Aerodynamic force modeling of quasi-steady stall phenomenon based on UKF-WNN

Author

Zhao, Liang ; Liu, Xiaodong ; Lei, Jing

Author_Institution

Eng. Coll., Air Force Eng. Univ., Xi´´an, China

fYear

2009

fDate

8-11 Aug. 2009

Firstpage

213

Lastpage

217

Abstract

The paper proposed an algorithm which can get over the BP algorithm´s shortcomings of slow convergence speed, computation complexity and local minimum by using the UKF to estimate the parameters of WNN. Then it takes the phenomenon of aerodynamic modeling of quasi-steady stall for ATTAS aircraft as applying background and uses the algorithm of BP, EKF and UKF to train the WNN respectively. From the simulation results we can see that the UKF algorithm is faster in training speed and more accurate in prediction when compared with BP and EKF and it is competent for modeling of complex nonlinear aerodynamic phenomenon as well.

Keywords

Kalman filters; aerodynamics; aircraft; backpropagation; computational complexity; minimisation; neural nets; nonlinear filters; parameter estimation; wavelet transforms; ATTAS aircraft; UKF-WNN; aerodynamic force modeling; backpropagation training algorithm; computation complexity; local minimum; parameter estimation; quasisteady stall phenomenon; unscented Kalman filter; wavelet neural network; Aerodynamics; Aircraft; Artificial neural networks; Convergence; Educational institutions; Land vehicles; Military computing; Neural networks; Predictive models; Road vehicles; aerodynamic force; kalman filter; neural network; wavelet;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Information Technology, 2009. ICCSIT 2009. 2nd IEEE International Conference on

Conference_Location

Beijing

Print_ISBN

978-1-4244-4519-6

Electronic_ISBN

978-1-4244-4520-2

Type

conf

DOI

10.1109/ICCSIT.2009.5234422

Filename

5234422