Defect feature recognition and diagnosis method for abnormal bearing in noise inspection

The invention discloses a defect feature recognition and diagnosis method for an abnormal bearing in noise inspection. The defect feature identification and diagnosis method for the abnormal bearing in noise inspection provided by the invention comprises the following steps: extracting various defect features of the abnormal bearing in noise inspection through an electron microscope analysis method, comprehensively diagnosing the identified defect feature information and a whole-flow production process, and determining the defect source of the abnormal bearing in noise inspection. According to the method, for a large number of bearing rings and steel balls with abnormal noise in sound detection, after sampling and processing, defect features causing abnormal noise of bearing parts are identified and diagnosed through a method of combining low-magnification magnifier defect positioning and high-magnification SEM scanning electron microscope with EDS energy spectrum analysis, the source of the defect features i