Assessment of model parameters to identify an equivalent wind power plant

An assessment of parameter observability to identify an equivalent wind power plant model is proposed in this paper. The trajectory sensitivity method is used as an updating rule to iteratively improve the estimation of the model parameters. The difficulties for obtaining the correct estimation are analyzed and a strategy to overcome those problems are highlighted. In general those problems are related to parameter divergence during the estimation process, specially when the initial parameter guesses are far away from the true values and when the sensitivities of the parameters with relation to the model output are low. Those problems become more evident as the number of parameter to be simultaneously estimated increases. The proposed strategy splits the estimation procedure into different stages of different window time, facilitating the estimation of parameters. The robustness of the proposed method with relation to initial parameter guesses is verified by estimating parameters in different scenarios and with the addition of noise in the measurements.