基于SVM方法的风电场短期风速预测

针对基于支持向量机的风电场短期风速预测进行研究．选择了不同的输入向量（历史风速时间序列，历史风速和温度．历史风速、温度和风向，历史风速、温度和时间）作为输入进行误差对比分析。实测数据及分析结果表明，采用历史风度和温度的二输入模型，预测效果最佳，为风速的短期预测和发电量预测提供了较好的参考价值。

Short-Term Wind Speed Forecasting of Wind Farm Based on SVM MethodPENG Huai-wu1, YANG Xiao-feng1, LIU Fang-rui2

In this paper, support vector machine (SVM) method is employed for the short term wind speed forcasting of wind farm. Various input vectors of SVM were generated and compared through error measures to guarantee the performance and accuracy of the chosen models. First a model with only historical wind speed data was chosen according to the traditional way. Nevertheless, the results were not sufficiently satisfactory. Therefore, three models, consisting of historical wind speed data and temperature, historical wind speed data, temperature and wind direction, historical wind speed data, temperature and time, were developed. The simplest model of two inputs with wind speed data and temperature, was the optimal for the short term wind speed forecasting. The developed model provides an alternative for short term wind speed forecasting with high pricision.