基于LSTM-POD的汽车湍流尾迹的高时间分辨速度场重构  

A High Time-Resolution Reconstruction on the Automotive Turbulent Wake Based on LSTM-POD

作　　者：杨志刚[1,2,3] 李俣静 夏超 王梦佳 余磊[1,2] Yang Zhigang;Li Yujing;Xia Chao;Wang Mengjia;Yu Lei(School of Automotive Studies,Tongji University,Shanghai 201804;Shanghai Automotive Wind Tunnel Center,Tongji University,Shanghai 201804;Beijing Aeronautical Science&Technology Research Institute,Beijing 102211)

机构地区：[1]同济大学汽车学院,上海201804 [2]同济大学上海地面交通工具风洞中心,上海201804 [3]北京民用飞机技术研究中心,北京102211

出　　处：《汽车工程》2024年第7期1302-1313,共12页Automotive Engineering

基　　金：国家自然科学基金(52372360);国家重点研发计划项目(2022YFE0208000);上海市地面交通工具空气动力与热环境模拟重点实验室(23DZ2229029);中央高校基本科研业务费专项资金资助。

摘　　要：本文针对方背Ahmed汽车标模的湍流尾迹,建立基于长短时记忆法(long short-term memory,LSTM)和本征正交分解(proper orthogonal decomposition, POD)相结合的深度学习模型LSTM-POD。通过建立非时间分辨平面速度场POD模态系数和若干离散点的时间分辨速度信号的映射关系,实现了方背Ahmed汽车标模湍流尾迹流场的高时间分辨率重构,并对比了不同时间步长配置,即单时间步长(LSTM-Sin)和多时间步长(LSTM-Mul)对重构效果的影响。研究表明:LSTM-POD模型在时间序列重构中具有较强的学习和泛化能力。另外,LSTM-Mul考虑到了时间上的连续性和相关性,相较于LSTM-Sin,其重构出的低阶模态系数和速度场与POD的重构结果更吻合。本研究提出的深度学习模型可以缓解通过实验及高精度数值模拟获取高时间分辨率流场数据资源消耗大、计算效率低等问题。A deep-learning LSTM-based POD model(LSTM-POD)based on long short-term memory(LSTM)and proper orthogonal decomposition(POD)is developed for the turbulent wake of the square-back Ahmed automotive general model.A high time-resolution reconstruction is achieved by establishing the mapping relation-ship between the POD modal coefficients of the non-time-resolved planar velocity field and the time-resolved veloci-ty signals at a number of discrete points,and the effect of different time-step configurations,i.e.,the single time step(LSTM-Sin)and multiple time steps(LSTM-Mul)on the reconstruction results is compared.The results show that the LSTM-POD model has strong learning and generalization ability in time series reconstruction,In addition,LSTM-Mul considers temporal continuity and correlation,the reconstructed mode coefficients(lower order)and ve-locity fields of which are more consistent with the POD reconstructed results compared with that of LSTM-Sin.The deep learning model proposed in this study can alleviate the problems of high resource consumption and low compu-tational efficiency in obtaining high time resolution flow field data through experiments and high-precision numerical simulation.

关 键 词：汽车湍流尾迹 深度学习 流场重构 本征正交分解 长短时记忆法 

分 类 号：U461.1[机械工程—车辆工程] TP18[交通运输工程—载运工具运用工程]