Abstract:A blind source separation method of machine fault diagnosis based on canonical correlation analysis was presented. Compared with the blind source separation method of machine fault diagnosis based on independent component analysis (ICA), only the statistical distribution of the sample values was considered in traditional blind source separation method, without regard to the time and spatial relationship between the source signals. However, in the proposed method, this defect was overcome, and the autocorrelation of source signal was used to separate the mixture signals. The simulation results show that using the proposed method, satisfactory separation effect and much more high computational efficiency were obtained than the traditional ICA method. Finally, the proposed method was applied to the fault diagnosis of rolling bearing. The experiment results validate the effectiveness of the proposed method.