Parameter estimation from power system disturbance measurements

Estimation of parameters from power system measurements is complicated by the nonlinear nature of system dynamics. The paper proposes a Gauss-Newton approach to computing model parameters that give the best fit between measurements and the model response. This approach involves trajectory sensitivities, i.e., the sensitivity of the model trajectory to small changes in parameters. An overview of trajectory sensitivity analysis is provided. A small example, which exhibits both soft and hard nonlinearities, is used to illustrate the estimation algorithm.