导航

Acta Aeronautica et Astronautica Sinica ›› 2025, Vol. 46 ›› Issue (19): 531850.doi: 10.7527/S1000-6893.2025.31850

• Special Issue: Aircraft Digital Twin Technology • Previous Articles     Next Articles

Aircraft attitude prediction model based on physical information neural networks

Yugang ZHANG1,2(), Zhe YANG1,3, Senpeng HE1,3, Wenqing YANG1,2   

  1. 1.School of Aeronautics,Northwestern Polytechnical University,Xi’an 710072,China
    2.National Key Laboratory of Aircraft Configuration Design,Xi’an 710072,China
    3.Chinese Flight Test Establishment,Xi’an 710089,China
  • Received:2025-01-27 Revised:2025-04-21 Accepted:2025-06-16 Online:2025-06-30 Published:2025-06-27
  • Contact: Yugang ZHANG E-mail:zhangyugang@nwpu.edu.cn
  • Supported by:
    Special Research on Civil Aircraft(MJZ5-1N22)

Abstract:

In order to avoid the flight test safety accidents caused by exceeding the flight envelope during flight test, and reduce the accident risk, the subsequent attitude evolution process of the aircraft was predicted according to the current state information of the aircraft and the angle data of rudder deflection, so as to provide a basis for pilot decision-making support. Combining flight dynamics equation and Physical Information Neural Networks (PINNs), a real-time aircraft attitude prediction model, Flight Dynamics-Physics Informed Neural Network (FD-PINN), was constructed to solve the problem of insufficient prediction accuracy and poor generalization ability of the Neural Network-based (NN)aircraft attitude prediction model. Considering the randomness of atmospheric environmental parameters and the uncertainty of aircraft control inputs, flight simulation data were obtained by FlightGear to verify the model. The calculation results show that FD-PINN model has stronger generalization ability and higher prediction accuracy than NN model, and the mean square error of angle of attack prediction results is reduced by 68.5%.

Key words: attitude prediction, physical information neural networks, flight dynamics, flight test safety, flight test

CLC Number: