基于小波包-神经网络的太阳逐时辐射预测

太阳辐射量受到季节、大气状况、云况、温度、湿度甚至沙尘等气象因素的影响,表现为强烈的时变性和随机性。对于非线性的辐射量预测,目前已提出了许多方法,但依然存在智能算法的选取不合理、网络结构泛化能力差、预测精度不理想等不足。针对光伏电站太阳逐时辐射强度数据特征不明显、普通BP网络难以完全映射其特征的缺点,提出了一种基于小波包-神经网络的预测模型(WPNN),利用小波包变换将辐射强度序列进行多尺度分解,并创建多个BP模型对各分量预测,最后通过重构得到最终的预测结果。结果表明,预测精度明显提高,满足预期效果,证明该模型的有效性和实际意义。

Study on Prediction of Solar Radiation Intensity Based on Wavelet Decomposition and BP Neural Network

The amount of solar radiation is affected by the season , atmospheric conditions , cloud conditions , tempera-ture , humidity and even dust and other weather factors with a strong time variability and randomness .For the predic-tion method of nonlinear radiation , many methods are put forward currently .However , the following problems are still presented , such as the intelligent algorithm selecting is unreasonable , network structure generalization ability is poor and prediction accuracy is not ideal .In view of the deficiency of that a photovoltaic power station solar radiation inten-sity of the original hourly data is not obvious and the normal BP neural network can not be completely mapped its fea -tures,this paper puts forward a prediction model based on Wavelet Packet Neural Network ( WPNN) .It used wavelet packet to transform the radiation intensity sequence multi-scale decomposition and established several BP neural net-work models to forecast each frequency components , and obtaining the complete prediction value with the wavelet packet reconstruction finally .The results show that the prediction accuracy is significantly improved to meet the expec-ted results , which demonstrates the effectiveness and practical value of the model .