Research and application of ensemble empirical mode decomposition principle and 1.5 dimension spectrum method

To extract the gear crack fault weak feature of locomotive running gear box on complex background, a new method of ensemble empirical mode decomposition (EEMD) and 1.5 dimension spectrum for fault feature extraction is proposed. The vibration signal is adaptively anti alias decomposed by EEMD method to get intrinsic mode function (IMF) of different frequency bands, then 1.5 dimension spectrum as a post processing method is adopted to process IMF which contains fault feature information. This method is endowed with characteristics of avoiding model mixing, suppressing Gaussian white noise, detecting the nonlinear coupling feature. Based on time-frequency character of the signal and the principle of EEMD, a criterion of adding Gaussian white noise in EEMD method is proposed, and the anti alias decomposing ability of EEMD method is verified by signal simulation experiment. EEMD and 1.5 dimension spectrum are introduced into monitoring diagnosis of a locomotive running gear box, and the results show that this method enables to successfully extract the early crack fault of the gear tooth root in gearbox.