Long-Time Trajectory Approximation via SA-NODEs: Model Predictive and Floquet Strategies

DOI10.48550/arXiv.2608.10738MaRDI QIDQ7284894arXivFDO

Authors Ziqian Li, Nikolaos M. Matzakos

Publication date 11 August 2026

Published in arXiv

Tldr We approximate long time trajectories using SA-NODEs with two strategies: model predictive and Floquet. The first partitions the horizon adaptively for adaptive resets, while the second uses a certified contraction of the learned return map to ensure linear growth in periods.

Full work available at URL https://doi.org/10.48550/arXiv.2608.10738

arXiv classification

cs.LGmath.NA

Cites work

• Activation Saturation and Floquet Spectrum Collapse in Neural ODEs
• Data-driven model identification using forcing-induced limit cycles
• Designing 2-D limit-cycle oscillators with prescribed trajectories and phase-response characteristics
• Differentiable Multiple Shooting Layers
• Differential Equations
• Error growth in the numerical integration of periodic orbits
• Error Growth in the Numerical Integration of Periodic Orbits, with Application to Hamiltonian and Reversible Systems
• Geometric Numerical Integration
• Learning Stable Deep Dynamics Models
• Model Predictive Control
• Neural ODE Control for Classification, Approximation, and Transport
• Neural Ordinary Differential Equations
• Norms and exclusion theorems
• Representation and regression problems in neural networks: relaxation, generalization, and numerics
• Sparsity in long-time control of neural ODEs
• Stability and convergence of a randomized model predictive control strategy
• Stability and Convergence of a Randomized Model Predictive Control Strategy
• Turnpike in optimal control of PDEs, ResNets, and beyond
• Universal approximation of dynamical systems by semiautonomous neural ODEs and applications
• Universal Approximation Theorems for Dynamical Systems with Infinite-Time Horizon Guarantees

This page was built for publication: Long-Time Trajectory Approximation via SA-NODEs: Model Predictive and Floquet Strategies

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7284894)