To address the issues of poor dynamic response and insufficient adaptability of controller parameters leading to significant output power fluctuations when traditional wind turbine pitch angle control strategies are confronted with wind speed variations
an improved linear active disturbance rejection pitch angle control strategy based on Deep Deterministic Policy Gradient (DDPG) algorithm is proposed. The proposed strategy introduces an additional free expansion dimension state variable on top of the linear extended state observer (LESO) and improved the parameters of the expanded system in a proportional-derivative form to enhance the capability of disturbance feedforward compensation. Then
an appropriate reward function is designed according to the generator speed error
and the improved linear active disturbance rejection control (LADRC) parameters can be adjusted adaptively by using DDPG algorithm
so as to achieve the optimal control effect. The simulation results show that the proposed strategy can make the pitch angle quickly adapt to the changes in wind speed and effectively deal with the violent fluctuation of wind speed
therefore maintaining the stable operation of wind turbine and the efficient output of electric energy.
GAO Z Q.Scaling and bandwidth-parameterization based controller tuning[C]//Proceedings of the 2003 American Control Conference. Denver, CO, USA, 2003: 4989-4996.
JAFARNEJADSANI H, PIEPER J, EHLERS J.Adaptive control of a variable-speed variable-pitch wind turbine using radial-basis function neural network[J]. IEEE transactions on control systems technology, 2013, 21(6): 2264-2272.
LIANG Y C, GUO C L, DING Z H, et al.Agent-based modeling in electricity market using deep deterministic policy gradient algorithm[J]. IEEE transactions on power systems, 2020, 35(6): 4180-4192.
SEHGAL A, LA H, LOUIS S, et al.Deep reinforcement learning using genetic algorithm for parameter optimization[C]//2019 Third IEEE International Conference on Robotic Computing (IRC). Naples, Italy, 2019: 596-601.
SHI Q, YING W D, LYU L, et al.Deep reinforcement learning-based attitude motion control for humanoid robots with stability constraints[J]. Industrial Robot: the international journal of robotics research and application, 2020, 47(3): 335-347.