基于EEMD-CatBoost的短期电力负荷预测  

Short-Term Power Load Forecasting Based on EEMD-CatBoost

作　　者：顾恩到 郭延鹏 GU Endao;GUO Yanpeng(School of Electrical Engineering,North China University of Water Resource and Electric Power,Zhengzhou,Henan 450011,China)

机构地区：[1]华北水利水电大学电气工程学院,河南郑州450011

出　　处：《自动化应用》2023年第5期221-224,228,共5页Automation Application

摘　　要：准确的负荷预测在电力调度、系统可靠性和规划中起着关键作用。针对各种不确定因素造成了电力需求的波动,本文提出了一种基于EEMD-CatBoost的短期负荷预测方法。模型利用集合经验模态分解(EEMD)对非平稳原始序列进行处理,将原始电力负荷数据分解为有限个固有模态函数(Intrinsic Mode Functions,IMF)和一个残差分量,以降低负荷序列的复杂度,再将分解后的各分量分别输入到CatBoost中预测,然后将每个分量的预测值重组,得到最终的负荷预测结果。以某地的实际数据为例,综合比较了该方法与现有电力负荷短期预测技术的性能。与现有基准相比,所提出的方法得到了相当精确的结果。Accurate load forecasting plays a key role in power dispatching,system reliability and planning.Aiming at the fluctuation of power demand caused by various uncertain factors,this paper proposes a short-term load forecasting method based on EEMD-CatBoost.In the model,the ensemble empirical mode decomposition(EEMD)is used to process the non-stationary original sequence,and the original power load data is decomposed into a finite number of intrinsic mode functions(IMF)and a residual component to reduce the complexity of the load sequence.Then,the decomposed components are respectively input into CatBoost for forecasting,and then the predicted values of each component are recombined to obtain the final load forecasting result.Taking the actual data of a certain place as an example,the performance of this method is comprehensively compared with the existing short-term power load forecasting technology.Compared with the existing benchmark,the proposed method produces quite accurate results.

关 键 词：集合经验模态分解 CatBoost 固有模态函数(IMF) 电力负荷短期预测 

分 类 号：TM74[电气工程—电力系统及自动化]