Nonlinear response prediction of cracked rotor based on EMD

Abstract The empirical mode decomposition (EMD) method is introduced, to improve the prediction accuracy of cracked rotor׳s nonlinear response during a long-term period. The EMD method was applied to decompose the nonlinear response into series of intrinsic mode functions (IMF). Consequently, the prediction results of IMF were obtained, based on the maximal local Lyapunov exponent (LLE). By adding all the prediction results of IMF, the nonlinear response of cracked rotor can be predicted, called the IMF prediction method. Compared with the response predicted directly by the maximal local Lyapunov exponent, when the forecasting step is less than the maximal prediction time which is calculated by the multiplicative inverse of maximal Lyapunov exponent, the IMF method has the same prediction accuracy. When the forecasting step is greater than maximal prediction time, the IMF prediction method is more advantageous than the Lyapunov prediction method. Bently RK4 rotor test is used to validate the IMF prediction method׳s advantage.