A deep learning-based approach for flow field prediction in a dual-mode combustor  

作　　者：Chen Kong Ziao Wang Fuxu Quan Yunfei Li Juntao Chang 

机构地区：[1]Harbin Institute of Technology,Harbin 150001,China

出　　处：《Propulsion and Power Research》2024年第2期178-193,共16页推进与动力（英文）

基　　金：supported by the National Natural Science Foundation of China (Grant No.11972139 and 52125603);the Fundamental Research Funds for the Central Universities (HIT.BRET.2021006 and FRFCU5710094620).

摘　　要：Accurate acquisition of the distribution offlow parameters inside the supersonic combustor is of great significance for hypersonicflight control.It is an interesting attempt to introduce a data-driven model to a supersonic combustor forflowfield prediction.This paper proposes a novel method for predicting theflowfield in a dual-mode combustor.Aflowfield prediction convolutional neural network with multiple branches is built.Numerical investiga-tions for a strut variable geometry combustor have been conducted to obtainflowfield data for training the network as aflowfield prediction model.Richflowfield data are obtained by changing the equivalent ratio,incomingflow condition and geometry of the supersonic combustor.The Mach number distribution can be obtained from the trainedflowfield prediction model using the combustor wall pressure as input with high accuracy.The accuracy offlowfield prediction is discussed in several aspects.Further,the combustion mode detection is im-plemented on the predictionflowfield.

关 键 词：Flowfield prediction Deep learning(DL) Convolutional neural networks(CNNs) Data-driven model Dual-mode combustor Variable geometry combustor 

分 类 号：V23[航空宇航科学与技术—航空宇航推进理论与工程]