基于数据的建筑能耗分析与建模

Databased Analysis and Modeling of Building Electricity Energy Consumption

摘要:

对建筑能耗数据进行深入分析,提出了建立建筑电力能耗模型的方法.首先对建筑能耗进行了分项计量,统计了建筑逐时照明能耗数据,办公设备能耗数据以及办公人数,同时调查了建筑管理控制方式.经过分析可知,办公设备能耗数据与办公人数线性相关,而照明能耗数据与办公人数以及太阳辐射强度相关,但为非线性关系.根据不同类型数据的特点,分别建立了线性回归模型以及决策树模型.该模型可以预测建筑能耗并评估建筑管理方式对能耗的影响.

关键词:建筑电力能耗; 线性回归算法; 决策树; 分类与回归树

Abstract:

The building electricity energy consumption models were established based on the analysis of building electricity energy consumption.Data were collected by hours about the office equipment energy consumption,lighting energy consumption and number of people.Investigations were made into the building management modes.Based on the collected data,a close relationship between equipment electricity energy consumption and number of people was found and the linear regression model was then developed.Lighting energy consumption was found to be related with number of people and solar radiation intensity,but the relationship was nonlinear.A classification model was developed with the decision tree algorithm.With the models,electricity consumption can be predicted hourly and the impact of building management mode on power demands can be assessed.

Key words:building electricity energy consumption; linear regression algorithm; decision trees; classification and regression trees