TY - JOUR
T1 - Input-to-State Stable Neural Ordinary Differential Equations with Applications to Transient Modeling of Circuits
AU - Yang, Alan
AU - Xiong, Jie
AU - Raginsky, Maxim
AU - Rosenbaum, Elyse
N1 - Funding Information:
This work was funded in part by the NSF under CNS 16-24811 and the industry members of the CAEML I/UCRC, and in part by the Illinois Institute for Data Science and Dynamical Systems (iDS2), an NSF HDR TRIPODS institute, under award CCF-1934986.
Publisher Copyright:
© 2022 A. Yang, J. Xiong, M. Raginsky & E. Rosenbaum.
PY - 2022
Y1 - 2022
N2 - This paper proposes a class of neural ordinary differential equations parametrized by provably input-to-state stable continuous-time recurrent neural networks. The model dynamics are defined by construction to be input-to-state stable (ISS) with respect to an ISS-Lyapunov function that is learned jointly with the dynamics. We use the proposed method to learn cheap-to-simulate behavioral models for electronic circuits that can accurately reproduce the behavior of various digital and analog circuits when simulated by a commercial circuit simulator, even when interconnected with circuit components not encountered during training. We also demonstrate the feasibility of learning ISS-preserving perturbations to the dynamics for modeling degradation effects due to circuit aging.
AB - This paper proposes a class of neural ordinary differential equations parametrized by provably input-to-state stable continuous-time recurrent neural networks. The model dynamics are defined by construction to be input-to-state stable (ISS) with respect to an ISS-Lyapunov function that is learned jointly with the dynamics. We use the proposed method to learn cheap-to-simulate behavioral models for electronic circuits that can accurately reproduce the behavior of various digital and analog circuits when simulated by a commercial circuit simulator, even when interconnected with circuit components not encountered during training. We also demonstrate the feasibility of learning ISS-preserving perturbations to the dynamics for modeling degradation effects due to circuit aging.
KW - Circuit simulation
KW - Input to State Stability
KW - Neural ODE
KW - Safe learning
UR - http://www.scopus.com/inward/record.url?scp=85141500172&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85141500172&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:85141500172
SN - 2640-3498
VL - 168
SP - 663
EP - 675
JO - Proceedings of Machine Learning Research
JF - Proceedings of Machine Learning Research
T2 - 4th Annual Learning for Dynamics and Control Conference, L4DC 2022
Y2 - 23 June 2022 through 24 June 2022
ER -