State recognition of viscoelastic sandwich structures based on permutation entropy and generalized Chebyshev support vector machine

Abstract

Viscoelastic sandwich structures are widely used in mechanical equipment, yet viscoelastic materials always suffer from aging which changes the dynamic characteristics of the structure and affects the whole performance of the equipment. Therefore, state recognition of viscoelastic sandwich structures is very necessary for monitoring structural health states and keeping the equipment running reliably. Considering the high nonlinearity on the dynamic characteristics and the strong nonstationarity of vibration response signals, a novel method based on permutation entropy (PE) and generalized Chebyshev support vector machine (GCSVM) is proposed in this paper. For obtaining more effective state information, redundant second generation wavelet packet transform is firstly used to process the non-stationary vibration response signals, and then PE features are extracted from the resultant wavelet packet coefficients to reveal the changes of the nonlinear dynamic characteristics. Aiming at improving the generalization ability of SVM, based on generalized Chebyshev kernel, the GCSVM is introduced to classify the various structural states. In order to demonstrate the effectiveness of the proposed method, different structural states are created by the accelerated aging experiment of viscoelastic material. The testing results show that the proposed method is effective for state recognition of viscoelastic sandwich structures, which has more strong generalization ability and can achieve higher recognition accuracy than that of the wavelet SVM. Keywords-redundant second generation wavelet packet transform; permutation entropy; generalized Chebyshev support vector machine; viscoelastic sandwich structures; state recognition I.This full text paper was peer-reviewed at the direction of IEEE Instrumentation and Measurement Society prior to the acceptance and publication.978-1-4799-6144-6/15/$31.00 ©2015 IEEE