Based on the research of LCD,this paper borrowed ideas from the sifting based signal decomposition methods such as EMD and LCD,and then presented a new mono-component signal with physically meaningful instantaneous frequencies, i.e.,ICC,for restraining the mode mixing problem of LCD and improving the deficiency in terms of the mean curve definition. Finally, a new signal decomposition method,ACD was proposed.Meanwhile, the evaluation criterion of ICC was also given.In the sifting procedure for separating certain order components,a set of ICCs was obtained by using different mean curves and compact coefficients. The optimal ICC for this order sifting would be selected from the candidate ICCs using the evaluation criterion of ICC, which guaranteed ACD outperforms LCD.The simulation results indicate that the decomposition effect of ACD is better than that of EMD,LCD,ensemble empirical mode decomposition(EEMD) and autonomous compact local characteristic-scale decomposition(ACLCD), experimental results of the experimental signals show its validity.Thus a new way for fault diagnosis of rotating machinery is provided.
