Adaptive Virtual Inertial Control of Wind Turbine Generators Considering Wind Speed and Load Variation

With the increasing proportion of wind power in the power grid, virtual inertia control has become one of the most important means for a wind turbine generator (WTG) to participate in frequency regulation. However, a single virtual inertia coefficient selected through engineering experience is difficult to apply in different conditions, which may result in a poor frequency modulation effect and even WTG instability. Considering the complex relationship between the virtual inertia coefficient setting and many factors, which are difficult to analyze and describe, this paper proposes a novel virtual inertia control, whose parameter set adaptively with BP neural network. In this method, the average wind speed and rate of change of frequency (ROCOF) are input variables of the neural network, the virtual inertia coefficient is output variable, and a mass of simulation results are used as training samples. Therefore, the selected virtual inertia coefficient can consider not only the frequency modulation capability of the WTG determined by the wind speed, but also the load change reflected by ROCOF. Finally, the effectiveness and superiority of this method are verified on MATLAB and DIgSILENT.