Journal of Shanghai Jiao Tong University ›› 2021, Vol. 55 ›› Issue (S2): 42-50.doi: 10.16183/j.cnki.jsjtu.2021.S2.007

Previous Articles     Next Articles

Optimization of DFIG Comprehensive Adaptive Frequency Regulation Parameters Based on Extreme Learning Machine

JIN Haochun1, GE Minhui1, XU Bo2()   

  1. 1. East Branch of State Grid Corporation of China, Shanghai 200120, China
    2. School of Electric Power Engineering, Shanghai University of Electric Power, Shanghai 200090, China
  • Received:2021-10-20 Online:2021-12-28 Published:2022-01-24
  • Contact: XU Bo


Aiming at the problem of insufficient system frequency regulation ability caused by wind turbines connected to the grid, doubly fed induction generators (DFIG) mostly use virtual inertia and droop control to participate the frequency regualtion of the power system. However, traditional control strategies cannot fully utilize the frequency regulation capability of DFIG. In order to further improve the frequency stability of the system, the adaptive control of the virtual inertia and the droop coefficient are realized by analyzing the effects of the virtual inertia and the droop coefficient in each stage of the frequency dynamic response. Then, based on the extreme learning machine to predict the various frequency regulation index at different levels of wind speed, the objective function of the frequency regulation index is established to achieve the optimization of the comprehensive adaptive frequency regulation parameters, and the variable load shedding rate active standby control scheme adapted to the wind speed is proposed. The simulation results show the effectiveness of the method.

Key words: doubly fed induction generator (DFIG), extreme learning machine, integrated adaptive control, frequency dynamic response

CLC Number: