制冷技术2025，Vol.45Issue(4)：52-59,8.DOI:10.3969/j.issn.2095-4468.2025.04.203
数据受限条件下的商业建筑制冷空调系统负荷预测研究
Research on Load Forecasting of Commercial Building Refrigeration and Air Conditioning System under Data Limitation
摘要
Abstract
In order to study the load forecasting of refrigeration and air conditioning system of commercial buildings under the condition that the original data collection is limited,different data are divided into three basic feature groups:cold station data,meteorological data and indoor data.Seven load prediction models are constructed through the combination of basic feature groups,and the model prediction results are quantitatively evaluated.The results show that when a single feature set is used as the model input,the prediction effect of cold station data is the best.In all models,when all features are taken as input,the prediction effect is the best:root mean square error is 3.03 and goodness of fit R2 is 0.87.The quantitative evaluation of the prediction results under different feature combinations is carried out to provide some help for the feature selection of load prediction data end and the collection of original data.关键词
商业建筑制冷系统/负荷预测/神经网络/特征分析Key words
Commercial building refrigeration system/Load forecasting/Neural network/Feature analysis分类
信息技术与安全科学引用本文复制引用
樊超,张鉴心,陈焕新,黎强,肖准..数据受限条件下的商业建筑制冷空调系统负荷预测研究[J].制冷技术,2025,45(4):52-59,8.基金项目
国家自然科学基金(No.51876070). （No.51876070）
