Offline Reinforcement Learning for Fluid Controls: Data-based Multi-observational Policy Extraction
For engineers in fluid dynamics, this work addresses the computational cost of online RL and sensor-specific retraining, but the results are demonstrated on simplified problems without concrete performance numbers.
This paper proposes an offline RL framework for active flow control that uses a sensor position-conditioned architecture to enable a single policy to adapt to multiple sensor arrangements, eliminating the need for retraining. The method is demonstrated on two problems, achieving flexible policy extraction from offline data.
Active flow control is a fundamental application in engineering. Recent advances in deep reinforcement learning have made progress in this field. However, the classical online RL approaches require extensive real-time interactions with the high fidelity environment, while each sensor configuration change necessitates whole policy retraining. All these factors result in prohibitive computational costs for real-world applications. In this work, we propose a novel offline RL framework that addresses both challenges through data-driven policy extraction. We develop a sensor position-conditioned architecture that enables a single policy network to adapt seamlessly to multiple sensor arrangements. The position-conditioned approach incorporated spatial relationship modeling through Point Attention layers to ensure the generalizability to varying sensor placements. We demonstrate the framework on two representative problems, mitigating chaoticity in the Kuramoto-Sivashinsky equation and flow control over airfoils governed by the Navier-Stokes equation. The result demonstrates that the policy extraction from the dataset provides unprecedented flexibility for sensor placement optimization. This approach represents a significant step towards adaptive, intelligent flow control systems.