Parameter Interval Optimization Of The Dbd Plasma Actuator Based On Orthogonal Experiment And Rbf Neural Network Approximation Model

To further improve the performance of the dielectric barrier discharge-plasma actuator (DBD-PA) and to ensure the convenience of excitation intensity adjustment, the parameters of the DBD-PA were subjected to interval optimization on the basis of an orthogonal experiment and the radial basis function (RBF) neural network approximation model. The parameters of the DBD-PA included electrode gap d1, exposed electrode width d2, covered electrode width d3, frequency f, and voltage peak-to-peak value Vpp. The maximum velocity U-max induced by DBD-PA was taken as the target variable. Orthogonal analysis results showed that the influence of Vpp on U-max was highly significant, whereas d1 had some influence and the other three parameters' influence was not significant. On the basis of the orthogonal experiment results, an RBF neural network approximate model was established. Through two groups of randomized experiments, the prediction error of the approximate model is verified to be within 3%. The interval optimization algorithm was used to optimize the parameters of the DBD-PA with Vpp as the uncertain variable. The optimal parameter combination of deterministic variables obtained by optimization is d1=0mm, d2=13mm, d3=20mm, and f=8.6kHz. Under different Vpp, the performance of the DBD-PA greatly improved in the optimal parameter combination, and the average increase in U-max was about 0.52 m/s.