Fault Diagnosis Method Based on Two-level Adaptive Chirp Mode Decomposition and Synchro-extracting Transform
Synchroextracting transform(SET)lacks adaptivity in handling strong interfering signal components,which leads to frequency ambiguity and makes it difficult to accurately extract the instantaneous frequencies with rapid fluctuations.In view of this problem,a fault diagnosis method is constructed by introducing adaptive chirp mode decomposition(ACMD)into the SET.The adaptive prior information of ACMD is combined with the advantages of the greedy algorithm.And a double grade fault diagnosis method is established based on ACMD-SET.In this method,the component selection recombination algorithm based on the Gini index(GI)maximization criterion and the Level 1 ACMD are combined to extract the patterns of multi-modal fault pulse signals with strong disturbances.The time-varying fault feature frequencies separated by level 2 ACMD are then represented in time-frequency with high accuracy using SET.To verify the effectiveness,high-resolution fault features of a simulated AM-FM signal are obtained by the proposed method.Finally,this method is applied to the vibration signal analysis of high-speed rolling bearings in aero-engines.The results show that the proposed method can effectively extract the time-varying fault characteristic frequencies of high-speed rolling bearing vibration signals,and the obtained results are significantly better than the SET method.
国家科技期刊平台
NSTL
万方数据
