摘要
近些年,风能成为世界上装机容量较大的可再生能源之一,风力发电的并网容量不断增加,给电网稳定运行带来不小挑战,风力输出功率预测精度的提升能够有效地减轻风电并网时对电网的冲击,同时为电网的调度和安全运营提供保障。为进一步提升的风电功率预测精度,借鉴分形理论并将其融合到风电功率预测模型中,同时结合自定义K最近邻算法(K-nearest-neighbor,KNN)。采用分形理论的基本思想,考虑基准功率曲线问题和气象特征值,利用分形插值可有效地获取相邻样本的局部信息,从而使得插值曲线更好地保留原采样信息的大部分特征,最后使用多评价指标维度对预测效果进行评估。以某风电场实测数据为例,与梯度提升决策树、随机森林、支持向量机预测模型进行比较,验证了提出的预测算法的有效性。
In recent years,wind energy has become one of the renewable energies with large installed capacity in the world.The grid-connected capacity of wind power generation has been gradually increasing,which brings great challenges to the stable operation of the power grid.The improvement of the prediction accuracy of wind power output can effectively reduce the impact of wind power grid.In order to achieve a better wind power prediction method,we study the application of fractal theory to wind power generation,apply the fractal theory to the wind power prediction model and combine the wind power prediction method of the custom KNN.Considering the problem of reference power curve and meteorological characteristic value,fractal interpolation can effectively obtain the local information of adjacent samples,so that the interpolation curve can better retain most of the features of the original sampling information and finally apply multiple evaluation indicators.The prediction results are evaluated and compared with the prediction models RFR,SVM and GBDT.The measured data of a wind farm is taken as an example to verify the effectiveness of the prediction model proposed.
作者
李昂儒
郑伟彦
赵京虎
杨勇
王辉东
汪李忠
邢海青
LI Ang-ru;ZHENG Wei-yan;ZHAO Jing-hu;YANG Yong;WANG Hui-dong;WANG Li-zhong;XING Hai-qing(NARI Group Corporation(State Grid Electric Power Research Institute),Nanjing 211106,China;NARI Research Institute Xi’an R&D Center,Xi’an 710000,China;State Grid Zhejiang Hangzhou Power Supply Co.,Ltd.,Hangzhou 310007,China;State Grid Zhejiang Power Supply Co.,Ltd.,Hangzhou 310012,China;State Grid Zhejiang Hangzhou Yuhang District Power Supply Co.,Ltd.,Hangzhou 311100,China)
出处
《计算机技术与发展》
2021年第3期191-195,共5页
Computer Technology and Development
基金
国家电网公司总部科技项目(5400-201919144A-0-0-00)。
作者简介
李昂儒(1991-),男,研究生,中级软件工程师,研究方向为大数据挖掘与智能数据处理。
