Flexible Wind Farm Power Distribution Method Considering Noise and Communication Delays

The effectiveness of wind power allocation is of importance for the safety and reliability of power systems. In this study, a flexible scheduling method based on the non-dominated sorting genetic algorithm (NSGA) is proposed, which dynamically adjusts the power reference values of wind turbines to avoid prolonged high-load operation of certain turbines. It can significantly improve the power allocation efficiency and reduce cumulative fatigue damage in wind turbine which could extend its service life and lower the maintenance costs. The three-point rain flow counting method is utilized to calculate the thrust-induced fatigue damage of 100 sets of wind turbine towers. By estimating turbine thrust and main shaft torque in real-time through power equations, the optimized result of power allocation can remain stable under noise and communication delays. The numerical results show that after the multi-objective optimization of wind power distribution, the cumulative damage of 65% of main shaft and 55% of tower are effectively reduced. It can help to reduce the fatigue damage of wind turbine and maintain the stable operation of the large-scale wind farm.