考虑时序二维变化的日前市场电价预测模型  

Forecasting Model of Day-ahead Market Electricity Price Considering 2D Variation of Time Series

作　　者：陈宇聪 白晓清[1] CHEN Yucong;BAI Xiaoqing(College of Electrical Engineering,Guangxi University,Nanning 530004,China)

机构地区：[1]广西大学电气工程学院,南宁530004

出　　处：《电力系统及其自动化学报》2024年第7期22-29,共8页Proceedings of the CSU-EPSA

基　　金：国家自然科学基金资助项目(51967001)。

摘　　要：电价预测对电力市场参与者的运营决策及电力系统安全稳定运行关系重大。针对日前市场电价预测问题,本文提出一种考虑时序二维变化的日前市场电价预测模型和方法。首先采用改进的带自适应噪声的完全集成经验模式分解对日前市场电价历史数据进行分解,然后对其高、低频子序列分别采用考虑时序二维变化的Ti⁃mesNet和基于统计分析的差分自回归移动平均进行预测,二者结果之和构成日前市场电价的预测值。仿真结果表明,所提方法相较于现有单一或组合模型具有较高的预测精度。Electricity price forecasting is of significance to the operation decision-making of power market participants and the safe and stable operation of power system.Aimed at the problem of day-ahead market electricity price forecast⁃ing,a day-ahead market electricity price forecasting model and the corresponding forecasting method are proposed in this paper considering the 2D variation of time series.First,the improved complete ensemble empirical mode decompo⁃sition with adaptive noise is used to decompose the historical data of day-ahead market electricity prices.Then,the high-and low-frequency sub-series are forecasted by TimesNet with the consideration of 2D variation of time series and autoregressive integrated moving average based on statistical analysis,respectively.The sum of the results constitutes the forecasted value of day-ahead market electricity price.Simulation results show that the proposed method has a high⁃er forecasting accuracy than the existing single or combined models.

关 键 词：日前市场电价预测 完全集成经验模式分解 差分自回归移动平均 TimesNet 时序二维变化 

分 类 号：TM715[电气工程—电力系统及自动化]