基于Prophet-LSTM组合模型的短期负荷预测方法  

Short-term Load Forecasting Method Based on Prophet-LSTM Combination Model

作　　者：彭湃 刘敏[1] PENG Pai;LIU Min(School of Electrical Engineering,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学电气工程学院,贵阳550025

出　　处：《电力系统及其自动化学报》2021年第11期15-20,共6页Proceedings of the CSU-EPSA

基　　金：国家自然科学基金资助项目(51967004)。

摘　　要：针对短期负荷预测精度问题,提出一种基于Prophet加法模型和长短期记忆LSTM(long-short termmemory)网络的组合模型的短期负荷预测方法。首先分别建立Prophet预测模型和LSTM预测模型,然后采用最小二乘法对两种方法取不同的权重组合,得到新的模型并进行预测。以2014年全球能源预测竞赛(GEFCom2014)的电力负荷数据作为算例验证。实验结果表明,与ARIMA模型预测方法、随机森林模型负荷预测模型方法和标准Prophet和LSTM负荷预测方法相比,所提负荷预测方法所得结果具有更高的预测精度。Aimed at the problem of short-term load forecasting accuracy,a short-term load forecasting method based on the combination of the Prophet addition model and long-short term memory(LSTM)network.First,the Prophet and LSTM prediction models are established,respectively.Then,the least squares method is adopted to take different weight combinations of the two methods to obtain a new model,which is further used to make predictions.The power load data of 2014 Global Energy Forecasting Competition(GEFCom2014)is used as an example to verify the proposed method.Experimental results show that compared with the autoregressive integrated moving average(ARIMA)model,stochastic forest model,and the standard Prophet and LSTM load forecasting methods,the proposed method can obtain results with higher forecasting accuracy.

关 键 词：PROPHET 神经网络 最小二乘法 短期负荷预测 

分 类 号：TM715[电气工程—电力系统及自动化] TP183[自动化与计算机技术—控制理论与控制工程]