A Method for on-Line Identification of Power System Model Parameters in the Presence of Noise

This paper describes a method for obtaining the parameters of power system controllers and generators while these devices are operating on-line or disconnected from the system. The technique has been developed to work in the presence of random disturbances acting on the system being studied. Wide-band noise is used as an excitation signal on the system and this enables the analyst to obtain a good estimate of time constants with magnitudes that are widely separated, Also the least-squares- parameter-estimator is quite stable and, for the examples used in this paper, convergence to within approximately 2% was obtained with initial estimates in excess of 400% of the true values.