基于SARIMA模型的用电负荷超短期自适应预测方法  

The Ultra-short-term Adaptive Electricity Load Forecasting Method Based on the SARIMA Model

作　　者：沈扬 沈泓 SHEN Yang;SHEN Hong(Electric Power Dispatching Control Center of Changzhou Power Supply Branch of State Grid Jiangsu Electric Power Co.,Ltd.,Changzhou 231000,China)

机构地区：[1]国网江苏省电力有限公司常州供电分公司电力调度控制中心,江苏常州231000

出　　处：《南京工程学院学报(自然科学版)》2024年第4期78-84,共7页Journal of Nanjing Institute of Technology(Natural Science Edition)

基　　金：国网常州供电公司用电负荷实时态势感知模型下的保供电辅助决策技术研究服务(SGJSCZ00KJJS2311049)。

摘　　要：针对超短期负荷预测模型的不稳定性,文章提出一种基于季节性自回归移动平均模型的超短期自适应用电负荷预测方法,旨在提高负荷预测的准确性和适应性.首先对用电负荷数据进行季节性分析,确定数据的季节性周期,并对数据进行季节性调整;然后根据季节性调整后的数据,建立季节性自回归移动平均模型.该模型综合考虑自回归、移动平均和季节性成分,能够捕捉数据的长期趋势和季节性变化.为了提高模型的自适应性,引入参数自适应策略.该策略基于模型的预测误差,自动调整模型的参数,以适应负荷数据的动态变化.In light of the instability of ultra-short-term load forecasting models,this paper proposed an adaptive electricity load forecasting method based on a seasonal autoregressive integrated moving average(SARIMA)model.The primary objective was to enhance both the accuracy and adaptability of load predictions.Firstly,a seasonal analysis of the electricity load data was undertaken to identify its seasonal cycles and implement seasonal adjustments accordingly.Subsequently,a SARIMA model that integrated autoregression utilizing the seasonally adjusted data were developed,moving average components and seasonal factors.This model effectively captured both long-term trends and seasonal variations in the data.To further improve the model's adaptability,a parameter adaptation strategy that automatically adjusts the model parameters based on prediction errors was proposed,facilitating improved alignment with dynamic fluctuations in load data.

关 键 词：季节性自回归移动平均模型 自适应 用电负荷预测 季节性分析 参数自适应 预测准确性 

分 类 号：TM615[电气工程—电力系统及自动化]