Fixed speed wind generator model parameter estimation using improved particle swarm optimization and system frequency disturbances

When planning power system operation it is important to have reliable models of the elements of the power system. Fixed speed wind turbines are a widely installed generation technology that use a single squirrel cage induction generator. The local wind profile and the properties of the induction machine constitute the main considerations when modeling these wind turbines. Existing methods for estimating the parameter values of induction machine models use a wide variety of parameter estimation algorithms but primarily use active and reactive power measurements made during start-up or direct mechanical testing to fit the model to. Proposed here is a parameter estimation method that applies improved particle swarm optimization to active and reactive power measurements made during a deviation in system frequency to estimate the parameter values of a induction machine model. This method has shown good accuracy and the use of on-line data may prove beneficial in future applications.