The depth neural network method is applied to the performance prediction of scramjet nozzle, and a network model based on the thrust coefficient corresponding to the pressure distribution unit is proposed. The convolution neural network model is established by parameterization of nozzle geometry feature extraction and pressure distribution curve, and the simulation results of computational fluid dynamics are used as training samples to predict the wall pressure distribution of scramjet nozzle. The results show that: The prediction model of nozzle performance based on neural network has good prediction accuracy, and the prediction error is less than 0.5% in training set and validation set, which is in good agreement with CFD data. The model presented in this paper is robust, and its performance will not be greatly influenced by the change of data set and initial parameters. The prediction model based on neural network takes only 1/500 of CFD, and has a great advantage in the prediction speed.
Thrust Performance Prediction of Scramjet Nozzle Based on Depth Neural Network
Lect. Notes Electrical Eng.
Asia-Pacific International Symposium on Aerospace Technology ; 2023 ; Lingshui, China October 16, 2023 - October 18, 2023
2023 Asia-Pacific International Symposium on Aerospace Technology (APISAT 2023) Proceedings ; Kapitel : 35 ; 400-410
02.07.2024
11 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Measurement of Scramjet Nozzle Thrust in a Hypersonic Impulse Facility
British Library Conference Proceedings | 1995
|Thrust Nozzle Design Study for a Quasi-Axisymmetric Scramjet-Powered Vehicle
Online Contents | 2011
|Wiley | 2022
|