摘要
为了解决风力发电和光伏发电随机性、波动性、间歇性造成的新能源功率预测建模和精度不高问题,基于深度学习模型变分自动编码器(Variational Auto⁃Encoder,VAE)在时间序列建模和非线性逼近特征提取方面的优异性能,开展新能源电站VAE模型功率短期预测研究,并与循环神经网络(RNN)、长短期记忆(LSTM)深度学习方法和支持向量回归(SVR)机器学习方法的预测结果进行了对比。光伏电站和风电场独立功率预测结果表明,深度学习模型较基线机器学习模型预测性能更好,基于VAE的预测方法能够学习更高级别的特征,其预测性能表现更佳,光伏功率预测模型的RMSE、MAE和R2值分别为1.593、1.098和0.973;风光一体化功率预测结果表明,VAE和RNN模型能够提高功率预测准确性,其一体化功率预测模型的R2值分别为0.96和0.97。
In order to solve the problem of new energy power prediction modeling and low precision caused by the randomness,volatility and intermittence of wind power generation and photovoltaic power generation,based on the excellent performance of variational auto⁃encoder(VAE)deep learning model in the aspect of time series modeling and nonlinear approximation feature extraction,the VAE model was used to carry out the short⁃term power prediction research of new energy power station.The prediction re⁃sult of VAE was compared with the prediction results of the recurrent neural network(RNN)and long short⁃term memory(LSTM)deep learning methods and the support vector regression(SVR)machine learning method.The independent power prediction results of photovoltaic power plants and wind farms show that the deep learning model has better prediction performance than the baseline machine learning model,and the VAE⁃based prediction method can learn higher⁃level features and its prediction perform⁃ance is better.The RMSE,MAE and R2 values of photovoltaic power prediction model are 1.593,1.098 and 0.973 respectively.The wind⁃solar integrated power prediction results show that the VAE and RNN models can improve the accuracy of power prediction,and R2 values of integrated power prediction model are 0.96 and 0.91 respectively.
作者
尹兆磊
刘嗣萃
于立强
毕圆圆
YIN Zhao-lei;LIU Si-cui;YU Li-qiang;BI Yuan-yuan(Chengde Power Supply Company,State Grid Jibei Electric Power Co.,Ltd.,Chengde,China,Post Code:067000)
出处
《热能动力工程》
CAS
CSCD
北大核心
2023年第6期137-146,共10页
Journal of Engineering for Thermal Energy and Power
基金
国家自然科学基金项目(71471070)。
关键词
一体化功率预测
深度学习
变分自动编码器
RNN
integrated power prediction
deep learning
variational auto⁃encoder(VAE)
recurrent neu⁃ral network(RNN)