은닉마르코프 모델과 경험모드분리법을 이용한 회전 블레이드의 크랙 위치 및 깊이 예측

Abstract

Crack location and depth prediction method of a cracked rotating blade employing Hidden Markov Model(HMM) and Empirical Mode Decomposition(EMD) is proposed in this study. To predict the location and depth employing HMM, appropriate feature vectors which represent characteristics about each location and depth should be extracted from transient responses of the system first. In this study, EMD and Fast Fourier Transform (FFT) are employed to obtain the feature vectors for HMM from transient responses of the system. Then, the crack location and depth are predicted by employing the feature vectors and HMM. Predicted results show that the crack location and depth can be identified accurately using the proposed method.