Identification of the standard parameters of a steam turbine-generator using wavelet denoising

Synchronous generator measurements obtained on-line from instrumentation at the machine terminals are contaminated by noise due to internal or external factors. A technique to remove the noise component of the measurements through decomposition, thresholding, and reconstruction of the signals is applied to actual measurements obtained from a steam turbine-generator. Multiple data sets at different operating points are used to ensure the applicability of the denoising technique and the accuracy of the estimation algorithm. These data can be used for the identification of machine parameters. The estimated parameters are compared to available manufacturer data and the accuracy of the method is assessed. The estimator is shown to have a superior performance as compared to estimators using traditional frequency based filtering techniques.