短期电价预测的机器学习方法:现状、挑战与展望
Machine Learning Methods for Short-Term Electricity Price Prediction:Current Status,Challenges and Perspectives
郑志勇 1陈田原 2胡哲 2许群2
作者信息
- 1. 河南省科学院数学研究所,郑州 450046;中国人民大学数学学院,北京 100872
- 2. 中国人民大学数学学院,北京 100872
- 折叠
摘要
随着全国统一电力市场的逐步建设,电价预测领域吸引了越来越多的学者关注.开发高效、精准的短期电价预测模型,不仅能够为电力现货市场的买卖双方提供决策支持,还能提升各方的风险控制能力,从而为建立高效统一的电力市场提供量化支撑.由于电力价格有随机性大、波动性高和时效性强等特点,将机器学习方法应用在短期电价预测领域能够综合考虑多方因素,能在较短的时间内拟合较为复杂的非线性电价数据,得到更为精准的预测结果.本文对基于机器学习的短期电力价格预测方法进行了综述,深入分析了各类方法在不同情境下的预测效果、适用性及其优势与不足,并对未来研究方向进行了展望.
Abstract
With the gradual construction of the national unified electricity market,the field of electricity price forecasting has attracted more and more attention from scholars.The development of efficient and accurate short-term electricity price prediction models can not only provide decision support for buyers and sellers in the electricity spot market,but also enhance the risk control ability of all parties,thus providing quantitative support for the establishment of an efficient and unified electricity market.Due to the large randomness,high volatility and strong timeliness of electricity prices,the application of machine learning methods in the field of short-term electricity price prediction can take into account a variety of factors,and can fit the more complex nonlinear price data in a shorter period of time to obtain more accurate prediction results.This paper provides an overview of the short-term electricity price prediction methods based on machine learning methods,analyzes the prediction effect and applicability of each type of method in different contexts and their advantages and shortcomings,and and looks forward to the future research direction.
关键词
电力现货市场/电价预测/机器学习/深度学习/Transformer模型Key words
electricity spot market/electricity price forecasts/machine learning/deep learning/Transformer modle引用本文复制引用
出版年
2025
NETL
NSTL
万方数据