基于单神经元PID的双馈风力发电机并网控制

Grid-connected Control of Doubly-fed Wind Power Generator Based on Single Neuron PID Controller

针对常规双馈风力发电机(DFIG)空载并网控制系统采用传统比例-积分(PI)控制器,系统响应慢,鲁棒性差的不足,将单神经元PID控制器与矢量控制相结合,提出基于定子磁链定向的空载并网控制策略。单神经元PID控制器可通过改变自身权重进行自学习、自组织,具备良好的鲁棒性和自适应性。在1.5 MW DFIG实验平台进行了实验,结果表明所述空载并网控制系统可实现柔性并网,具有良好的动态和稳态性能。

In the conventional no-load grid-connected control scheme of doubly-fed wind power generator （DFIG）, us- ing proportional-integral （PI） controller has deficiencies of slow system response and poor robustness.In order to get a good performance in no-load grid-connected control, a single neuron PID controller is presented to take place of the traditional PI controller.And no-load grid-connected control system of DFIG is designed based on the stator flux ori- entation vector control technique.Single neuron PID controller can be self-learning, self-organized by changing their own weight,and it has good robustness and adaptability.Experiments on the platform of 1.5 MW DFIG system are made to verify the proposed control strategy.The experimental results prove no surge current grid-connected and verify its good dynamic and steady state performance.