Nonlinear Aerodynamics Reduced-Order Model Based on Multi-Input Volterra Series

Abstract

This paper gives a low-order approximations of multi-input Volterra series as a nonlinear reduced-order model (ROM) based on wavelet multiresolution analysis. The band-limited pseudorandom multilevel sequence (PRMS) is used as the identification signal and the QR decomposition recursive least square (QRD-RLS) method is utilized as identification method. At last, the ROM is applied to model the nonlinear aerodynamic moment of an airfoil undergoing simultaneous forced pitch and plunge harmonic oscillation. The results show that including the second-order Volterra cross kernels in ROM can capture the coupling effect which significantly improves accuracy in predicting nonlinear aerodynamics under simultaneous motion. And the wavelet multiresolution analysis efficiently reduces the number of identification coefficients for Volterra kernels.