基于支持向量机权系数的独立变桨距学习控制

Individual Pitch Learning Control Based on Support Vector Machine Weight Coefficient

  • 摘要: 由于风速的随机性和不均衡性易造成高风速段机组输出功率波动,同时给整个风电机组带来挥舞震动等不平衡载荷. 为降低疲劳载荷和改进系统性能,在分析风力发电系统恒功率运行区域内载荷动态模型的基础上,将支持向量机和权系数控制相结合,提出一种基于在线支持向量机学习的权系数的独立变桨距学习控制方案. 然后构建永磁直驱风力发电系统独立变桨距控制系统模型并在RT-LAB实时仿真系统上进行仿真. 仿真结果表明该方法在保证稳定功率输出的同时,实现了桨距角的平稳调节,降低了不平衡载荷和减轻了机组疲劳度和组件间的磨损,从而验证了该控制策略的正确性和有效性.

     

    Abstract: Because of the random and irregular nature of wind speed, the output power generated by wind turbines can be unsteady in high-speed regions; thus, an entire wind turbine experiences unbalanced loads and wave vibration. To mitigate load fatigue and improve system performance, we propose an individual pitch control strategy for wind turbines based on an analysis of dynamic load models of wind turbine systems when operating at constant power combined with a support vector machine with weight coefficient control. We then developed an individual pitch control system mode for wind turbine systems with a permanent magnet synchronous generator and simulated it on an RT-LAB real-time simulation system. The simulation results show that the proposed control strategy ensures a stable power output, achieves a smooth regulation of the pitch angle, and reduces the unbalanced load as well as the fatigue and wear on components. Thus, the correctness and effectiveness of the proposed control strategy are verified.

     

/

返回文章
返回