Deep Reinforcement Learning in Action: Real-Time Control of Vortex-Induced Vibrations

2025-09-29

AI Summary

This study addresses the challenge of real-time active flow control for vortex-induced vibration (VIV) of a circular cylinder at high Reynolds number (Re = 3000). For the first time, deep reinforcement learning (DRL) is applied under realistic experimental constraints—including actuator delay—to suppress vibration via rotary actuation. To overcome delay-induced limitations, a state representation incorporating historical control actions is introduced. A policy network is designed to jointly consider displacement, velocity, and past inputs, enabling autonomous learning of a synergistic low- and high-frequency control law. Experimental results demonstrate >95% vibration attenuation—substantially outperforming a baseline low-frequency strategy relying solely on instantaneous state information. The method exhibits strong robustness, adaptability, and engineering feasibility in complex physical environments.

Technology Category

• Machine Learning: Reinforcement Learning
• Intelligent Robots: Behavior Learning & Control
• Computer Vision: Low Level & Physics-based Vision

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

This study showcases an experimental deployment of deep reinforcement learning (DRL) for active flow control (AFC) of vortex-induced vibrations (VIV) in a circular cylinder at a high Reynolds number (Re = 3000) using rotary actuation. Departing from prior work that relied on low-Reynolds-number numerical simulations, this research demonstrates real-time control in a challenging experimental setting, successfully addressing practical constraints such as actuator delay. When the learning algorithm is provided with state feedback alone (displacement and velocity of the oscillating cylinder), the DRL agent learns a low-frequency rotary control strategy that achieves up to 80% vibration suppression which leverages the traditional lock-on phenomenon. While this level of suppression is significant, it remains below the performance achieved using high-frequency rotary actuation. The reduction in performance is attributed to actuation delays and can be mitigated by augmenting the learning algorithm with past control actions. This enables the agent to learn a high-frequency rotary control strategy that effectively modifies vortex shedding and achieves over 95% vibration attenuation. These results demonstrate the adaptability of DRL for AFC in real-world experiments and its ability to overcome instrumental limitations such as actuation lag.

Problem

• Real-time active flow control of vortex-induced vibrations at high Reynolds numbers
• Overcoming actuator delay limitations in experimental deep reinforcement learning
• Achieving over 95% vibration suppression through modified vortex shedding control

Innovation

• Deep reinforcement learning controls vortex-induced vibrations experimentally
• Low-frequency strategy achieves 80% vibration suppression via lock-on
• High-frequency control with past actions reaches 95% attenuation