支持向量回归机在风电系统桨距角预测中的应用

Application of SVR for pitch angle forecasting in wind turbine

为解决风力发电系统中随着风速的变化,桨距角也随之发生不确定变化的问题,运用支持向量回归机算法对桨距角预测和仿真检验,并可将预测误差达到最小。该方法主要包括支持向量机中的回归分析技术,针对有限样本情况得到现有信息下的最优解。应用结果表明,此算法精度高,泛化能力强,可提高整个变桨距系统的控制精度和效率。

To solve the existing problems of pitch angle change uncertainty with change of wind speed in wind turbine, the SVR is studied to predict pitch angle and simulation test, and the calculate error is minimized in this paper. The methods include regression analysis technology that can find the best data by limited samples. The results show that the accurate and great generalization ability algorithm can improve the control accuracy and efficiency in pitch-controlled wind turbine.