Open AccessNeural Partial Differentiation Based Nonlinear Parameter Estimation from Noisy Flight Data
Author(s) -
Majeed Mohamed,
Guruprasad Madhavan,
R. Manikantan,
PS Lal Priya
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1215/1/012025
Subject(s) - nonlinear system , control theory (sociology) , extended kalman filter , aerodynamics , kalman filter , thrust , computer science , engineering , aerospace engineering , physics , artificial intelligence , control (management) , quantum mechanics
This paper focuses on the application of neural partial differentiation (NPD) approach to estimate the longitudinal parameters of an aircraft HFB 320 from noisy flight data. By exciting both short period and phugoid modes of an aircraft with thrust variation, the aircraft system dynamics becomes highly nonlinear and aerodynamic parameters appears nonlinear to the state trajectories of velocity, AOA, pitch rate and pitch angle. This paper highlights the application of NPD for such a class of nonlinear dynamics; previously it was used only for the estimation of parameter appearing linear to the states. The extracted the nonlinear longitudinal parameters of HFB 320 aircraft are compared with the parameters estimated by using adaptive Unscented Kalman Filter (UKF) approach. Finally, the estimation results are validated by comparing with flight data and the responses obtained from the estimates by adaptive UKF.