Study on fault diagnosis of rolling bearing based on time-frequency generalized dimension

Combining with the multi-fractal theory, a multi-fractal fault diagnosis theory based on signals in time and frequency domain has been provided, and the concrete method and numerical algorithm are supplied. According to the definition of multi-fractal and the related principle of fractal geometry, by math method and the related theory of signal processing, the research of the effect to the calculation of the selection of mesh type J and q parameter in algorithm has been done, and the value range of J and the cut off condition of q are optimized. By LabVIEW simulation experiment, the multi-fractal method in time and frequency domain can be proved. The method can complete effective signal identification, which is related to the factors such as signal energy and distribution. It has specific law. According to the law, the further fault diagnosis experiment of bearings can be done. The experiment has shown that not only the multi-fractal method in time and frequency domain can complete the fault diagnosis, but also in the early stage of fault, the fault detection can be done. Therefore, that the multi-fractal method in time and frequency is used into fault diagnosis of bearing is a practicable method.