基于周期性ARMA-SVR模型的空调冷热负荷预测  

Air-conditioning Cooling and Heating Load Prediction Based on Periodic ARMA-SVR Model

作　　者：甘中学 喻想想 许裕栗 李德伟[2] GAN Zhong-xue;YUXiang-xian;XU Yu-li;LI De-wei(ENN Science&Technology Development Co.Ltd,Langfang 065001,China;Department of Automation,Shanghai Jiao Tong University,Shanghai 200240,China)

机构地区：[1]新奥科技发展有限公司,河北廊坊065001 [2]上海交通大学自动化系,上海200240

出　　处：《控制工程》2020年第2期380-385,共6页Control Engineering of China

基　　金：国家重点基础研究发展计划(973计划)(2014CB249200)。

摘　　要：工业用户的空调冷热负荷预测对于有目的的节能减排有重要作用。由于用户的冷热负荷数据具有非线性、外界干扰多且呈日周期性的特点,采用传统的ARMA和SVR方法不能取得良好的预测效果。因此提出一种利用日周期性特点的组合ARMA模型和SVR模型的预测方法:首先结合原始数据的日周期性特点,采用ARMA模型进行线性预测;对于ARMA模型的预测残差中保留的原始数据的非线性特征,利用SVR模型对残差进行非线性部分的预测,修正原来的预测结果,得到最终的预测值。采用真实数据的实验结果显示,新提出的预测方法可以显著改善预测效果。For industrial users, the prediction of air-conditioning load can be great helpful to reduce the energy consumption. Besides the nonlinear character, the user’s load data is also very sensitive to the external disturbance and with daily periodic characteristics. This makes the traditional ARMA and SVR methods hard to achieve good prediction. Therefore, a method of combining ARMA model and SVR model with daily periodic characteristics is presented. Firstly, combined with the daily periodic characteristic of the raw data, the ARMA model is used for linear prediction. For the nonlinear characteristics of original data which retained in the prediction residuals of ARMA model, the SVR model is used to predict the nonlinear part of the residuals and modify the prediction result, obtaining the final predicted value. Experimental results using actual data show that the proposed method can significantly improve the prediction performance.

关 键 词：自回归移动平均 支持向量机 空调冷负荷 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]