Objective The background noise in the engine room during actual ship navigation leads to poor accuracy in fault diagnosis methods. To address this issue, proposes a ship motor fault diagnosis method based on complete ensemble empirical mode decomposition （CEEMDAN）and RECAN.

Methods First, the noisy motor fault signal is decomposed into multiple intrinsic mode components （IMFs） through adaptive noise CEEMDAN, and the noise dominant signal and information dominant signal in the IMF are divided based on detrended fluctuation analysis. Empirical wavelet transform （EWT） is used to de-noise the noise dominant signal; Then, construct a Bayesian residual efficient channel attention network （BRECAN） model, use network parameters to model and fit the interference of noise on model training, and guide the network to extract fault differential features through the RECA module; Finally, the effectiveness of the proposed method was verified through a motor fault simulation experimental platform.

Results The results show that the proposed method can achieve accurate diagnosis of ship motor faults under strong noise, and still maintain a diagnostic accuracy of over 90% underSNR = - 12.

Conclusion The research results provide reference for the diagnosis of ship motor faults under strong noise.