Aircraft geometric feature and parameter joint modeling method based on deep learning

Abstract

The invention discloses an aircraft geometrical characteristic, layout parameter and flight state parameter combined modeling method based on deep learning, which belongs to the field of aircraft aerodynamic characteristic prediction, and comprises the following steps: a, carrying out geometrical characteristic extraction based on Bessel manifold aiming at layout parameters of an aircraft, constructing a Bezier manifold by adopting a Bezier curve or a curved surface, and extracting geometric features of the Bezier manifold to express a geometric structure of the aircraft; b, by introducing CNN, RBFNN and FCN networks, aircraft layout parameters, geometric features and flight state parameters are learned respectively, the three parameters are combined through another FCN to form a unified deep neural network, accurate prediction of the aerodynamic characteristics of the aircraft is achieved, the aircraft geometric feature extraction can be met, and the aircraft aerodynamic characteristic prediction accuracy is improved. And the modeling requirement for accurate prediction of the aerodynamic characteristics of the aircraft is met, and the aerodynamic characteristics of the aircraft can be well optimized.

Description

technical field [0001] The invention relates to the field of aircraft aerodynamic characteristic prediction, in particular to a joint modeling for predicting the aerodynamic characteristics of an aircraft by using aircraft geometric characteristics, layout parameters and flight state parameters as inputs. Background technique [0002] Aircraft layout parameters and flight state parameters are two important parameters that need to be considered in aerodynamic data modeling. However, the joint modeling technology that considers the geometric features, layout parameters and flight state parameters of the aircraft shape at the same time is not mature: on the one hand, under the same flight state parameters, different aircraft shapes will have different effects on the aerodynamic characteristics of the aircraft; On the one hand, the geometry of the aircraft is complex and difficult to describe. Therefore, it is difficult to fully reflect the details of the geometric shape of the...

AI Technical Summary

Problems solved by technology

[0006] The invention discloses a deep neural network modeling method for aerodynamic data with large differences, which can meet the modeling requirements when the aerodynamic shape and flight state change simultaneously, and can give the impact of the change of flight state and aerodynamic shape on the aerodynamic characteristics of the aircraft. The degree of influence can optimize the aerodynamic characteristics of the aircraft, but it does not solve how to further extract the geometric features that can reflect the shape change for the complex aircraft layout parameters, and jointly model the extracted geometric features with the layout parameters and flight state parameters , the problem of improving the modeling accuracy of the aerodynamic model

Images