摘要
研究了风电场超短期风电功率预测问题,提出一种基于最佳预测步长的超短期风电功率预测模型。为减小模型原始输入数据所导致的预测误差,对原始输入数据进行预处理。通过对比分析基于不同预测步长的超短期风电功率预测模型得到超短期功率预测结果,采用均方根误差作为选择模型最佳预测步长的判据,确定最佳预测步长,建立基于最佳预测步长的超短期风电功率预测模型。利用陕西地区某风电场2014年10~11月的实测功率和短期预测功率数据进行建模和验证,结果验证了该方法的有效性。
Prediction on ultra-short-term wind power of wind power field was studied and a kind of prediction model for ultra-short-term wind power based on optimal predictive time length was proposed.In order to reduce prediction error caused by original input data of the model,pretreatment for original data was conducted.By means of comparing and analyzing pre-diction model for ultra-short-term wind power based on different predictive time length,prediction result of ultra-short-term power was obtained.Root mean square error was used for selecting and determining optimal predictive time length of the model as well as building a prediction model for ultra-short-term wind power based on optimal predictive time length.On the basis of data of measured power and short-term predictive power from October to November in 2014 of some wind power field in Shaanxi region,modeling and verification was carried on and results verified effectiveness of this method.
出处
《广东电力》
2015年第8期19-22,54,共5页
Guangdong Electric Power
基金
国家电网公司科技项目"基于集群划分的新能源功率预测技术研究和示范"资助
关键词
风电功率
超短期预测
预测步长
预测模型
均方根误差
wind power
ultra-short-term prediction
predictive time length
prediction model
root mean square error
