Load Forecast-Based Power Reserve Control Strategy for Primary Frequency Regulation of Variable Speed Wind Power Generators

The large-scale application of variable-speed wind power generators (VSWPGs) has introduced stochastic power fluctuations to power systems and reduced the system inertia, thus it is an urgent requirement that VSWPGs should also regulate frequency in high-penetration renewable energy power systems. VSWPGs can operate under the derating state during frequency regulation through power reserve control (PRC). However, the efficiency of the VSWPGs in PRC is reduced compared with the maximum power point tracking (MPPT) mode during normal operation, because the conventional PRC strategy is immediate regulation. Therefore, this paper proposes a novel load forecast-based power reserve control (FB-PRC) strategy of VSWPGs for frequency regulation, which adjusts wind power generation in advance based on load forecasting results. The long short-term memory (LSTM) neural network is adopted to forecast the very short-term load fluctuations, and then the wind power generators operating in MPPT mode can reserve power for frequency regulation in advance according to the real-time load forecasting results, which thus improves the frequency regulation performance and increases the operating efficiency of the wind power generators. Meanwhile, in order to take full advantage of the frequency regulation capability of VSWPGs, a novel frequency dead band response method specially adapted for the high-penetration renewable energy system is designed in this paper. Finally, the time-domain simulation results of the system frequency response model verify that the proposed strategy has better frequency dynamic performance.