Year: 2025
Type: article
Abstract: Model Predictive Control has emerged as a popular tool for robots to generate complex motions. However, the real-time requirement has limited the use of hard constraints and large preview horizons, which are necessary to ensure safety and stability. In practice, practitioners have to carefully design cost functions that can imitate an infinite horizon formulation, which is tedious and often results in local minima. In this work, we study how to approximate the infinite horizon value function of constrained optimal control problems with neural networks using value iteration and trajectory optimization. Furthermore, we experimentally demonstrate how using this value function approximation as a terminal cost provides global stability to the model predictive controller. The approach is validated on two toy problems and a real-world scenario with online obstacle avoidance on an industrial manipulator where the value function is conditioned to the goal and obstacle.
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
Awards: G1190173965, Artificial and Natural Intelligence Toulouse Institute, Next generation of AI-powered robotics for agile production, PR[AI]RIE-PSAI - Paris School of Artificial Intelligence, CPS: Medium: Accurate and Efficient Collective Additive Manufacturing by Mobile Robots, CISE-ANR: RI: Small: Numerically efficient reinforcement learning for constrained systems with super-linear convergence (NERL), Learning and Control of Sensorimotor Models in Robotics, PaRis Artificial Intelligence Research InstitutE, FW-HTF-R: US-Ireland R&D Partnership: ARISE: Assembly and Robotics Innovation in Steel Building Erection, W-HTF-RL: Collaborative Research: Improving the Future of Retail and Warehouse Workers with Upper Limb Disabilities via Perceptive and Adaptive Soft Wearable Robots, G7073499455
OpenAlex ID: W4411086224