Year: 2025
Type: preprint
Abstract: This paper presents a transferable solution method for optimal control problems with varying objectives using function encoder (FE) policies. Traditional optimization-based approaches must be re-solved whenever objectives change, resulting in prohibitive computational costs for applications requiring frequent evaluation and adaptation. The proposed method learns a reusable set of neural basis functions that spans the control policy space, enabling efficient zero-shot adaptation to new tasks through either projection from data or direct mapping from problem specifications. The key idea is an offline-online decomposition: basis functions are learned once during offline imitation learning, while online adaptation requires only lightweight coefficient estimation. Numerical experiments across diverse dynamics, dimensions, and cost structures show our method delivers near-optimal performance with minimal overhead when generalizing across tasks, enabling semi-global feedback policies suitable for real-time deployment.
Source: arXiv (Cornell University)
Authors: Li, Xingjian, Kan, Kelvin, Verma, Deepanshu, Kumar, Krishna, Osher, Stanley, Drgoňa, Ján
Institution: The University of Texas at Austin
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: SCIPE: Chishiki.ai: A sustainable, diverse, and integrated CIP community for Artificial Intelligence in Civil and Environmental Engineering, CAREER: HayaRupu: Accelerating Natural Hazard Engineering with AI-Driven Discovery Loops, Collaborative Research: Algorithms, Theory, and Validation of Deep Graph Learning with Limited Supervision: A Continuous Perspective
OpenAlex ID: W7083169234
Cites:
Cited by:
Related to: