Improving active power regulation for wind turbine by phase leading cascaded error-based active disturbance rejection control and multi-objective optimization

With the escalating global demand for renewable energy, the active coordinated control of wind turbine is poised to become a crucial factor in ensuring the stable operation of new power system. However, existing coordinated control strategies for permanent magnet wind turbine remain inadequate in addressing the coupling effects between torque control and variable pitch control. These strategies require further development to enhance their effectiveness in practical applications. In response to this challenge, a phase leading cascaded error-based active disturbance rejection control and multi-objective optimization strategy are proposed to determine reference signals for pitch angle and torque, facilitating rapid and stable power command tracking. Firstly, the significant phase lag issue inherent in traditional extended state observer is examined. To improve the precision of system perturbation estimation, a phase leading cascaded error-based active disturbance rejection controller is designed, with its stability is theoretically proven. Secondly, an enhanced snow ablation optimization algorithm is utilized to identify the optimal solution for controller parameters, balancing power tracking accuracy with fatigue load mitigation. Additionally, to address the challenge of calculating fatigue loads during wind turbine operation, a data-driven fatigue modelling method based on bidirectional long and short-term memory is proposed, enabling real-time estimation of fatigue loads. Finally, a simulation model of a 5 MW wind turbine is used to validate the effectiveness of the presented strategy. Experimental results show that the proposed strategy can effectively perform power regulation tasks under three scenarios: power command tracking, actuator fault and model mismatch, while minimizing tracking error and reducing fatigue loads.