Machine Learning

Physics-Embedded Neural ODEs for Sim2Real Edge Digital Twins of Hybrid Power Electronics Systems
PENODE enhances edge digital twins for power electronics by unifying event-based switching and physics-informed neural modeling, enabling efficient and interpretable real-time control on FPGA.
Jialin Zheng, Haoyu Wang, Yangbin Zeng, Di Mou, Xin Zhang, Hong Li, Sergio Vazquez, Leopoldo G. Franquelo
PDF Cite Code DOI
Physics-Embedded Neural ODEs for Sim2Real Edge Digital Twins of Hybrid Power Electronics Systems
Neural Substitute Solver for Efficient Edge Inference of Power Electronic Hybrid Dynamics
NSS enables fast, low-cost edge inference of hybrid PES dynamics by replacing traditional solvers with parallel-friendly neural networks, achieving 23× speedup and 60% resource savings.
Jialin Zheng, Haoyu Wang, Yangbin Zeng, Han Xu, Di Mou, Hong Li, Sergio Vazquez, Leopoldo G. Franquelo
PDF Cite DOI
Neural Substitute Solver for Efficient Edge Inference of Power Electronic Hybrid Dynamics