Aerodynamic Thermal Data Analysis Of Hypersonic Aircraft Based On Neural Network And Database Research

The rapid development of near-space hypersonic vehicles makes the demand for aerodynamic heat research more urgent.The use of existing data to quickly calculate aerodynamic heat based on machine learning methods is a useful way to study aerodynamic heat.This paper studies the aerodynamic thermal database and its management system,and studies the aerodynamic thermal research from the aspects of data association and data fusion by means of machine learning methods such as neural network,and integrates the data analysis function into the aerodynamic thermal database system to establish the technical aerodynamic research ideas and methods based on machine learning.Firstly,this paper studies the data correlation technology of using neural network to study wall heat flow,which is based on the idea of combining the non-viscous numerical solution outside the boundary layer with the engineering method in the boundary layer,and develops the idea of using neural network to correlate wall heat flow with Euler's calculation results.The specific path is to extract the correlation parameters needed for the associated wall heat flow from the dimensionless equation;then design a reasonable neural network model to correlate the relationship between the heat flow and the associated parameters;then use the neural network model to correlate the calculated wall heat flow according to the Euler results.The neural network model is built using the Python language and the Keras library.Finally,this paper uses the blunt cone and HTV-2 model to verify the aerodynamic thermal correlation method,and proposes a method of partitioned machine learning for the complex shape.The neural network-based aerodynamic thermal correlation method is validated.Secondly,this paper studies the data fusion method based on the improved VCM(Variable Complexity Modeling)method and the neural network method.By using the difference of precision and quantity between different aerodynamic thermal data sources,the fusion of different data sources is used to study the acquisition accuracy and more reliable aerodynamic thermal data path.On this basis,the experimental data and CFD data of-the blunt biconical meridian region/whole surface region under different angles of attack are used as data sources for fusion verification to ensure the validity and applicability of the proposed method.Finally,this paper builds a aerodynamic thermal database management system to manage and store a large amount of aerodynamic data,and provides support for data association research and data fusion research,and integrates data analysis and data fusion data analysis functions on the system.The database system of this paper mainly adopts C#language to develop based on.NET platform,adopts PostgreSQL as the underlying database support,adopts ASP.NET as the front-end implementation mode,and the whole architecture adopts MVC mode for development,and finally gets an integrated aerodynamic data management platform.