基于ARIMA-LSTM模型的综合能源系统负荷与风光资源预测  

Multivariate load prediction and wind-solar resource characteristic quantity prediction of integrated energy system based on ARIMA-LSTM model

作　　者：王鑫[1] 李安桂[1] 李扬[1] 卜令晨 彭怀午[2] 牛东圣 许晨琛 韩欧 WANG Xin;LI Angui;LI Yang;BU Lingchen;PENG Huaiwu;NIU Dongsheng;XU Chenchen;HAN Ou(School of Building Services Science and Engineering,Xi′an Univ.of Arch.&Tech.,Xi′an 710055,China;Institute of Solar Engineering Technology,Northwest Engineering Corporation Limited,Power China,Xi′an 710065,China)

机构地区：[1]西安建筑科技大学建筑设备科学与工程学院,陕西西安710055 [2]太阳能利用工程技术研究所,中国电建集团西北勘测设计研究院有限公司,陕西西安710065

出　　处：《西安建筑科技大学学报（自然科学版）》2022年第5期762-769,共8页Journal of Xi'an University of Architecture & Technology(Natural Science Edition)

基　　金：十二五国家科技支撑课题合作单位基金资助项目(2011BAJ03B03-5)。

摘　　要：在能源互联网快速发展的背景下,研究分析了综合能源系统的多元负荷预测模型及理论方法.针对传统ARIMA(Autoregressive Moving Average Model,ARMA)模型仅能处理线性关系的问题,将ARIMA模型与LSTM(Long-Short Term Memory,LSTM)网络模型结合,提出并建立了ARIMA-LSTM模型.该模型不仅兼容冷、热、气、电等多元负荷的预测,并且可以用于风速、辐射照度等数据的预测,有较好的适应性和预测精度.Under the background of the rapid development of energy Internet,this paper studies and analyzes the multivariate load prediction and theoretical method of integrated energy system.Aiming at the problem that the traditional ARIMA model can only deal with the linear relationship,the ARIMA-LSTM model is proposed and established by combining ARIMA model with LSTM(Long-Short Term Memory,LSTM)network model.The model is not only compatible with the prediction of multiple loads such as cold,heat,gas and electricity,but also can be used for the prediction of wind speed,radiation illumination and other data,and has good adaptability and prediction accuracy.

关 键 词：综合能源系统 负荷预测 构建模型 误差分析 

分 类 号：TU111.1954[建筑科学—建筑理论] TK01[动力工程及工程热物理]