| 引用本文: | 王娟娟,赵闻蕾,王兴强,靳小钊.基于Johnson分布直接转换法的风速预测[J].电力自动化设备,2014,34(6): |
| WANG Juanjuan,ZHAO Wenlei,WANG Xingqiang,JIN Xiaozhao.Wind speed prediction based on Johnson direct transformation[J].Electric Power Automation Equipment,2014,34(6): |
|
| 摘要: |
| 大部分风速呈现非正态分布特性，因此时间序列法不能准确预测风速。提出一种基于Johnson分布直接转换法的风速预测方法，首先采用时间序列法对平稳的风速数据进行初步预测；然后根据风速数据本身所具有的数字特征，构建合适的Johnson转换曲线；最后应用Johnson转换曲线对初步预测结果进行修正。实例分析结果表明，时间序列法的预测结果经由Johnson变换曲线修正之后更加接近具有非正态分布特性的原始数据，预测精度更高。 |
| 关键词: 风电 Johnson分布直接转换 时间序列法 风速预测 非正态分布 数据转换 |
| DOI： |
| 分类号: |
| 基金项目:大连市科学技术基金资助项目（2011J21dw007） |
|
| Wind speed prediction based on Johnson direct transformation |
|
WANG Juanjuan, ZHAO Wenlei, WANG Xingqiang, JIN Xiaozhao
|
|
Dalian Jiaotong University，Dalian 116028，China
|
| Abstract: |
| Since the wind speed obeys abnormal distribution in most conditions，it cannot be precisely predicted by the time series method. A method of wind speed prediction is proposed based on the Johnson direct transformation，which applies the time series method to predict the preliminary wind speeds based on the stable wind speed data，constructs a suitable Johnson transformation curve according to the inherent digital features of original wind speed data，and then corrects the preliminary wind speeds by the constructed Johnson transformation curve. The analytical results of an actual case show that，with higher prediction accuracy，the distribution of the wind speeds predicted by the time series method and corrected by the Johnson transformation curve is much closer to the abnormal distribution of original wind speed data. |
| Key words: wind power Johnson direct transformation time series method wind speed prediction abnormal distribution data transformation |
