Statistical modeling of grid voltage fluctuations at wind turbine generator output by extreme value analysis

Wind turbine generators (WTGs) are required to operate at different levels of voltage uctuations in the power grid. When the uctuations exceed the standard levels, i.e. over and under voltage, they affect the performance of WTGs. In this study, we estimate the distribution of over and under voltages by utilizing Extreme Value Analysis. Assuming voltage as a random variable, we show that the over and under voltages can be considered as extreme deviations from the probability distribution median. The results approve that extreme fluctuations in the voltage of one WTG follow Generalized Extreme Value distribution. Since different WTGs do not necessarily experience the same extreme events, the model is generalized to the wind farm scale to cover all probable variations. This is accomplished by assuming the parameters of the generalized model as random variables. The generalized model will be used in the WTG test center, in order to determine WTG's capability to withstand the fluctuations.