17 June 2020 in Journal of Fluid Mechanics
An artificial intelligence (AI) control system is developed to maximize the mixing rate of a turbulent jet. This system comprises of six independently operated unsteady minijet actuators, two hot-wire sensors placed in the jet and genetic programming for the unsupervised learning of a near-optimal control law. The ansatz of this law includes multi-frequency open-loop forcing, sensor feedback and nonlinear combinations thereof. Mixing performance is quantified by the decay rate of the centreline mean velocity of the jet. Intriguingly, the learning process of AI control discovers the classical forcings, i.e. axisymmetric, helical and flapping achievable from conventional control techniques, one by one in the order of increased performance, and finally converges to a hitherto unexplored forcing. Careful examination of the control landscape unveils typical control laws, generated in the learning process, and their evolutions. The best AI forcing produces a complex turbulent flow structure that is characterized by periodically generated mushroom structures, helical motion and an oscillating jet column, all enhancing the mixing rate and vastly outperforming others. Being never reported before, this flow structure is examined in various aspects, including the velocity spectra, mean and fluctuating velocity fields and their downstream evolution, and flow visualization images in three orthogonal planes, all compared with other classical flow structures. Along with the knowledge of the minijet-produced flow and its effect on the initial condition of the main jet, these aspects cast valuable insight into the physics behind the highly effective mixing of this newly found flow structure. The results point to the great potential of AI in conquering the vast opportunity space of control laws for many actuators and sensors and in optimizing turbulence.
31 July 2018 in Experiments in Fluids
1 December 2020 in Physics of Fluids
A machine learning control (MLC) is proposed based on the explorative gradient method (EGM) for the optimization and sensitivity analysis of actuation parameters. This technique is applied to reduce the drag of a square-back Ahmed body at a Reynolds number Re = 1.7 × 105. The MLC system consists of pulsed blowing along the periphery of the base, 25 pressure taps distributed on the vertical base of the body, and an EGM controller for unsupervised searching for the best control law. The parameter search space contains the excitation frequency fe, duty cycle α, and flow rate blowing coefficient Cm. It is demonstrated that the MLC may cut short the searching process significantly, requiring only about 100 test runs and achieving 13% base pressure recovery with a drag reduction of 11%. Extensive flow measurements are performed with and without control to understand the underlying flow physics. The converged control law achieves fluidic boat tailing and, meanwhile, eliminates the wake bistability. Such simultaneous achievements have never been reported before. A machine-learned response model is proposed to link the control parameters with the cost function. A sensitivity analysis based on this model unveils that the control performance is sensitive to fe and α but less so to Cm. The result suggests that a small sacrifice on performance will give a huge return on actuation power saving, which may provide important guidance on future drag reduction studies as well as engineering applications.
12 May 2023 in Journal of Fluid Mechanics
This work proposes a machine-learning or artificial intelligence (AI) control of a low-drag Ahmed body with a rear slant angleφ = 35° with a view to finding strategies for efficient drag reduction (DR). The Reynolds numberReinvestigated is 1.7 × 105based on the square root of the body cross-sectional area. The control system comprises of five independently operated arrays of steady microjets blowing along the edges of the rear window and vertical base, twenty-six pressure taps on the rear end of the body and a controller based on an ant colony algorithm for unsupervised learning of a near-optimal control law. The cost function is designed such that bothDRand control power input are considered. The learning process of the AI control discovers forcing that produces aDRup to 18 %, corresponding to a drag coefficient reduction of 0.06, greatly exceeding any previously reportedDRfor this body. Furthermore, the discovered forcings may provide alternative solutions, i.e. a tremendously increased control efficiency given a small sacrifice inDR. Extensive flow measurements performed with and without control indicate significant alterations in the flow structure around the body, such as flow separation over the rear window, recirculation bubbles and C-pillar vortices, which are linked to the pressure rise on the window and base. The physical mechanism forDRis unveiled, along with a conceptual model for the altered flow structure under the optimum control or biggestDR. This mechanism is further compared with that under the highest control efficiency.
1 November 2017 in AIAA Journal
A modified extremum-seeking closed-loop control system is proposed for jet mixing enhancement. The system introduces a variable perturbation component that may act to adjust adaptively the searching range of the optimal control signal. The actuator is a single pulsed minijet injected radially into the main jet before its issue. The excitation frequency for the minjet is controlled by an electromagnetic valve. The open-loop control measurements indicate that, given the duty cycle and the mass flow rate ratio of the minijet to the main jet, the decay rate of jet centerline mean velocity, which provides a measure for jet mixing, displays a maximum at the optimum frequency ratio , where is the preferred-mode frequency of the main jet. Comparison is made between the modified and classical systems. Although both may find automatically the optimal and hence a maximum , the modified system achieves a contraction in the fluctuating , when the control target is reached, by 70%, and subsequently an increase in by 12%. The new system further exhibits a better robustness than the classical.
1 January 2024 in Physics of Fluids
An experimental investigation is conducted on the active drag reduction (DR) of an Ahmed body with a rear slant angle of 35°, corresponding to the low-drag regime, using single and combined actuations at the Reynolds number Re = 1.7 × 105. Five different actuations, produced by steady blowing jets, are deployed independently around the edges of the rear slant surface and vertical base, achieving the maximum DR of 1%–9%. An artificial intelligence control system based on ant colony algorithm is used for finding near-optimal control laws of the combined jets. With both DR and control power input considered in the cost function, the maximum DR obtained reaches 18%, though the corresponding control efficiency η (≡ES/EI, where ES and EI are the saved power due to DR and the total input power of the actuations, respectively) is only 0.13. However, η may go up greatly, climbing to 5.8, given a 3% sacrifice of DR. Extensive flow measurements are conducted, with and without control, to understand the flow physics and mechanisms under the control of individual and combined actuations. A linear regression model is established to describe the correlation between the control efficiency and parameters under the combined actuations.