Overview

Optimal control asks how to steer a system over time toward goals while minimizing costs such as energy, time, or tracking error and respecting limits on actions and states. Classical approaches include Pontryagin’s maximum principle, dynamic programming and Hamilton–Jacobi–Bellman equations, and numerical methods such as direct collocation, differential dynamic programming, and model predictive control. Recent work also uses reinforcement learning, differentiable simulators, and neural surrogates to speed up planning or handle complex models, with applications ranging from robotics and aerospace to energy systems and PDEs. A central challenge is making these methods fast enough for real-time, high-dimensional tasks without losing reliability when models are uncertain or safety guarantees are needed.

Latest papers

CardsList

154 papers12 papers in the last four weeks, up 20% on the four weeks before. 0.1% of all new papers.

  1. Safe Learning Control with Optimality and Stability Guarantees

    Jan 26, 2025Xinyang Wang, Hongwei Zhang, Shimin Wang +2Control Barrier FunctionsRobust Control

  2. Indirect Optimal Control for Constructing Libraries of Periodic Trajectories: Application to Bipedal Walking

    Oct 12, 2024Maximilian Raff, Kathrin Flaßkamp, C. David RemyBipedal LocomotionTrajectory Optimization

  3. An Efficient Parallel Solver for Linear-Equality-Constrained LQR Problems

    Jul 7, 2024João Sousa-Pinto, Dominique OrbanLinear Quadratic RegulatorOptimal Control

  4. Entropic perspective on AI resilience, data-minimal learning and optimal control

    Oct 19, 2023Illia HorenkoData-Efficient LearningAdversarial Robustness