Denoising of test data of sudden three-phase short based on wavelet algorithm

In order to eliminate noise and singularities from dynamical test data, this paper employs spatial correlation of wavelet coefficients to remove noise by the feature, which noise-decomposed coefficients using wavelet transform are not correlative at the different finest scales, and uses the characteristic of the singularities through decomposing wavelet to eliminate these singularities, that is denoising and deleting singularities in one step using wavelet algorithm. And the parameters of a 650 MW turbo- generator are identified accurately by LM algorithm based on improved models.