AI-Based Real-Time Control of Power Electronic Circuits on Microcontrollers

#powerelectronics #artificial-intelligence #reinforcement-learning #microcontrollers #control

Title:
AI-Based Real-Time Control of Power Electronic Circuits on Microcontrollers

Abstract:
Artificial intelligence is increasingly entering the field of power electronics, but its use for real-time control remains challenging due to strict requirements on timing, robustness, and embedded implementation.

This webinar will introduce the motivation for AI-based controllers with a focus on reinforcement learning and compare their potential and limitations with established control approaches.

A practical workflow for developing and deploying reinforcement-learning-based controllers on resource-constrained microcontrollers will be presented, covering simulation-based training, embedded implementation, real-time constraints, and experimental validation.

Measurement results from power electronic circuits will be used to discuss the transition from simulation to hardware.

The webinar will conclude with open challenges and future directions, including robustness, safety, hybrid AI-classical control, and the role of AI at the edge for next-generation power electronic systems.

Speaker:
Prof. Christian Schumann
Kaiserslautern University of Applied Sciences, Germany