基于PCA-MLR的公共建筑能耗报警阈值计算
Calculation of public building energy consumption alarm threshold based on PCA-MLR

为提高建筑能耗监测系统的运行管理效率，提出一种基于PCA-MLR的公共建筑能耗报警阈值计算方法. 首先根据建筑的实际使用情况对能耗进行分类处理，接着以温度、 含湿量、 风速等8个环境因素作为初始变量，通过主成分分析，提取5个主元. 在此基础上，分别建立不同类型能耗的多元线性回归模型. 对模型进行F检验，并与单一元线性回归模型进行对比，模型精度较高. 将方法用于公共建筑能耗监测系统，系统运行正常可靠，误报警少，强化了系统的报警性能.
In order to improve the operational efficiency of building energy consumption monitoring system，a reasonable set of alarm threshold is very important. First，the actual situation of the building is used to classification of the energy consumption. Then，8 environmental factors including temperature，humidity and wind speed as the initial variables，5 principal components were put forward through the principal component analysis(PCA)，and the multiple linear regression(MLR) models with different energy consumption patterns were established. F test is carried out on the model，and compared with the single MLR，the model has higher precision. Using this method，the alarm performance of building energy consumption monitoring system can be enhanced