Rapidly validate systems performance using system simulation and reduced order modeling techniques
Accelerate validation of systems and controls during integration with neural networks
Time-to-market has become an even more critical element for OEMs to differentiate. The strong push toward electrification requires managing the growing complexity of systems to design to meet strict targets for all performance attributes and be the fastest to do it. It involves engineering departments to deploy and implement new simulation methodologies widely.
Validating the subsystems' choice and related controls involve the virtual evaluation of their performance levels during integration. Detailed models used for components sizing can no longer be used. New techniques are available to rapidly adapt model levels to integration constraints and speed up model usage using reduced-order model techniques.
In this webinar, our experts will introduce how to use neural networks techniques to create and integrate surrogate models with system simulation to accelerate the validation of systems and controls during integration.
Speed up simulation model usage using reduced order model techniques
What will you learn in this webinar?
- Identify when and on which subsystem surrogate modeling can bring benefit
- Generate a reference dataset for the subsystem
- Train and validate neural network and optimize surrogate model response using engineer knowledge
- Application case: integrate into a complete vehicle model an HVAC loop connected to a chiller for BEV thermal management
About the speaker
Business Development Manager, Siemens Digital Industries Software
Thomas is the thermal applications business development manager for the Simcenter portfolio at Siemens Digital Industries Software. He became the manager of the 1D application engineers’ team in France for over five years with a successive specific focus on territory, train and the automotive market.