摘要
由于风电出力具有随机性、波动性的特点,风电功率预测技术的研究对电力系统的安全稳定运行具有重要意义。提出一种基于PCC-GRU-Attention组合风电功率超短期预测方法,首先使用皮尔逊相关系数(pearson correlation coefficient,PCC)对数据预处理,选出高度相关性的特征作为输入,针对长短期记忆(long short term memory,LSTM)网络与门控循环单元(gated recurrent unite,GRU)网络处理长时序列易丢失序列信息的问题,通过GRU网络处理时间序列信息以及注意力(Attention)机制优化输出权重,与单一LSTM网络和GRU网络以及其它组合模型相比,有效提高了预测精度。
Due to the randomness and volatility of wind power output,the research on wind power prediction technology is significant to the safe and stable operation of the power system.A method of ultra short term forecast of wind power based on PCC-GRU-Attention combination is proposed.Firstly,Pearson correlation coefficient(PCC)is used to preprocess the data,and the features with significant correlation are selected as input.Secondly,an ultra-short-term wind power forecasting method based on PCC-GRU-Attention combined model for the current problem that long short term memory(LSTM)network and gated recurrent unit(GRU)network is easy to lose sequence information when dealing with long-term series is proposed.The GRU network processes time series information and the attention mechanism optimizes the output weight,comparing with a single LSTM network,a GRU network and other combined models,the prediction accuracy is effectively improved.
作者
陈德余
张玮
王辉
房栋
CHEN De-yu;ZHANG Wei;WANG Hui;FANG Dong(School of Information and Automation Engineering,Qilu University of Technology(Shandong Academy of Sciences),Jinan 250353,China;School of Electrical Engineering,Shandong University,Jinan 250061,China;Jinan Water Conservancy Engineering Service Center,Jinan 250013,China)
出处
《齐鲁工业大学学报》
CAS
2022年第6期1-8,共8页
Journal of Qilu University of Technology
基金
国家自然科学基金(2018YFE0208400)。
