Equivalent Modeling of DFIG Wind Farm Based on PCA and Improved ABC-K Clustering Algorithm

In view of the shortcomings of the existing doubly-fed induction generator(DFIG) wind farm dynamic equivalent model, such as the accuracy is not high enough and the adaptability is not strong enough, a dynamic equivalent modeling method based on principal component analysis(PCA) and improved artificial bee colony(IABC) is presented. Firstly, by modeling and analyzing the doubly-fed wind turbine, all the state variables that can represent the wind turbine are obtained. Secondly, principal component analysis is used to deal with the redundancy and correlation between variables, and four dominant variables, namely stator q-axis current, rotor q-axis current, electromagnetic torque and generator speed, are extracted as clustering indexes. IABC is used to search for the best clustering center. Finally, the K-means algorithm is used to cluster the wind farms, and the simulation is carried out on the Matlab/Simulink platform. The simulation results show that the equivalent model has good accuracy and adaptability under wind speed disturbance and system side fault conditions.