电网技术2017，Vol.41Issue(12)：3797-3802,6.DOI:10.13335/j.1000-3673.pst.2017.1657
基于长短期记忆网络的风电场发电功率超短期预测
Short-Term Wind Power Forecasting Based on LSTM
摘要
Abstract
Accurate ultra-short-term wind power forecasting is essential to economic dispatching and secure operation of power system with large-scale wind power integrated. To further enhance accuracy of ultra-short-term wind power forecasting, a multivariate method for ultra-short-term wind power forecasting based on long short-term memory (LSTM) was presented in this paper. It sifted multivariate time series highly relevant to wind power with distance analysis, and modeled the time series from the viewpoint of time with LSTM. The real-world data collected from a wind farm in California was applied to verify the conclusions. Results show that the proposed method can forecast wind power using multivariate time series, and outperform artificial neural network and support vector machine with respect to forecasting accuracy.关键词
风电功率预测/长短期记忆/多变量时间序列/距离分析法Key words
wind power forecasting/long short-term memory/multivariate time series/distance analysis分类
信息技术与安全科学引用本文复制引用
朱乔木,李弘毅,王子琪,陈金富,王博..基于长短期记忆网络的风电场发电功率超短期预测[J].电网技术,2017,41(12):3797-3802,6.基金项目
国家重点研发计划项目(2016YFB0900100).Project Supported by The National Key Research and Development Program (2016YFB0900100). （2016YFB0900100）
