A cooperative diagnosis technique for rolling bearing faults was proposed for aero-engine rolling bearing fault diagnosis based on casing measuring point signal. Firstly

the minimum entropy deconvolution was used to eliminate the influence of the signal transmission path and enhance the impulsive component in the signal.Then

the resonance band was extracted by applying wavelet transform. Finally

the non-periodic signal components in the resonance band were suppressed by using autocorrelation analysis while the signal-to-noise ratio was further improved. Two bearing tests were carried out respectively on the artificial fault bearing and the real fault bearing on the rotor tester with casing. Test results showed that compared with other typical methods

the spectrum peaks corresponding to the fault characteristic frequencies in the envelope spectrum obtained by the proposed cooperative diagnosis method were more clear and obvious.