||In the condition monitoring and fault diagnosis of dual-rotor bearings, the signal transmission path is complicated, and it is difficult to obtain signals directly through the accelerometer. Acoustic signal contains a large amount of characteristic information, and has the advantage of non-contact measurement. In order to accurately and effectively achieve fault diagnosis through acoustic signals, a rolling bearing fault diagnosis method based on adaptive noise complete set empirical mode decomposition (CEEMDAN) and Multipoint Optimal Minimum Entropy Deconvolution Adjusted (MOMEDA) is proposed. Firstly, the MOMEDA method is used to enhance the impact in the acoustic signal, then the CEEMDAN processing is performed on the processed signal, the kurtosis value of each empirical modal component is calculated, and the optimal component is selected according to the kurtosis value. Hilbert envelope spectrum is used to accurately extract fault feature frequencies. This method is used to realize the fault diagnosis of rolling bearing based on acoustic signal, which provides an optimal component selecting standard for the separated signals, which can reduce the complexity of fault diagnosis and has a good adaptability.