超短期太阳辐照度多模型预测

Multi-model prediction for super short term solar irradiance

为了提高太阳辐射预测精度,提出了1种超短期混合预测模型。该模型用非线性方法集成多个子模型,从而克服了单一模型的局限性,提高了整体预测性能。预测时,首先将历史太阳辐射值转换为晴空系数,消除太阳位置对辐射值的影响;然后根据相关性分析,将转换的晴空系数输入各子模型,并将子模型的预测结果非线性集成。最后,将预测的晴空系数转换为目标水平总辐照度。实验结果表明:所提出的混合模型性能优于单一预测模型,非线性集成方法优于传统的平值加权集成方法,且整体预测均方差相对持续模型提高了约60%。

To improve the prediction accuracy of global horizontal irradiance(GHI),a hybrid model has been proposed.The model is a non linear combination of different sub models,overcomes limitations of each single model and improves the overall predictive performance.The historical GHIs are firstly transformed into the clear sky indexes to remove the impact of the sun's location,and then the clear sky indexes are set as inputs of forecasting sub models based on correlation analysis.Finally,the outputs of sub models are combined non linearly to achieve the forecasting clear sky index which is transformed into the forecasting target.The experimental results show that the performance of the proposed model is superior to that of signal sub model.The non linear combination does better than the traditional averaging integration method,and the performance of the proposed model is improved by about60%over the persistent model.