Special Issue: Aircraft Digital Twin Technology

Aircraft attitude prediction model based on physical information neural networks

  • Yugang ZHANG ,
  • Zhe YANG ,
  • Senpeng HE ,
  • Wenqing YANG
Expand
  • 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 date: 2025-01-27

  Revised date: 2025-04-21

  Accepted date: 2025-06-16

  Online published: 2025-06-27

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%.

Cite this article

Yugang ZHANG , Zhe YANG , Senpeng HE , Wenqing YANG . Aircraft attitude prediction model based on physical information neural networks[J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2025 , 46(19) : 531850 -531850 . DOI: 10.7527/S1000-6893.2025.31850

References

[1] WANG G Z, PEI B B, XU H J, et al. Risk quantification and visualization method for loss-of-control scenarios in flight[J]. Aerospace202310(5): 416.
[2] WANG G Z, XU H J, PEI B B. An intelligent approach for flight risk prediction under icing conditions[J]. Chinese Journal of Aeronautics202336(6): 109-127.
[3] KANG C, WOOLSEY C A. Model-based path prediction for fixed-wing unmanned aircraft using pose estimates[J]. Aerospace Science and Technology2020105: 106030.
[4] 袁修久, 赵学军, 李嘉林. 基于航迹的飞机姿态角建模与仿真[J]. 系统工程与电子技术201638(4): 889-894.
  YUAN X J, ZHAO X J, LI J L. Modeling and simulation of aircraft attitude angle based on air-path[J]. Systems Engineering and Electronics201638(4): 889-894 (in Chinese).
[5] 孙博, 高永卫, 魏斌斌. 无人机离机轨迹与姿态的高准度快速预测方法研究[J]. 弹箭与制导学报202343(6): 97-104.
  SUN B, GAO Y W, WEI B B. Research on a fast and precise prediction method for UAV separation trajectory and attitude[J]. Journal of Projectiles, Rockets, Missiles and Guidance202343(6): 97-104 (in Chinese).
[6] 李海泉, 陈小前, 张嘉图, 等. 随机激励作用下飞机飞行姿态动力学研究[J]. 航空学报202243(3): 225232.
  LI H Q, CHEN X Q, ZHANG J T, et al. Study on aircraft attitude dynamics under random excitation[J]. Acta Aeronautica et Astronautica Sinica202243(3): 225232 (in Chinese).
[7] 杨丽梅, 郭立红. 一种序列图像飞行器姿态的预测方法[J]. 计算机测量与控制200614(6): 797-799.
  YANG L M, GUO L H. Aircraft pose prediction method of image sequences[J]. Computer Measurement & Control200614(6): 797-799 (in Chinese).
[8] PASHA A ALI, SANKARALINGAM L, RAHMAN M M, et al. MEMS fault-tolerant machine learning algorithm assisted attitude estimation for fixed-wing UAVs[J]. Engineering Applications of Artificial Intelligence2024129: 107608.
[9] 曹咏弘, 张慧, 马铁华, 等. 基于神经网络的无陀螺捷联惯导系统姿态预测[J]. 中国惯性技术学报200816(2): 159-161, 170.
  CAO Y H, ZHANG H, MA T H, et al. Attitude forecast of gyroscope-free SINS based on neural network[J]. Journal of Chinese Inertial Technology200816(2): 159-161, 170 (in Chinese).
[10] 梁少军, 郑幸, 谢礼鹏, 等. 固定翼无人机纵向控制回路多工况状态预测[J]. 兵器装备工程学报202041(10): 203-209.
  LIANG S J, ZHENG X, XIE L P, et al. Muti-condition state prediction of longitudinal control loop of fixed-wing UAV[J]. Journal of Ordnance Equipment Engineering202041(10): 203-209 (in Chinese).
[11] 路晶, 史宇, 任洲. 基于增量式在线学习飞行训练姿态预测研究[J]. 航空计算技术202252(5): 5-8.
  LU J, SHI Y, REN Z. Research on attitude prediction of flight training based on incremental online learning[J]. Aeronautical Computing Technique202252(5): 5-8 (in Chinese).
[12] RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational Physics2019378: 686-707.
[13] CAO W B, SONG J H, ZHANG W W. A solver for subsonic flow around airfoils based on physics-informed neural networks and mesh transformation[J]. Physics of Fluids202436(2): 027134.
[14] 李野, 陈松灿. 基于物理信息的神经网络: 最新进展与展望[J]. 计算机科学202249(4): 254-262.
  LI Y, CHEN S C. Physics-informed neural networks: Recent advances and prospects[J]. Computer Science202249(4): 254-262 (in Chinese).
[15] PANG G F, LU L, KARNIADAKIS G E. fPINNs: Fractional physics-informed neural networks[J]. SIAM Journal on Scientific Computing201941(4): A2603-A2626.
[16] HE G Y, ZHAO Y X, YAN C L. MFLP-PINN: A physics-informed neural network for multiaxial fatigue life prediction[J]. European Journal of Mechanics-A/Solids202398: 104889.
[17] BAI Z W, SONG S F. Physics-informed neural network for first-passage reliability assessment of structural dynamic systems[J]. Computers & Structures2023289: 107189.
[18] MOHAJERIN N, MOZIFIAN M, WASLANDER S. Deep learning a quadrotor dynamic model for multi-step prediction[C]∥2018 IEEE International Conference on Robotics and Automation (ICRA). Piscataway: IEEE Press, 2018: 2454-2459.
[19] GU W B, PRIMATESTA S, RIZZO A. Physics-informed neural network for quadrotor dynamical modeling[J]. Robotics and Autonomous Systems2024171: 104569.
[20] 王康晋. 基于PINN的细长体飞行器气动特性预测研究[D]. 绵阳: 西南科技大学, 2024: 13-15.
  WANG K J. Research on predicting aerodynamic characteristics of slender aircraft based on PINN[D]. Mianyang: Southwest University of Science and Technology, 2024: 13-15 (in Chinese).
[21] 付军泉, 钟伯文, 钟运琴, 等. 基于物理信息神经网络的飞机气动参数辨识方法[J]. 空气动力学学报202341(9): 30-37.
  FU J Q, ZHONG B W, ZHONG Y Q, et al. A physics informed neural network based method for aircraft aerodynamic parameter identification[J]. Acta Aerodynamica Sinica202341(9): 30-37 (in Chinese).
[22] 方振平, 陈万春, 张曙光. 航空飞行器飞行动力学[M]. 北京: 北京航空航天大学出版社, 2005: 22-24, 174-181.
  FANG Z P, CHEN W C, ZHANG S G. Aerodynamic flight dynamics of aircraft[M]. Beijing: Beijing University of Aeronautics & Astronautics Press, 2005: 22-24, 174-181 (in Chinese).
[23] 金长江. 飞行动力学: 飞机飞行性能计算[M]. 北京: 国防工业出版社, 1983: 29-34.
  JIN C J. Flight dynamics: Aircraft flight performance calculation[M]. Beijing: National Defense Industry Press, 1983: 29-34 (in Chinese).
Outlines

/