基于SVM回归方法的风速预测研究

Wind Speed Forecasting Based on SVM-regression Method

根据气象评价和大量的实地调查,以及风速对于风力发电的决定性作用,在简要介绍SVM回归方法的基础上,对RBF核函数的参数C和g进行参数寻优,建立预测模型,以误差百分比(APE)、平均百分比误差(MAPE)和平均绝对误差(MAE)来评价,其中平均百分比误差(MAPE)为1.705%,平均绝对误差(MAE)为0.1705较为理性,可以为风电场风速预测提供较好的参考。

On the basis of Meteorological evaluation and much on-the-spot investigation, wind speed plays a decisive role in wind power generation.The papcr optimizes the parameters of tt.BF： kernel parameter C and g based on the Brief introduction of the SVM regression method, then, a forecasting model is founded and gets a better performance by taking average percentage error（APlZ）, mean average percentage error（MAPE）, mean absolute error（MAE）figr evaluating and MAPE is 1.705%, MAE is 0.1705.The method has important value for wind speed forecastinK for wind power farm.