Closed-Loop Turbulence Control: Progress and Challenges

By Steven L. Brunton, Bernd R. Noack. 2015

627 citations recorded by the source corpus.

Closed-loop turbulence control is a critical enabler of aerodynamic drag reduction, lift increase, mixing enhancement, and noise reduction. Current and future applications have epic proportion: cars, trucks, trains, airplanes, wind turbines, medical devices, combustion, chemical reactors, just to name a few. Methods to adaptively adjust open-loop parameters are continually improving toward shorter response times. However, control design for in-time response is challenged by strong nonlinearity, high-dimensionality, and time-delays. Recent advances in the field of model identification and system reduction, coupled with advances in control theory (robust, adaptive, and nonlinear) are driving significant progress in adaptive and in-time closed-loop control of fluid turbulence. In this review, we provide an overview of critical theoretical developments, highlighted by compelling experimental success stories. We also point to challenging open problems and propose potentially disruptive technologies of machine learning and compressive sensing.

Subfields: Aerospace Engineering, Computational Mechanics, Statistical and Nonlinear Physics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computational Mechanics

In Statistical and Nonlinear Physics

Citations in the corpus

Discovering governing equations from data by sparse identification of nonlinear dynamical systems (2015, background): “erning equations, including neuroscience, climate science, epidemiology, and ﬁnancial markets. Fields that already use genetic programming, such as machine learning control for turbulent ﬂuid systems [5, 34], may also beneﬁt. Finally, normal forms may be discovered by including parameters in the opt”
Sparse Identification of Nonlinear Dynamics with Control (SINDYc) (2016, background): “D S] 2 1 M ay 2 01 cations (Brunton and Noack, 2015).”
Sparse identification of nonlinear dynamics for model predictive control in the low-data limit (2017, background): “The major drawback of model-based control, such as MPC, is in the development of a suitable model via existing system identification or model reduction [7], which may require expensive and timeconsuming data collection and computations.”
Machine Learning for Fluid Mechanics (2019, background): “However, simulations and experiments are expensive for iterative optimization, and simulations are often too slow for real-time control (Brunton & Noack 2015).”
Machine Learning Control – Taming Nonlinear Dynamics and Turbulence (2016, methodology): “Modified from Brunton and Noack, Applied Mechanics Reviews, 2015 [43].”
Recursive dynamic mode decomposition of transient and post-transient wake flows (2016, background): “In addition, a particularly attractive opportunity for RDMD is the unsupervised extraction of generalized mean-field models with few dominant frequencies (Brunton & Noack 2015).”
Modal Analysis of Fluid Flows: An Overview (2017, methodology): “Such models can be useful in closed-loop flow control [54,59,84,85,90,93,217] as well as aerodynamic design [218,219].”
Multi-scale proper orthogonal decomposition of complex fluid flows (2018, methodology): “…for adaptive masking or image pre-processing (Mendez et al. 2018a, 2017), flow sensing (Venturi & Karniadakis 2004; Willcox 2006), flow control (Brunton & Noack 2015; Bergmann & Cordier 2008), reduced order modeling (Deane et al. 1991), experimental validation of CFD data (Blanchard et al.…”
Data-Driven Sparse Sensor Placement for Reconstruction: Demonstrating the Benefits of Exploiting Known Patterns (2017, methodology): “There are myriad complex systems that would beneﬁt from principled, scaleable sensor and actuator placement, including ﬂuid ﬂow control [9], power grid optimization [10], epidemiological modeling and suppression [11], bio-regulatory network monitoring and control [12], and high-performance…”
Chaos as an intermittently forced linear system (2016, background): “The motion of planets [76], weather and climate [63, 65, 66, 38, 83, 67], population dynamics [9, 94, 102], epidemiology [93], financial markets, earthquakes, solar flares, and turbulence [53, 51, 52, 95, 12], all provide compelling examples of chaos.”
Data-driven discovery of Koopman eigenfunctions for control (2017, background): “A wide range of data-driven and nonlinear control approaches exist in the literature, including model-free adaptive control [67], extremum-seeking [6], gain scheduling [97], feedback linearization [27], describing functions [115], sliding mode control [37], singular perturbation [62], geometric cont”
Constrained sparse Galerkin regression (2016, background): “Over the past decades, identifying robust, accurate and efficient reduced-order models has thus become a central challenge in fluid dynamics and closed-loop flow control (Fabbiane et al. 2014; Brunton & Noack 2015; Sipp & Schmid 2016; Rowley & Dawson 2017).”
Artificial neural networks trained through deep reinforcement learning discover control strategies for active flow control (2018, methodology): “Drag reduction and flow control are techniques of critical interest for the industry (Brunton & Noack, 2015).”