Experimental Data-Driven Flow Field Prediction for Compressor Cascade based on Deep Learning and l_(1)Regularization

作     者：LIU Tantao GAO Limin LI Ruiyu 

作者机构：School of Power and EnergyNorthwestern Polytechnical UniversityXi'an 710012China School of Aerospace EngineeringXi'an Jiaotong UniversityXi'an 710049China 

出 版 物：《Journal of Thermal Science》 (热科学学报（英文版）)

年 卷 期：2024年第33卷第5期

页      面：1867-1882页

核心收录：

学科分类：080704[工学-流体机械及工程] 080103[工学-流体力学] 08[工学] 0807[工学-动力工程及工程热物理] 0802[工学-机械工程] 0801[工学-力学（可授工学、理学学位）] 

基　　金：the support of the National Natural Science Foundation of China(No.52106053 No.92152301) 

主　　题：experimental data-driven compressor cascade deep learning l_(1)regularization 

摘      要：For complex flows in compressors containing flow separations and adverse pressure gradients,the numerical simulation results based on Reynolds-averaged Navier-Stokes(RANS)models often deviate from experimental measurements more or *** improve the prediction accuracy and reduce the difference between the RANS prediction results and experimental measurements,an experimental data-driven flow field prediction method based on deep learning and l_(1)regularization is proposed and applied to a compressor cascade flow *** inlet boundary conditions and turbulence model parameters are calibrated to obtain the high-fidelity flow *** Saplart-Allmaras and SST turbulence models are used independently for mutual *** contributions of key modified parameters are also analyzed via sensitivity *** results show that the prediction error can be reduced by nearly 70%based on the proposed *** flow fields predicted by the two calibrated turbulence models are almost the same and nearly independent of the turbulence *** corrections of the inlet boundary conditions reduce the error in the first half of the *** turbulence model calibrations fix the overprediction of flow separation on the suction surface near the tail edge.