Active power and reactive power dispatch of wind farm based on wavelet learning

During normal operation, the doubly-fed induction generator (DFIG) generates certain range of reactive power. The DFIG based wind farm can participate in reactive power control of grid as a reactive power supply. In order to get a more stable input wind speed of the DFIG, wavelet multi-resolution analysis method is used. This paper proposes a kind of power dispatch model which considers a learning mechanism of minimum copper loss of all DFIGs in wind farm as an objective function. An active power and reactive power allocation optimization model is established. This power dispatch model makes the working condition of DFIGs and the PCC running in the optimum state. The active power and reactive power generated by wind farm satisfy the power gird requirements of both active power and reactive power. The advantage of the proposed method is verified by a case study which successfully demonstrates the learning mechanism.