Due to the abundant information contained in multichannel data, a multichannel fusion method based on coupled hidden Markov models (CHMMs) is presented for rolling- element-bearing’s fault diagnosis. A double-chain CHMM is adopted to combine the horizontal and vertical vibration signals from bearings. An experiment was carried out to validate the proposed method. The experimental results show that the proposed method can improve the diagnostic accuracy when compared with the conventional hidden Markov model (HMM) based method.

The phenomenon of stochastic resonance of an over-damped second-order linear system subject to dichotomic noise (DN) is investigated. Based on linear-system theory and the correlation deletion method, the explicit expression of the output amplitude gain (OAG) of the system is obtained. It is shown that the OAG is a non-monotonic function of the strength, the correlation time of the DN, the system damping coefficient, as well as the frequency of the driving signal. In addition, by choosing appropriate parameters of the noise and the system, the OAG of the noisy system can be larger than that of the noise-free system.