基于数据分解与重构的光伏发电功率超短期预测

为进一步提高光伏发电超短期预测的精度,以数据分解重构和深度学习技术为依托,提出一种基于 CEEMDAN-DBN-Seq2Seq 的光伏发电功率超短期预测方法。首先利用具有自适应噪声的完整经验模态分解算法( CEEMDAN )将原始发电数据分解成在时域内特征更加明显的模态函数序列,以提取发电序列在时间尺度上的特征;随后引入影响光伏出力的主要气象因素,利用深度信念网络( DBN )对重构后的高、低频分量和序列对序列( Seq2Seq )方法对残差分量进行预测。实验表明,所提模型在光伏发电预测研究中精确度更高。

Ultra-short-term Prediction of Photovoltaic Power Generation Based on Data Decomposition and Deconstruction

In order to further improve the accuracy of ultra-short-term photovoltaic power generation prediction , based on data decomposition and reconstruction and deep learning technology , an ultra short-term photovoltaic power prediction method based on CEEMDAN-DBN-Seq2Seq was proposed. Firstly , Complete Ensemble Empirical Mode Decomposition with Adaptive Noise( CEEMDAN ) is use to decomposes the original generation data into a series of modal functions with more obvious characteristics in the time domain to extract the characteristics of the generation sequence on the time scale.Then , Deep Belief Network( DBN ) and Sequence to Sequence( Seq2Seq ) methods were used to predict the reconstructed modal components respectively.The experiments show that the proposed model has higher accuracy in the prediction research of photovoltaic power generation.