Optimized Active Power Dispatching of Wind Farms Considering Data Driven Fatigue Load Suppression

The active power fluctuation of wind turbines is not only related to their friendliness to the grid, but also to their fatigue damage. In this paper, the active power of wind turbines in wind farms is optimally scheduled to achieve the suppression of fatigue load of wind turbines. Considering the complexity of fatigue load calculation, it is difficult to apply to real-time active scheduling using metrics that directly characterize fatigue load. To address this problem, a data-driven modeling method for wind turbine fatigue based on deep neural network (DNN) is proposed in this paper, and the relationship between wind speed, power and other easily measurable parameters and fatigue load is established. Further, an improved multi-objective grey wolf optimizer (MOGWO) is designed to achieve the wind farm active scheduling process with the data-driven fatigue calculation results as the optimization objective. The results show that: The fatigue load prediction model of data-driven fatigue calculation proposed in this paper has a satisfactory effect, and the fatigue load of wind turbines can be effectively reduced by adjusting the active power.