Instantaneous feature extraction of non-stationary signal based on complex Morlet wavelet

Abstract

Put forward the method of Instantaneous feature extraction of non-stationary signal based on complex wavelet Morlet. Aiming at the characteristics of non-stationary signal signal, we construct two groups of combination information suitable for feature extraction combining amplitude and phase information which complex wavelet transform provide. The example of linear frequency-modulated signals and application examples of bearing vibration signal show that the model and the method have better antinoise ability, and have higher precision to the Instantaneous feature of non-stationary signal, and it can extract Instantaneous feature of non-stationary signal This method can realize its fast algorithm by Fourier transform, and has the simple and fast characteristics of algorithm, and realize the real-time analysis of the instantaneous characteristics of non-stationary signal