摘要
根据空调日总负荷和日平均气温之间的较强相关性,对经典的季节性指数平滑法预测模型中的水平因子项进行修正,并去掉趋势因子项,得到了改进的季节性指数平滑预测模型.以上海某医学中心空调系统作为该空调负荷预测模型的应用实例,在整个预测期内模型平均预测误差为8.8%.这表明改进的季节性指数平滑法适合于办公类建筑空调负荷的预测.
According to the strong correlation between daily average ambient temperature and daily total air conditioning load, a modified seasonal exponential smoothing model for air conditioning load prediction is established. Comparing with traditional seasonal exponential smoothing model, it modifies level factor, and gets rid of trend factor. The modified prediction model is applied to the air conditioning system of a medical central in Shanghai. The average error of this prediction is 8.8 %, showing that the modified seasonal exponential smoothing model is fit for the load prediction of office buildings.
出处
《同济大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2005年第12期1672-1676,共5页
Journal of Tongji University:Natural Science
关键词
空调负荷
季节性指数平滑法
预测
air conditioning load
seasonal exponential smoothing
prediction
