基于EEMD-SVR模型的光伏功率预测

Photovoltaic Power Prediction Based on EEMD-SVR Model

受天气因素的影响,光伏出力具有较强的不确定性和波动性,导致光伏功率的预测难度较大。针对光伏功率的预测问题,提出一种集合经验模态分解(EEMD)和支持向量回归(SVR)模型的预测方法。首先,对光伏功率进行集合经验模态分解,将光伏功率信号分解为多个特征模态分量以及一个残余分量;其次,使用SVR模型训练数据,实现分量预测;最后,结合预测分量预测光伏功率。结果表明,该模型可以实现对非平稳序列的可靠分解,并能有效提高光伏预测的平均绝对百分比误差(MAPE)性能。

Due to weather factors,photovoltaic output has strong uncertainty and volatility,making it difficult to predict photovoltaic power.A prediction method based on ensemble empirical mode decomposition(EEMD) and support vector regression(SVR) models is proposed for the prediction of photovoltaic power.Firstly,the photovoltaic power is decomposed into multiple characteristic mode components and a residual component through ensemble empirical mode decomposition.Secondly,use SVR model to train data and achieve component prediction.Finally,predict the photovoltaic power by combining the predicted components.The results indicate that the model can achieve reliable decomposition of non-stationary sequences and effectively improve the average absolute percentage error(MAPE) performance of photovoltaic prediction.