A new discrete spectrum correction method for ultra-low frequency signals with negative frequency contribution

The traditional analysis and correction method for discrete spectrum has greater error for the extreme frequency signal which widely existing in vibration engineering such as ultra low frequency signal, short time signal and ultra high frequency signal(close to the Nyquist frequency). In this paper we analyze the typical situation of extreme frequency signal and put forward a new discrete spectrum method based on accounting the influence of ultra low frequency. This method based on Blackman window sets up a discrete spectrum correction model involving the contribution of positive-negative frequency by utilizing the three spectral line near the local spectral peak. The frequency, amplitude and phase can be got by the solution of the model. The result coming from the simulation verification of the spectrum correction formula using frequency scanning indicates that this method efficiently lowers the influence of negative spectrum, and has higher precision for the frequency, amplitude and phase of the ultra low frequency.