摘要
光伏发电出力与太阳辐射强度和温度等气象条件关系密切,在出力预测模型中,气象因素与出力数据存在着非线性关系,同时各气象因素之间也存在非线性的关系。核偏最小二乘(KPLS)方法具有较强的处理非线性问题的能力,本文基于KPLS方法建立光伏发电出力预测模型。利用光伏电站的历史气象数据和出力数据对KPLS预测模型进行训练,训练的模型用以预测光伏电站的出力。通过实际光伏电站运行数据的验证,并将预测结果同偏最小二乘(PLS)方法和人工神经网络(ANN)方法的预测结果相比较,实验结果显示KPLS预测模型具有较准确的预测能力和较强的适用性。
The output ofphotovoltaic （PV） power is closely related to the solar radiation and ambient temperature. For the output forecasting model, there is a nonlinear relationship between meteorological data and output data. Also, this nonlinear relationship exists among the meteorological variables. Kernel partial least squares （KPLS） can handle the nonlinear problem when the modeling data contain nonlinear information, so this paper proposes an output forecasting model for PV power based on KPLS. This model is built based on the historical meteorological data and output data, and then it is used to forecast the PV power output. The perfbrmance of the KPLS model is validated through actual operation data in PV power station and the forecasting data are compared with that of the partial least squares （PLS） mode/and the artificial neural network （ANN） model. These results show that the KPLS model is more accurate and has better applicability.
作者
胡益
HU Yi(East China Electric Power Design Institute Co. Ltd., Shanghai 200063, Chin)
出处
《电力勘测设计》
2018年第1期70-74,共5页
Electric Power Survey & Design
