点击次数：

论文类型：会议论文

发表时间：2008-01-01

收录刊物：CPCI-S

页面范围：1971-1978

关键字：building energy consumption; PLSR; weather conditions

摘要：The building energy consumption (BEC) is influenced by many factors. For the existing building, the local weather condition is the most important influenced factor. The present prediction studies on BEC mainly apply two kinds of models. One is the model based on the mathematical statistics by mean of Multiple Linear Regression (MLR), which can not better deal with the problem of parameter relativities. Another is the engineering model based on the technology analysis, which is complicated and is difficult to master by general researchers. I order to solve these problems; this paper presents a prediction model of BEC by PLSR (Partial Least Square Regression). Through an calculation example of a building in Dalian, the BEC is predicted by use of this model taking weather condition as main factor. The result of 1.08% mean relative error shows that this model has better forecast precision compared with other studies, and could be considered as a new making decision tool for retrofitting and re-optimization design of air-conditioning systems in large-scale buildings.