A novel windowed multi-resolution Dynamic Mode Decomposition (wmrDMD) technique was developed to obtain a data-driven model from experimental measurements and allow for an increase in temporal resolution of the measured data. The novel data-driven analysis technique was applied to two-dimensional motion-capturing Pressure Sensitive Paint (PSP) data of the pressure field over a wing in a flutter. The new data-driven analysis technique was validated by reconstructing the pressure information from down-sampled time series of surface pressure data. The reconstruction was compared with the original pressure data and was in excellent agreement, with error values less than 0.1% for all data points. In addition, the data-driven model reconstruction allows for an increase in temporal resolution to double that of experimental measurements. The multi-resolution analysis differentiated between wing motion, shock wave location oscillation, and shock strength amplitude fluctuations. The proposed wmrDMD is a general technique applicable to a wide range of problems that include unsteady fields and moving domains.

开发了一种新颖的窗口多分辨率动态模式分解 (wmrDMD) 技术，以从实验测量中获得数据驱动的模型，并允许提高测量数据的时间分辨率。新颖的数据驱动分析技术被应用于扑翼中机翼上压力场的二维运动捕捉压敏涂料 (PSP) 数据。通过从地面压力数据的下采样时间序列重建压力信息，验证了新的数据驱动分析技术。将重建结果与原始压力数据进行比较，结果非常吻合，所有数据点的误差值均小于 0.1%。此外，数据驱动的模型重建允许将时间分辨率提高到实验测量的两倍。多分辨率分析区分了机翼运动、冲击波位置振荡和冲击强度幅度波动。所提出的 wmrDMD 是一种通用技术，适用于包括不稳定场和移动域在内的广泛问题。