Features of Singular Value Decomposition and Its Application to the Vibration Monitoring of Turboprop Engine

The singular value decomposition (SVD) can decompose an original signal into a series of component signals linearly. By means of analyzing deeply the fundamental principle and existing problems of Hankel matrix-based SVD, This paper reveals the three basic features of SVD, including linear decomposition, reconstruction component frequency domain disorder and band-pass filtering. Based on those features a new SVD method is put forward. Numerical simulation results show that the proposed method not only solve the frequency domain disorder problem of traditional SVD, and can achieve a given linear band-pass filter bandwidth, complete recovery of original signal amplitude, frequency and phase characteristics in any given frequency nearby. Other signal processing methods have no such advantages. The proposed method has been successfully applied to the vibration signal extraction of a certain type of turbofan engine, and the results show that the method has excellent in the feature extraction.