A DIAGNOSIS METHOD OF ROLLING BEARING FAULT BASED ON ADAPTIVE MATHEMATICAL MORPHOLOGY

The invention relates to a diagnosis method of rolling bearing fault based on adaptive mathematical morphology, which belongs to the field of fault diagnosis technology. It includes the following steps: correlation analysis and rolling bearing fault diagnosis. The invention adopts the above-mentioned diagnosis method of rolling bearing fault based on adaptive mathematical morphology, which has significant advantages such as amplitude non-offset and non-attenuation. The morphology opening operation and morphology closing operation are allocated in a more reasonable proportion, which effectively improves the feature extraction ability of the signal. Combined with the spectral correlation analysis method, the fault type and fault grade of the unknown signal are identified according to the cross-correlation coefficient of the unknown signal to be tested and the trained signal. The diagnosis method in this paper can not only identify different types of faults, but also identify faults with different damage levels. Compared with traditional methods, it has higher recognition rate and reliability.