A methodology for parameter estimation of synchronous generators based on trajectory sensitivity and synchronization technique

In this paper, a new approach combining synchronization techniques and trajectory sensitivity analysis is proposed to estimate the parameters of electrical synchronous generator models. The main contribution of this combination is the improvement on the numerical robustness of the estimation algorithm. The synchronization technique consists of coupling the real system with the mathematical model. The coupling is of unilateral type, that is, some outputs of the real system are considered inputs of the mathematical model. With a convenient choice of coupling, parameters that would have difficulties to be estimated by the traditional approach (due in general to very low sensitivity) can be more easily estimated. The proposed methodology is appropriate for on-line applications and does not require special operating conditions. Indeed, it can naturally deal with generator model nonlinearities. The methodology was successfully applied to estimate the parameters of a two-axis synchronous generator model. Synchronization improved the numerical robustness of the estimation algorithm allowing the simultaneous estimation of the d and q axis parameters even for bad initial guesses of parameters