Rotating machinery is widely used in each field of national production, and the rolling bearing is one of the important parts in rotating machinery, the working state of rolling bearing is directly related to the operation of rotating machinery. Under normal circumstances,rotating machinery and several other more equipment together constitute a system, if the rotating mechanical failure, is likely to trigger a series of chain reaction, causing damage to the equipment, and even the emergence of casualties. As can be seen, the systematic research of vibration signal of rolling bearing, will be able to enhance the stability of equipment operation greatly, has certain theoretical and practical significance.The HHT method proposed by Huang since it’s adaptability is widely used in signal processing field, many scholars do research on it, this method is thought to be a major breakthrough to the traditional signal analysis methods with the Fourier transform is the core of them. According to the HHT method, empirical mode decomposition(Empirical Mode Decomposition, EMD) would generate mode cracking phenomenon, and Hilbert transform would generate cannot explain the negative frequency, the HHT was improved, the specific contents are as follows:Starting from the structure of the rolling bearing and rolling bearing type, summarized and analyzed the vibration mechanism and vibration type of rolling bearings, and based on the analysis of the development process, main failure forms of rolling bearing faults, and the frequency characteristics of vibration signals of rolling bearing were analyzed.According to the modal cracking phenomenon existing in the EMD method, and Hilbert transform has the negative frequency problem, put forward to extract fault characteristic frequency of rolling bearing with morphological filtering and EMD-AR algorithm, this paper describes the basic principle of morphology filter, the basic principle of EMD algorithm and the proposed morphological filter and EMD-AR algorithm, and a simulation of signal characteristic frequency extracted by morphological filtering and EMD-AR algorithm, prove the validity of the algorithm.A brief description of the basic principle of FFT, HHT algorithm, FFT method, and respectively by HHT method and morphological filtering and EMD-AR method to extract fault characteristic frequency of rolling bearing inner ring, through the experiment to prove that the morphological filtering method using the algorithm of this paper can reduce the modal cracking phenomena in the EMD to a certain extent, overcome the disadvantage of negative frequency Hilbert transform, obtain the spectrum of a smoother, and the higher theaccuracy of extraction of the fault characteristic frequency. |