We present the first closed-loop separation control experiment using a novel, model-free strategy based on genetic programming, which we call 'machine learning control'. The goal is to reduce the recirculation zone of backward-facing step flow at Re h = 1350 manipulated by a slotted jet and optically sensed by online particle image velocimetry. The feedback control law is optimized with respect to a cost functional based on the recirculation area and a penalization of the actuation. This optimization is performed employing genetic programming. After 12 generations comprised of 500 individuals, the algorithm converges to a feedback law which reduces the recirculation zone by 80 %. This machine learning control is benchmarked against the best periodic forcing which excites Kelvin-Helmholtz vortices. The machine learning control yields a new actuation mechanism resonating with the low-frequency flapping mode instability. This feedback control performs similarly to periodic forcing at the design condition but outperforms periodic forcing when the Reynolds number is varied by a factor two. The current study indicates that machine learning control can effectively explore and optimize new feedback actuation mechanisms in numerous experimental applications.
The separated flow downstream of a backwardfacing step is controlled using visual information for feedback. This is done when looking at the flow from two vantage points. Flow velocity fields are computed in real time and used to yield inputs to a control loop. This approach to flow control is shown to be able to control the detached flow in the same way as has been done before by using the area of the recirculation region downstream of the step as the input for a gradient descent optimization scheme. Visual feedback using real-time computations of two-dimensional velocity fields also allows for novel inputs in the feedback scheme. As a proof of concept, the spatially averaged value of the swirling strength λ Ci is successfully used as the input for an automatically tuned proportional-integral-derivative controller.
Closed-loop control of an amplifier flow is experimentally investigated. A feed-forward algorithm is implemented to control the flow downstream of a backward-facing step (BFS) perturbed by upstream perturbations. Upstream and downstream data are extracted from real-time velocity fields to compute an ARMAX model used to effect actuation. This work, done at Reynolds number 430, investigates the practical feasibility of this approach which has shown great promise in a recent numerical study by Hervé et al. (J. Fluid Mech., vol. 702, 2012, pp. 26–58). The linear nature of the regime is checked, two-dimensional upstream perturbations are introduced, and the degree to which the flow can be controlled is quantified. The resulting actuation is able to effectively reduce downstream energy levels and fluctuations. The limitations and difficulties of applying such an approach to an experiment are also discussed.
We propose a novel closed-loop control strategy of turbulent flows using machine learning methods in a model-free manner. This strategy, called Machine Learning Control (MLC), allows -for the first time -to detect and exploit all enabling nonlinear actuation mechanisms in an un-supervised automatic manner. In this communication, we focus on MLC applications for in-time control of experimental shear flows and demonstrate how it outperforms state-of-the-art control. In particular, MLC is applied to three different experimental closed-loop control setups: (1) the TUCOROM mixing layer tunnel, (2) the Görtler PMMH water tunnel with a backward facing step, and (3) the LML Boundary Layer wind tunnel with a separating turbulent boundary layer. In all three cases, MLC finds a control which yields a significantly better performance with respect to the given cost functional as compared to the best previously tested open-loop actuation. We foresee numerous potential applications to most nonlinear multiple-input multiple-output (MIMO) flow control problems, particularly in experiments. In particular, the model-free architecture of MLC enables its application to a large class of complex nonlinear systems in all areas of science. *
Circular flush Jets In Cross-Flow were experimentally studied in a water tunnel using Volumetric Particle Tracking Velocimetry, for a range of jet to cross-flow velocity ratios, r, from 0.5 to 3, jet exit diameters d from 0.8 cm to 1 cm and cross-flow boundary layer thickness δ from 1 to 2.5 cm. The analysis of the 3D mean velocity fields allows for the definition, computation and study of Counter-rotating Vortex Pair trajectories. The influences of r, d and δ were investigated. A new scaling based on momentum ratio r m taking into account jet and cross-flow momentum distributions is introduced based on the analysis of jet trajectories published in the literature. Using a rigorous scaling quality factor Q to quantify how well a given scaling successfully collapses trajectories, we show that the proposed scaling also improves the collapse of CVP trajectories, leading to a final scaling law for these trajectories.
This paper presents a high-speed implementation of an optical flow algorithm which computes in real-time planar velocity fields in an experimental flow. Real-time computations of the flow velocity field allow the experimentalist to have instantaneous access to quantitative features of the flow. This can be very useful in many situations: fast evaluation of the performances and characteristics of a new setup, design optimization, easier and faster parametric studies, etc. It can also be used as a visual sensor for an input in closed-loop flow control experiments where fast estimation of the state of the flow is needed. The algorithm is implemented on a graphics processor unit. The accuracy of the computation is demonstrated. Computation speed and scalability of the processing are highlighted along with guidelines for further improvements. The system architecture is flexible, scalable and can be adapted on the fly in order to process higher resolutions or achieve higher precision. The setup is applied on a backward-facing step flow in a hydrodynamic channel. For validation purposes, classical particle image velocimetry (PIV) is used to compare with instantaneous optical flow measurements. The important flow characteristics like the dynamics of the recirculation bubble, computed in real time for the first time, are well recovered. The accuracy of real-time optical flow measurements is comparable to off-line PIV computations.
In this study, a simple model based closed-loop algorithm is used to control the separated flow downstream a backward-facing step. It has been shown in previous studies that the recirculation bubble can be minimized when exciting the shear layer at its natural Kelvin-Helmholtz instability frequency. In this experiment, the natural shedding frequency is identified through real-time analysis of 2D velocity fields. Actuation (pulsed jet) is then locked on this frequency. If flow characteristics stray too far from a set point, shedding frequency is updated and actuation changed. The present work demonstrates the efficacy and robustness of this approach in reducing recirculation while Reynolds number is randomly varied between 1400 and 2800. arXiv:1311.0696v1 [physics.flu-dyn]
The flow downstream a backward-facing step is controlled using a pulsed jet placed upstream of the step edge. Experimental velocity fields are computed and used to the recirculation area quantify. The effects of jet amplitude, frequency and duty cycle on this recirculation area are investigated for two Reynolds numbers (Re h = 2070 and Re h = 2900). The results of this experimental study demonstrate that upstream actuation can be as efficient as actuation at the step edge when exciting the shear layer at its natural frequency. Moreover it is shown that it is possible to minimize both jet amplitude and duty cycle and still achieve optimal efficiency. With minimal amplitude and a duty-cycle as low as 10% the recirculation area is nearly canceled.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
334 Leonard St
Brooklyn, NY 11211
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.