摘要
随着新型电力系统的不断发展,电力系统对负荷预测提出了更高的要求。负荷预测作为电力调度日常工作的一部分,在方式调整、供电规划以及用电平衡方面有着重要的作用,而负荷预测的模型选择很大程度上决定了预测准确性的上下限。为了进一步分析对比传统回归模型与深度学习模型在电力负荷预测中的优缺点,首先对差分自回归移动平均模型(ARIMA)与长短期记忆神经网络模型(LSTM)进行理论介绍,然后通过比较两种模型在电力负荷预测中的实际应用效果,总结两种方法各自的优缺点,最后对负荷预测的发展方向进行了可行性讨论。
With the continuous development of new power systems, power systems have put forward higher requirements for load forecasting. As a part of the daily work of power dispatch, load forecasting plays an important role in mode adjustment, power supply planning, and power consumption balance. The choice of load forecasting models largely determines the upper and lower limits of forecast accuracy. In order to further analyze and compare the advantages and disadvantages of traditional regression models and deep learning models in power load forecasting, First the theory of autoregressive integrated moving average model(ARIMA) and the long short-term memory model(LSTM) are introduced, and the advantages and disadvantages of each method are summarized, the feasibility of the development direction of load forecasting is discussed.
作者
朱剑飞
孙锦涛
卫科
刘贺龙
李政宇
ZHU Jianfei;SUN Jintao;WEI Ke;LIU Helong;LI Zhengyu(State Grid Jincheng Power Supply Company,Jincheng 048000,China)
出处
《电气应用》
2022年第2期27-31,共5页
Electrotechnical Application
关键词
新型电力系统
负荷预测
回归模型
深度学习
new power system
load forecasting
regression model
deep learning