Method for extracting bearing fault feature frequency based on singular value decomposition and optimized frequency band entropy and application thereof

摘要: The invention relates to a method for extracting bearing fault feature frequency based on singular value decomposition and optimized band entropy the application thereof, belongs tothe field of mechanical diagnosis signal processing. concept singularity kurtosis relative change rate is provided indexes, SVD reconstruction order determined by adopting rate. simple in principle takes as theoretical basis has solid when compared with other methods. denoising effect can be better than that After determined, reconstructed obtained, an band-pass filter designed utilizing entropy, further noise reduction processing carried out signals, analysis result good effect. According method, effectively extracted. applied simulation signals actual analysis, wide practicability.