| 基于时序动态回归的超短期光伏发电功率预测方法 |
| |
| 引用本文： | 解振学,林帆,王若谷,张耀,高欣,王建学.基于时序动态回归的超短期光伏发电功率预测方法[J].陕西电力,2022,0(7):45-51. |
| |
| 作者姓名： | 解振学 林帆 王若谷 张耀 高欣 王建学 |
| |
| 作者单位： | （1.国网陕西物资公司，陕西西安 710054；2.陕西省智能电网重点实验室，西安交通大学陕西西安 710049；3.国网陕西电力科学研究院，陕西西安 710100） |
| |
| 摘 要： | 光伏发电功率预测对于电力系统安全可靠运行以及提高光伏发电产业经济效益具有重要意义。提出一种基于时序动态回归的超短期光伏发电功率预测方法，仅需要历史光伏发电功率数据与数值天气预报作为输入。首先建立光伏发电功率与地表太阳辐射累计值的回归模型，再建立ARIMA模型预测回归残差序列，最后引入傅里叶谐波序列刻画日季节性。根据线性形式与对数形式的回归公式提出两种预测模型，综合二者形成最终的混合预测方法。算例结果表明，与一般时序模型相比，该方法在超短期预测方面预测精度更高。
|
| 关 键 词： | 超短期预测 光伏发电预测 时序动态回归 ARIMA 谐波序列 |
Very Short-term Photovoltaic Power Forecasting Method Based on Time Series Dynamic Regression |
| |
| XIE Zhenxue,LIN Fan,WANG Ruogu,ZHANG Yao,GAO Xin,WANG Jianxue.Very Short-term Photovoltaic Power Forecasting Method Based on Time Series Dynamic Regression[J].Shanxi Electric Power,2022,0(7):45-51. |
| |
| Authors: | XIE Zhenxue LIN Fan WANG Ruogu ZHANG Yao GAO Xin WANG Jianxue |
| |
| Affiliation: | （1. State Grid Shaanxi Electric Power Materials Company，Xi’an 710054，China； 2. Shaanxi Province Key Laboratory of Smart Grid，Xi’an Jiaotong University，Xi’an 710049，China； 3. State Grid Shaanxi Electric Power Research Institute， Xi’an 710100， China） |
| |
| Abstract: | Photovoltaic（PV） power forecasting is of great significance for power system reliability and economic benefits of PV industry. A very short-term PV power forecasting method based on time-series dynamic regression model is proposed. Only historical PV power output and numerical weather prediction （NWP） are needed as input data. A regression model is firstly established between PV power output and surface solar radiation （SSRD）. ARIMA model is used to forecast the regression residual term. Finally， Fourier harmonic time-series is introduced to characterize the daily seasonality of PV power output. Two concrete forecasting models are proposed respectively based on linear regression and logarithmic regression， which are combined to form a hybrid forecasting method. The results of our experiment show that compared with traditional time-series models， the proposed method is of higher accuracy for very short-term forecasting. |
| |
| Keywords: | very short-term forecasting photovoltaic power forecasting time series dynamic regression ARIMA harmonic time series |
|
| 点击此处可从《陕西电力》浏览原始摘要信息 |
|
点击此处可从《陕西电力》下载全文 |
