Abstract: Considering the difficulty of the extracting for the fault feature of blade in service, an intelligent recognition method for defects in blade based on Complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) combined with support vector machine (SVM) is proposed. First of all, the signals originated from defects of the blade were decomposed by Complete Ensemble Empirical Mode Decomposition with adaptive noise, and the intrinsic mode functions (IMF) containing the main feature information were selected by cross-correlation number. Then, the main Energy Entropy of the intrinsic mode functions was calculated and the vectors of Complete Ensemble Empirical Mode Decomposition with adaptive noise energy entropy were constructed. Finally, in order to verify the reliability of vector of energy entropy, the pattern recognition was carried out by the support vector machine (SVM) for different defects. The results show that the average recognition rate for different faults were as high as 96.7%, and it suggests that the method of Complete ensemble empirical mode decomposition with adaptive noise energy entropy combined with SVM can be supplied well for recognizing and warning in blade during the early process.