Improving Active Power Regulation for Wind Turbines: A Data-Driven MPC Approach

This paper presents a data-driven model predictive control (DD-MPC) strategy to improve active power regulation (APR) for variable-speed, variable-pitch wind energy conversion systems (WECSs). The analysis focuses on key performance parameters, including generator output power, generator speed, and the rate of change of the pitch angle, all evaluated within the above-rated wind speed region. These parameters are essential for maintaining the stability, reliability and power quality of the WECS. To show the effectiveness of the DD-MPC strategy, the simulation results are compared with the conventional Multiple Model Predictive Control (MMPC) strategy. DD-MPC effectively minimizes power fluctuations, delivering a stable and consistent output-critical for preventing grid instability and ensuring superior power quality. Additionally, DD-MPC maintains rotor speed more accurately than MMPC, which is crucial for reducing mechanical stress on the wind turbine. The results also show that DD-MPC offers smoother and more precise pitch angle control, reducing actuator wear compared to the larger changes observed with MMPC. These results show that the DD-MPC has better performance compared to MMPC by addressing the dynamic challenges of WECS operation, providing a robust solution for APR and the long-term durability of WECS.