The use of a digital twin of mechatronic systems enables design optimizations, virtual tests and validations during the development phase as well as continuous monitoring during operation. By integrating trustworthy data sources, the theoretical behavior of the digital twin or the assembly-specific digital fingerprint is analyzed and compared with the real application. The AI-based self-validation of electronic systems based on innovative hybrid model approaches is being continuously expanded for this purpose. For functional structures in the automotive, rail technology and aviation sectors, corresponding hybrid models are realized using both open source and commercial software environments. With the grey-box models developed to complement the hardware, the system level can be fully mapped right down to the assembly and connection technology.
- Multi-domain system simulation: Electrical, thermal, mechanical modeling
- Hybrid grey-box modeling (combination of physical and data-driven model approaches)
- Integration of failure mechanisms to evaluate the remaining useful lifetime (RUL)
- Development of efficient dynamic simulations using model order reduction (MOR, ROM)