Opt-ODENet: Neural ODE Controller Design with Differentiable Optimization Layers for Safety and Stability

Keyan Miao, Liqun Zhao, Han Wang, Konstantinos Gatsis, Antonis Papachristodoulou
Proceedings of the 7th Annual Learning for Dynamics \& Control Conference, PMLR 283:1217-1229, 2025.

Abstract

Designing controllers that achieve task objectives while ensuring safety is a key challenge in control systems. This work introduces Opt-ODENet, a Neural ODE framework with a differentiable Quadratic Programming (QP) optimization layer to enforce constraints as hard requirements. Eliminating the reliance on nominal controllers or large datasets, our framework solves the optimal control problem directly using Neural ODEs. Stability and convergence are ensured through Control Lyapunov Functions (CLFs) in the loss function, while Control Barrier Functions (CBFs) embedded in the QP layer enforce real-time safety. By integrating the differentiable QP layer with Neural ODEs, we demonstrate compatibility with the adjoint method for gradient computation, enabling the learning of the CBF class-$\mathcal{K}$ function and control network parameters. Experiments validate its effectiveness in balancing safety and performance.

Cite this Paper

BibTeX
@InProceedings{pmlr-v283-miao25a, title = {Opt-ODENet: Neural ODE Controller Design with Differentiable Optimization Layers for Safety and Stability}, author = {Miao, Keyan and Zhao, Liqun and Wang, Han and Gatsis, Konstantinos and Papachristodoulou, Antonis}, booktitle = {Proceedings of the 7th Annual Learning for Dynamics \& Control Conference}, pages = {1217--1229}, year = {2025}, editor = {Ozay, Necmiye and Balzano, Laura and Panagou, Dimitra and Abate, Alessandro}, volume = {283}, series = {Proceedings of Machine Learning Research}, month = {04--06 Jun}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v283/main/assets/miao25a/miao25a.pdf}, url = {https://proceedings.mlr.press/v283/miao25a.html}, abstract = {Designing controllers that achieve task objectives while ensuring safety is a key challenge in control systems. This work introduces Opt-ODENet, a Neural ODE framework with a differentiable Quadratic Programming (QP) optimization layer to enforce constraints as hard requirements. Eliminating the reliance on nominal controllers or large datasets, our framework solves the optimal control problem directly using Neural ODEs. Stability and convergence are ensured through Control Lyapunov Functions (CLFs) in the loss function, while Control Barrier Functions (CBFs) embedded in the QP layer enforce real-time safety. By integrating the differentiable QP layer with Neural ODEs, we demonstrate compatibility with the adjoint method for gradient computation, enabling the learning of the CBF class-$\mathcal{K}$ function and control network parameters. Experiments validate its effectiveness in balancing safety and performance.} }
Endnote
%0 Conference Paper %T Opt-ODENet: Neural ODE Controller Design with Differentiable Optimization Layers for Safety and Stability %A Keyan Miao %A Liqun Zhao %A Han Wang %A Konstantinos Gatsis %A Antonis Papachristodoulou %B Proceedings of the 7th Annual Learning for Dynamics \& Control Conference %C Proceedings of Machine Learning Research %D 2025 %E Necmiye Ozay %E Laura Balzano %E Dimitra Panagou %E Alessandro Abate %F pmlr-v283-miao25a %I PMLR %P 1217--1229 %U https://proceedings.mlr.press/v283/miao25a.html %V 283 %X Designing controllers that achieve task objectives while ensuring safety is a key challenge in control systems. This work introduces Opt-ODENet, a Neural ODE framework with a differentiable Quadratic Programming (QP) optimization layer to enforce constraints as hard requirements. Eliminating the reliance on nominal controllers or large datasets, our framework solves the optimal control problem directly using Neural ODEs. Stability and convergence are ensured through Control Lyapunov Functions (CLFs) in the loss function, while Control Barrier Functions (CBFs) embedded in the QP layer enforce real-time safety. By integrating the differentiable QP layer with Neural ODEs, we demonstrate compatibility with the adjoint method for gradient computation, enabling the learning of the CBF class-$\mathcal{K}$ function and control network parameters. Experiments validate its effectiveness in balancing safety and performance.
APA
Miao, K., Zhao, L., Wang, H., Gatsis, K. & Papachristodoulou, A.. (2025). Opt-ODENet: Neural ODE Controller Design with Differentiable Optimization Layers for Safety and Stability. Proceedings of the 7th Annual Learning for Dynamics \& Control Conference, in Proceedings of Machine Learning Research 283:1217-1229 Available from https://proceedings.mlr.press/v283/miao25a.html.

Related Material