改进神经网络下的电力负荷短期预测方法
Short-Term Forecasting Method for Power Load Under Improved Neural Networks
戴雯菊 1严炼1
作者信息
- 1. 贵州电网有限责任公司贵阳供电局,贵州 贵阳 550001
- 折叠
摘要
随着我国电力供给压力的逐步增大,电力负荷短期预测工作变得尤为重要.其不仅可为电力调度工作提供一定参考,还能正确指导电价拟定.基于此,以神经网络的特点为切入点,分析改进神经网络下的电力负荷短期预测方法,并对RNN、LSTM、GRU进行实验对比分析,旨在证明通过改进神经网络,可有效提高电力负荷短期预测精度,为电力部门决策提供一定参考.
Abstract
With the gradual increase of power supply pressure in our country,short-term forecasting of power load is extremely important.It can not only provide certain reference for power dispatching work,but also guide the formulation of electricity prices correctly.Based on this,taking the characteristics of neural networks as the starting point,this paper analyzes and improves the short-term power load forecasting method under neural networks,and conducts experimental comparative analysis on RNN,LSTM,and GRU,aiming to prove that improving neural networks can effectively improve the accuracy of short-term power load forecasting and provide certain reference for decision-making in the power sector.
关键词
神经网络/电力负荷/短期预测Key words
neural network/power load/short-term forecasting引用本文复制引用
出版年
2024
NETL
NSTL
万方数据