Avoidance of an unexpected obstacle without reinforcement learning: Why not using advanced control-theoretic tools?

2025/09/03 by Join, Cédric, Fliess, Michel · 1 citation
#FOS: Computer and information sciences #FOS: Electrical engineering #FOS: Mathematics #Optimization and Control (math.OC) #Robotics (cs.RO) #Systems and Control (eess.SY) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.2509.03721

Abstract

This communication on collision avoidance with unexpected obstacles is motivated by some critical appraisals on reinforcement learning (RL) which "requires ridiculously large numbers of trials to learn any new task" (Yann LeCun). We use the classic Dubins' car in order to replace RL with flatness-based control, combined with the HEOL feedback setting, and the latest model-free predictive control approach. The two approaches lead to convincing computer experiments where the results with the model-based one are only slightly better. They exhibit a satisfactory robustness with respect to randomly generated mismatches/disturbances, which become excellent in the model-free case. Those properties would have been perhaps difficult to obtain with today's popular machine learning techniques in AI. Finally, we should emphasize that our two methods require a low computational burden.

Citations

• Linear Quadratic Regulators: A New Look
• Guidance and Control of Unmanned Surface Vehicles via HEOL
• Model-Free Predictive Control: Introductory Algebraic Calculations, and a Comparison with HEOL and ANNs
• Synchronization of Kuramoto oscillators via HEOL, and a discussion on AI
• The Euler-Lagrange equation and optimal control: Preliminary results
• Flatness-based control revisited: The HEOL setting
• Deep Reinforcement Learning for Robotics: A Survey of Real-World Successes
• Reinforcement Learning of the Prediction Horizon in Model Predictive\n Control
• A Tour of Reinforcement Learning: The View from Continuous Control
• Reinforcement learning in robotics: A survey
• Model-free control

Cited by

• Linear Quadratic Regulators: A New Look

Related

• Model-Reference Reinforcement Learning for Collision-Free Tracking Control of Autonomous Surface Vehicles
• Guiding Reinforcement Learning with Incomplete System Dynamics
• Sampling-based Safe Reinforcement Learning for Nonlinear Dynamical Systems
• Disturbance Observer-based Control Barrier Functions with Residual Model Learning for Safe Reinforcement Learning
• Online Nonstochastic Model-Free Reinforcement Learning