This paper presents a novel adaptive control pattern (ACP) for real-time feedback control of flow separation over airfoils using sparse processing particle image velocimetry (SPPIV) and dielectric barrier discharge plasma actuators. The study addresses the challenge of suppressing separation in deep stall at low Reynolds numbers, where previous feedback-based methods often fail to maintain effective flow attachment. Unlike conventional feedback control methods such as single-step and multiple-step prediction, where control inputs are directly decided on the basis of state predictions, ACP determines the modulation frequency of the actuation. This is done according to threshold-based flow state detection, enabling the selection of effective actuation patterns for the estimated flow features. Experiments were conducted on a NACA0015 airfoil at an angle of attack of 18 degrees and a Reynolds number of \ (\) 66, 000 using a plasma actuator positioned at the leading edge. The SPPIV system acquired data with a sampling frequency of 2, 000 Hz, processed PIV at optimized limited interrogation windows, and estimated the state within 20% of the time between samples. Linear models for state estimation were generated via dynamic mode decomposition with control of the flow field representations from proper orthogonal decomposition. Results demonstrate that ACP control successfully achieves flow reattachment at 6 kV actuation voltage where other methods fail, whereas at higher voltages, ACP control combines the fast and reliable reattachment speeds of lower actuation burst frequencies of open-loop control with stable quasisteady attached conditions of higher burst frequencies. It shows that the optimal actuation frequency for driving the reattachment process is not the same as that for maintaining a reattached condition, and that guiding the actuation frequency as the reattachment process evolves can provide substantial improvements in control authority. This breakthrough brings the plasma actuator one step closer to practical control applications, with highly effective yet robust control parameters.
Viguera et al. (Sat,) studied this question.