基于EEMD和BP神经网络的短期光伏功率预测模型  

A Hybrid Model for Short-Term Photovoltaic Power Forecasting Based on EEMD-BP Combined Method

作　　者：于群[1] 朴在林[1] 胡博[1] 

机构地区：[1]沈阳农业大学信息与电气工程学院,辽宁沈阳110866

出　　处：《电网与清洁能源》2016年第7期132-137,共6页Power System and Clean Energy

基　　金：十二五国家科技支撑项目(2012BAJ26B00)~~

摘　　要：为了实现对并网型光伏电站调度,提出了一种基于集合经验模态能分解(EEMD)与BP神经网络的短期光伏出力的组合预测模型。利用集合经验模态分解将光伏出力序列分解,得到本征模函数分量IMF和剩余分量Res,降低序列的非平稳性。采用游程检验法优化因IMF分量数量多造成的建模过程复杂的问题,针对优化后的分量分别建立相应的BP神经网络预测模型。利用该方法对额定容量为40 k W的光伏系统进行预测,并与EMD-BP神经网络和传统的BP神经网络模型进行比较分析。结果表明,所提出的方法有效地提高了预测精度。In order to schedule the large-scale grid- connected PV generation, a combined forecasting model for grid-connected photovohaic generation system output power is proposed based on ensemble empirical mode decomposition （EEMD） and BP neural network. By using the EEMD of the set, the PV output sequence is decomposed, the intrinsic mode function component IMF and the residual component Res are obtained, and the non stationary of the sequence is reduced. The run test method optimization leads to complexity of the modeling process due to the huge number of IMF components, thus corresponding BP neural network prediction models are established respectively for the optimized components. By using this method, the PV system with rated capacity of 40 kW is predicted, and compared with the EMD-BP neural network and the traditional BP neural network model. The results show that the proposed method can effectively improve the prediction accuracy.

关 键 词：光伏功率预测 集合经验模态分解 BP神经网络 游程检验法 组合预测模型 

分 类 号：TM615[电气工程—电力系统及自动化]