journal article Towards Modelica Models with Credibility Information.
Example for the kind of information that can be added to a scalar parameter value (image is from the article above; min/max/unit/description are standard Modelica attributes; the remaining information is new and is added via Modelica parameter records of the Credibility library):
The Credibility library contains an example of a controlled drive train with scalar and 1D-table parameters with credibility information (Credibility.Examples.SimpleControlledDriveNonlinear). The models of this example are then used to demonstrate (for details see the article above):
- optimization based calibration (using the commercial DLR Optimization Modelica library),
- Monte Carlo simulation (demonstrating in particular a new approach for utilizing table-based uncertainty in a Monte Carlo simulation),
- generation of an FMU (Functional Mock-up Unit) where the credibility information of the Modelica model is included in the FMU (generate the FMU from Credibility.Examples.SimpleControlledDriveNonlinear.SimpleDrive_forFMU with any Modelica tool that supports FMU export). With the many available FMI tools, the credibility information can be utilized.
The current approach has the drawback that data needs to be partially manually copied to perform calibration and Monte Carlo simulation in the many available tools for these tasks. The goal is to start a discussion in the Modelica Association how to improve this.
Martin Otter, Matthias Reiner, Jakub Tobolar, DLR - Institute of System Dynamics and Control.
The development of the Library was organized within the European ITEA3 Call6 project UPSIM – Unleash Potentials in Simulation (number 19006).The work was partially funded by the German Federal Ministry of Education and Research (BMBF, grant numbers 01IS20072H and 01IS20072G).
The development of this library is based on work carried out together with Leo Gall and Matthias Schäfer (both LTX Simulation GmbH) in the UPSIM project.