基于长短期记忆网络的风电场发电功率超短期预测

发布时间：2018-07-26 17:12

【摘要】：精确的风电场发电功率超短期预测对电力系统的经济调度和安全稳定运行意义重大。为充分利用多数据源中的有效信息来进一步提高风电场超短期发电功率的预测精度,提出了一种基于长短期记忆(long short-term memory,LSTM)网络的多变量风电场超短期发电功率预测方法。该方法首先利用距离分析法筛选出与风电功率相关程度高的变量,进而降低数据的规模和复杂程度。然后利用LSTM网络对多变量时间序列进行动态时间建模,最终实现对风电功率的预测。采用美国加州某风电场的实测数据进行验证,结果表明,文中方法能够有效利用多变量时间序列进行风电场发电功率的超短期预测,较人工神经网络和支持向量机拥有更高的预测精度。
[Abstract]:The accurate short-term prediction of wind power generation power is of great significance to the economic dispatch and safe and stable operation of the power system. In order to make full use of the effective information in the multi data source to further improve the prediction accuracy of the ultra short term power generation power of the wind farm, a kind of long short-term memory (LSTM) network based on the long and short period memory is proposed. This method first uses distance analysis method to select variables with high correlation with wind power, and then reduces the scale and complexity of the data, then uses LSTM network to model the dynamic time series of multivariable time series, and finally realizes the prediction of wind power. The measured data of a wind farm in the state are verified. The results show that the method can effectively use the multivariable time series to predict the power generation power of the wind farm effectively, and has a higher prediction accuracy than the artificial neural network and the support vector machine.
【作者单位】：强电磁工程与新技术国家重点实验室(华中科技大学);国网湖南省电力公司经济技术研究院;国网湖北省电力公司经济技术研究院;
【基金】：国家重点研发计划项目(2016YFB0900100)~~
【分类号】：TM614

【相似文献】