Short-circuit test based maximum likelihood estimation of stability model of large generators

This paper presents a time-domain statistical identification method for synchronous-machine linear parameters from the standard line-to-line short-circuit test. The measurements are recorded on a 13.75-MVA hydrogenerator at Hydro-Quebec's Rapide-des-Quinze generating station. A complete mathematical model for synchronous machine asymmetrical test analysis is proposed. An efficient algorithm is built to accurately calculate the standard equivalent circuit from time-constants and operational inductances. The maximum likelihood estimator derived from the generalized least-squares method is then used for parameter identification. Validation of the estimated model response against the measured running-time domain data confirms the effectiveness of the proposed estimation technique.