Correlation based mode shape determination of a power system

In this paper, the concept of identification of modal frequencies and mode shape on the basis of cross-correlation (autocorrelation) is introduced. This method uses the cross power spectral density (auto power spectral) of the appropriate time domain signals, FFT and curve fitting in the frequency domain. It is found that a disturbance in power systems as well as customer load variations can be used to identify the mode shape and the contribution of power systems controllers in oscillatory modes. The disturbance is modeled by an impulse and the customer load variation is considered by white noise. The method is applied to a simulated network and the resonant frequency, damping and modes shape are identified. The results of the simulation validate the algorithm.