| 非晴空条件下光伏发电短期功率预测方法 |
| |
| 引用本文： | 王清亮,杨博,应欣峰,宋曦,高梅.非晴空条件下光伏发电短期功率预测方法[J].太阳能学报,2022,43(3):188-196. |
| |
| 作者姓名： | 王清亮 杨博 应欣峰 宋曦 高梅 |
| |
| 作者单位： | 西安科技大学电气与控制工程学院,西安 710054 |
| |
| 基金项目： | 陕西省自然科学基础研究计划面上基金（2017JM5138）； |
| |
| 摘 要： | 针对非晴空条件下光伏发电短期功率预测精度不高的问题,提出一种基于自适应混合核的相关向量机光伏发电短期功率预测方法。通过构造混合核函数和自适应寻优核参数来增强预测模型的泛化和学习能力,建立对多尺度多模态变化数据的映射关系,实现光伏发电功率随机性波动规律的机器学习和有效捕捉。采用关联系数筛选历史相似日,通过历史相似日数据自动确定最优预测模型。最后,采用美国俄勒冈州某光伏电站的实测数据进行实验,对比该文方法与其他预测方法的功率预测精度。结果表明该文预测方法在非晴空条件下对光伏发电短期功率预测精度最高。
|
| 关 键 词： | 光伏发电 功率预测 机器学习 非晴空 核函数 |
| 收稿时间： | 2020-06-23 |
SHORT-TERM PHOTOVOLTAIC POWER FORECASTING METHOD UNDER NON-CLEAR SKY CONDITION |
| |
| Wang Qingliang,Yang Bo,Ying Xinfeng,Song Xi,Gao Mei.SHORT-TERM PHOTOVOLTAIC POWER FORECASTING METHOD UNDER NON-CLEAR SKY CONDITION[J].Acta Energiae Solaris Sinica,2022,43(3):188-196. |
| |
| Authors: | Wang Qingliang Yang Bo Ying Xinfeng Song Xi Gao Mei |
| |
| Affiliation: | School of Electrical and Control Engineering, Xi’an University of Science and Technology, Xi’an 710054, China |
| |
| Abstract: | To solve the problem that the short-term power forecasting accuracy of a photovoltaic generator is not high under non-clear sky conditions, a new short-term power forecasting method is proposed based on relevance vector machine model with an adaptive hybrid kernel. By constructing hybrid kernel function and adaptive optimization kernel parameters, the generalization and learning abilities of the forecasting model are enhanced. Then, the mapping relation of multi-scale and multi-mode data is established to realize the machine learning and effective capture of the random fluctuation rules of photovoltaic power generation. The correlation coefficient was used to screen the historical similarity days, and the optimal forecasting model was automatically determined by the data of historical similarity days. Finally, the short-term power forecasting accuracy of the proposed method and other forecasting methods is compared with the measured data from a photovoltaic power station in Oregon, USA. The results show that the proposed method has the highest accuracy for short-term power forecasting of the photovoltaic generator under non-clear sky conditions. |
| |
| Keywords: | photovoltaic generators power forecasting machine learning non-clear sky kernel function |
|
| 点击此处可从《太阳能学报》浏览原始摘要信息 |
|
点击此处可从《太阳能学报》下载全文 |
|
