摘要
在可再生能源中,随着光伏发电量逐年增加,光伏发电的波动性可能对电网系统带来多种不利影响,准确的光伏发电量预测有益于确保电网的稳定运行。对光伏发电的直接预测进行了系统的阐述,讨论了输入输出数据的相关性以及模型输入数据预处理的重要性,基于不同类别的几种光伏功率预测模型进行性能分析,考虑了不同预测模型的优劣势并进行了评估。研究结果表明,经过优化的算法显著提高了模型的预测精度,遗传算法被认为是最可行的优化方法之一。
In renewable energy,with the increase of photovoltaic power generation capacity year by year,the volatility of photovoltaic power generation may have a variety of adverse effects on the power grid system,and accurate photovoltaic power generation capacity prediction is beneficial to ensure the stable operation of the power grid.In this paper,the direct prediction of photovoltaic power generation is systematically expounded,the correlation of input and output data and the importance of model input data preprocessing are discussed,and the performance analysis is carried out based on several photovoltaic power prediction models of different categories,the advantages and disadvantages of different prediction models are considered and evaluated.The study results show that the optimized algorithm significantly improves the prediction accuracy of the model,and the genetic algorithm is considered to be one of the most feasible optimization methods.
作者
杨海亭
白伟
胡运冲
YANG Hai-ting;BAI Wei;HU Yun-chong(Inner Mongolia Jingneng Kangbashi Thermal Power Co.,Ltd,Ordos 017000,China)
出处
《电工电气》
2024年第4期1-9,共9页
Electrotechnics Electric
基金
内蒙古京能康巴什热电有限公司厂区分布式光伏发电项目(ZBA172101515)。
关键词
分布式光伏发电
功率预测模型
模型优化
清洁能源
distributed photovoltaic power generation
power prediction model
model optimization
clean energy
