Adaptive WNN aerodynamic modeling based on subset KPCA feature extraction

Abstract

In order to accurately describe the dynamic characteristics of flight vehicles through aerodynamic modeling, an adaptive wavelet neural network (AWNN) aerodynamic modeling method is proposed, based on subset kernel principal components analysis (SKPCA) feature extraction. Firstly, by fuzzy C-means clustering, some samples are selected from the training sample set to constitute a sample subset. Then, the obtained samples subset is used to execute SKPCA for extracting basic features of the training samples. Finally, using the extracted basic features, the AWNN aerodynamic model is established. The experimental results show that, in 50 times repetitive modeling, the modeling ability of the method proposed is better than that of other six methods. It only needs about half the modeling time of KPCA-AWNN under a close prediction accuracy, and can easily determine the model parameters. This enables it to be effective and feasible to construct the aerodynamic modeling for flight vehicles.

Keywords

wavelet / neural network / fuzzy C-means clustering / kernel principal components analysis / feature extraction / aerodynamic modeling

References

[1] LinseD J, StrengelR F. Identification of aerodynamic coefficients using computational neural networks [J]. Journal of Guidance, Control and Dynamics, 1993, 16(6): 1018-1025

[2] PeyadaN K, GhoshA K. Longitudinal aerodynamic parameter estimation using neural network and Gauss Newton method [J]. Journal of Aerospace Sciences and Technology, 2009, 61(2): 295-304

[3] KumarR, GanguliR, OmkarS N, KumarM V. Rotorcraft parameter estimation from real time flight data using radial basis function networks [J]. Journal of Aircraft, 2008, 45(1): 333-341

[4] ZhaoL, LiuX-d, LeiJing. UKF-WNN and its applications to the quasi-steady stall aerodynamic force modeling [J]. Acta Aerodynamica Sinica, 2010, 28(3): 353-357

[5] GanX-s, DuanmuJ-S, CongWei. SSUKF-WNN algorithm and its applications in aerodynamic modeling of flight vehicles [J]. Control and Decision, 2011, 26(2): 187-190

[6] MengY-b, ZouJ-h, GanX-s, ZhaoLiang. Research on wnn aerodynamic modeling from flight data based on improved pso algorithm [J]. Neurocomputing, 2012, 83: 212-221

[7] GanX-s, DuanmuJ-S, MengY-b, CongWei. Aerodynamic modeling from flight data based on WNN optimized by particle swarm [J]. Acta Aeronautica et Astronautica Sinica, 2012, 33(7): 1209-1929

[8] BaudatG, AnouarF. Feature vector selection and projection using kernels [J]. Neurocomputing, 2003, 55(1/2): 21-38

[9] FanY-g, LiP, SongZ-huan. KPCA based on feature samples for fault detection [J]. Control and Decision, 2005, 20(12): 1415-1422

[10] ScholkopfB, SmolaA, MullerK R. Nonlinear component analysis as a kernel eigenvalue problem [J]. Neural Computation, 1998, 10(5): 1299-1319

[11] WASHIZAWA Y. Subset kernel principal component analysis [C]// Proceedings of 2009 IEEE International Workshop on Machine Learning for Signal Processing. Grenoble, France, 2009: 1–6.

[12] WASHIZAWA Y. Subset basis approximation of kernel principal component analysis [R]. Technical Report 2012–03. The University of Electro-Communications.

[13] WangH-y, YaoZ-a, LiLei. The application of feature extraction on using kernel principle component analysis based on clustering [J]. Computer Science, 2005, 32(4): 64-66

[14] ZhangL, HuangS-g, GuoHao. A fast kernel supervised locality preserving projection algorithm [J]. Journal of Electronics & Information Technology, 2011, 33(5): 1049-1054

[15] BezdekJ C. Recent convergence results for the fuzzy C-means clustering algorithm [J]. Classification, 1998, 5(2): 237-247

[16] WangH-m, KimJ-h, JungD-y. LEE Sang-min. Power interconnected system clustering with advanced fuzzy C-mean algorithm [J]. Journal of Central South University of Technology, 2011, 18: 190-195

[17] ZhangQ-h, BenvenisteA. Wavelet Network [J]. IEEE Transactions on Neural Networks, 1992, 3(6): 889-898

[18] LiJ-P, WangF-T, YangBo. Fast training algorithm of BP wavelet neural network [J]. Systems Engineering and Electronics, 2001, 23(8): 72-75

[19] JategaonkarR V. Flight vehicle system identification: A time domain methodology [M]. Progress in Astronautics and Aeronautics Series, 20061st Edition204-207

[20] KleinV, MorelliE A. Aircraft system identification: theory and practice [M]. Progress in Astronautics and Aeronautics Series, 20061st ed333-343

[21] ZengQ-yongFaint signal detection [M], 1994HangzhouZhejiang University Press153-162

[22] TippingM E. Sparse kernel principal component analysis [J]. Advances in Neural Information Processing Systems, 2001, 13: 633-639