Reduced Order Modeling with AI: Accelerating Simulink Analysis and Design

With Model-Based Design, you use virtual models to design, implement, and deliver complex systems. Creating high-fidelity virtual models that accurately capture hardware behavior is valuable and can be time consuming. However, these high-fidelity models are not suitable for all stages of the development process. For example, a computational fluid dynamics model that is useful for detailed component design will be too slow to include in system-level simulations to verify your control system or to perform system analysis that requires many simulation runs. A high-fidelity model for analyzing NOx emissions will be too slow to run in real time in your embedded system. Does this mean you have to start from scratch to create faster approximations of your high-fidelity models? This is where reduced order modeling (ROM) comes to help. ROM is a set of automated computational techniques that helps you reuse your high-fidelity models for creating faster-running, lower-fidelity approximations.

In this talk, learn about different ROM techniques and methods, covering AI-based approaches, linear-parameter varying (LPV) modeling, and strategies for bringing large-scale sparse state-space matrices from FEA tools into MATLAB^® and Simulink^® for applications such as flexible body modeling and control. The focus of the talk, however, will be on AI-based ROM. See how you can perform a thorough design of experiments and use the resulting data to train AI models using LSTM, neural ODE, and nonlinear ARX algorithms. Learn how to integrate these AI models into your Simulink simulations, whether for hardware-in-the-loop testing or deployment to embedded systems for virtual sensor applications. Learn about the pros and cons of different ROM approaches to help you choose the best one for your next project.