A new approach for fast identification of distribution system dynamic model from measured data

Study of distribution system´s dynamic response from the perspective of M.V and H. V networks, especially in presence of DGs is one of the important issues in integrating distributed generation. This paper presents the modeling of distribution system dynamics during voltage drops, from the data provided by phasor measurement units, through a seventh order induction motor and a ZIP load model and initializes the dynamic parameters of this model using a linear state space model. Using subspace methods for estimation of linear state space model, the initial guess for nonlinear state space model parameters is obtained and then the model parameters are tuned for the closest possible resemblance to the actual measured response. This method is verified on IEEE 37 bus distribution test system and the results shows the effectiveness of this approach.