In this study, the drag minimization problem around a circular cylinder was considered. Two plasma actuators (PAs) were installed on the upper and the lower side of the cylinder. Thus, the design problem is the multi-variate problem. A Kriging based genetic algorithm (GA) was employed to optimize the parameters of the operating conditions of PAs and the design knowledge discovery is also carried out for the multi-variate problem. The aerodynamic performance was evaluated in wind tunnel testing to overcome the disadvantages of time-consuming numerical simulations. This optimization methodology explores the optimum waveform of parameters for AC voltage by changing the waveform automatically. Based on these results, optimum designs and global design information were obtained while drastically reducing the number of experiments required compared to a full factorial experiment. An analysis of variance (ANOVA) and a scatter plot matrix (SPM) were introduced for design knowledge discovery. According to the discovered design knowledge, it was found that the modulation frequency for two PAs is an important parameter to reduce drag. In addition, the duty ratio for a PA on the model has two optimum points. Flow visualization is also carried out by particle image velocimetry (PIV). According to this result, the flow separation around the model is reduced by optimum design compared with the flow of the model without PAs.
展开▼
