Deep reinforcement learning for turbulent drag reduction in channel flows

TL;DR

This paper develops a reinforcement learning environment for turbulent drag reduction in channel flows, enabling testing of advanced control strategies and demonstrating DRL's superior performance over traditional methods.

Contribution

It introduces a high-fidelity, parallelized RL environment for turbulent flow control and benchmarks DRL against classical opposition control, showing significant improvements.

Findings

DRL achieves 43-46% drag reduction in channel flows.

The environment facilitates testing and benchmarking of flow control strategies.

DRL outperforms opposition control by approximately 20 percentage points.

Abstract

We introduce a reinforcement learning (RL) environment to design and benchmark control strategies aimed at reducing drag in turbulent fluid flows enclosed in a channel. The environment provides a framework for computationally-efficient, parallelized, high-fidelity fluid simulations, ready to interface with established RL agent programming interfaces. This allows for both testing existing deep reinforcement learning (DRL) algorithms against a challenging task, and advancing our knowledge of a complex, turbulent physical system that has been a major topic of research for over two centuries, and remains, even today, the subject of many unanswered questions. The control is applied in the form of blowing and suction at the wall, while the observable state is configurable, allowing to choose different variables such as velocity and pressure, in different locations of the domain. Given the complex nonlinear nature of turbulent flows, the control strategies proposed so far in the literature are physically grounded, but too simple. DRL, by contrast, enables leveraging the high-dimensional data that can be sampled from flow simulations to design advanced control strategies. In an effort to establish a benchmark for testing data-driven control strategies, we compare opposition control, a state-of-the-art turbulence-control strategy from the literature, and a commonly-used DRL algorithm, deep deterministic policy gradient. Our results show that DRL leads to 43% and 46% drag reduction in a minimal and a larger channel (at a friction Reynolds number of 180), respectively, outperforming the classical opposition control by around 20 percentage points.