Efficient Flow Field Reconstruction Based on Ensemble Transform Kalman Filter

Accurate estimation of turbulence field is of great importance in aerospace, and existing means of obtaining turbulence field is inadequate in terms of resolution or accuracy. Experimental measurements are accurate but often have a limited number of observation points, and numerical computations can obtain full-field data but the accuracy is difficult to guarantee. The data assimilation method integrates experimental observation and numerical simulation, which is an effective tool for flow field reconstruction. This paper explored the effectiveness of data assimilation method based on ensemble transform Kalman filter(ETKF) in spatial flow field reconstruction, and also discussed the reconstruction accuracy and computational efficiency of different iterative updates, namely ETKF-M and ETKF-D, which update state variables based on turbulence model and flow field data respectively. Using the ONERA M6 airfoil as a numerical example, the algorithm was experimented by combining the pressure measurement data from 271 pressure holes on the airfoil surface from the wind tunnel test. The results show that different iterative updates of ETKF method can effectively modify the prediction of the turbulence model, and the ETKF-D improves the computational efficiency by 83% compared with the ETKF-M. In addition, two groups of 1/4 experimental observation points at different locations were selected for assimilation experiments, and results with different accuracies were obtained. It indicates that the accuracy of reconstruction is closely related to the location and number of assimilated observation points.