Research on Theoretical Power of Wind Farm Based on Wind Turbine Grouping Method

The research on theoretical power of wind farm is significant to evaluate curtailment of renewable energy. Therefore, we propose a new model to calculate theoretical power of wind farm called wind turbine grouping method, Firstly, to avoid measurement deviation, the abnormal data are eliminated by multi-step k-clustering algorithm. Secondly, the wind turbines are divided into two groups by calculating the variance inflation factor of wind speed, namely, strong correlative wind turbine group and weak correlative wind turbine group, and the multiple mutual linear problem of the wind speed then gets a better solution. Finally, two artificial neural nets are built respectively for two wind turbine groups, one is dependent on median wind speed of the group and the other is dependent on all the wind speed series of the group. A simulation is performed based on the actual output of a certain wind farm in the northwest of China. The test results show that the new model can make a superior performance compared with other traditional methods. The average absolute deviation of theoretical power calculated by the new model is less than the rated capacity of a single wind turbine, the correlation coefficient is close to 0.98, and the average power error is 0.47%. © 2019, High Voltage Engineering Editorial Department of CEPRI. All right reserved.