关键词: 气动力建模, 集成神经网络, 大迎角飞行仿真, 过失速机动, 非定常气动力

Abstract: Due to the significant nonlinear and unsteady effects, it is difficult to accurately simulate the high angle of attack maneu-vering characteristics of aircraft by existing wind tunnel experiments and numerical methods. In order to improve the ac-curacy of high angle of attack maneuver flight simulation, the physical-model-embedding ensemble neural network was developed to accurately model the unsteady aerodynamics of aircraft at high angles of attack, and the flight dynamics equations were further coupled in time domain to realize high angle of attack maneuver flight simulation. Taking the F-16 fighter as the research object, the open loop broadband excitation, open loop Due to the significant nonlinear and unsteady effects, it is difficult to accurately simulate the high angle of attack maneu-vering characteristics of aircraft by existing wind tunnel experiments and numerical methods. In order to improve the ac-curacy of high angle of attack maneuver flight simulation, the physical-model-embedding ensemble neural network was developed to accurately model the unsteady aerodynamics of aircraft at high angles of attack, and the flight dynamics equations were further coupled in time domain to realize high angle of attack maneuver flight simulation. Taking the typi-cal fighter as the research object, the open loop broadband excitation, open loop harmonic excitation and post stall ma-neuver flight data of high angle of attack longitudinal flight are taken as the data sample of aerodynamic modeling. Be-sides, the traditional dynamic derivative model, black box neural network model and ensemble neural network model are constructed and compared, and the flight characteristics of coupled simulation are further compared. Besides, the idea of using the flight simulation method to test the robustness of the aerodynamic model is proposed. The results show that the aerodynamic modeling error of the physical-model-embedding ensemble neural network is 57% lower than that of the traditional dynamic derivative model, and has better robustness and stability in the coupling process. The aircraft re-sponse error is 63% lower than that of the black box neural network model, which proves the advantages and engineering potential of the proposed modeling framework in small sample flight data identification.

Key words: aerodynamic modeling, ensemble neural networks, high angle of attack fight simulation, post stall maneuver, unsteady aerodynamics