摘要
电网运行电量数据呈现出海量化高维化的发展趋势,为有效利用这些大数据建立电量预测模型,提出了一种分时建模、经验模态分解和极限学习机相结合的方法。将每日输电量划分为峰、平、谷3个时段,并对每个时段的输电量曲线进行经验模态分解,再采用极限学习机训练得到各分量的预测模型,最后通过叠加合并得到最终的输电量预测值。通过与传统的极限学习机进行比较,结果表明所提方法可有效地提高模型的预测精度。
With the development trend of high-dimensional quantification of power grid operation data,in order to effectively use these big data to establish power forecasting model,a method combining time division modeling,empirical mode decomposition and extreme learning machine is proposed.The daily transmission capacity is divided into three periods:peak,flat and valley.The transmission capacity curve of each period is decomposed by empirical mode decomposition,and then the prediction model of each component is obtained by extreme learning machine training.Finally,the final prediction value of transmission capacity is obtained by superposition and combination.Compared with the traditional extreme learning machine,the results show that the proposed method can effectively improve the prediction accuracy of the model.
作者
庞红旗
高飞翎
程国开
罗玉鹤
陈静
温步瀛
PANG Hongqi;GAO Feiling;CHENG Guokai;LUO Yuhe;CHEN Jing;WEN Buying(Ningbo Electric Power Design Institute Co.Ltd.,Ningbo 315000,China;College of Electrical Engineering and Automation,Fuzhou University,Fuzhou350108,China)
出处
《智慧电力》
北大核心
2021年第9期63-69,共7页
Smart Power
基金
国家自然科学基金资助项目(61973085)。
关键词
输电量预测
峰平谷分时建模
经验模态分解
极限学习机
transmission capacity forecast
time division modeling of peak-plateau-valley
empirical mode decomposition
extreme learning machine
