Study On Driving Characteristics Of Air-conditioning Beaviour

With the increase of residential air conditioners popularity and the improvement of Household air conditioned environment of comfort requirements, air conditioning energy consumption also increased significantly in summer. Therefore, occurate grasp of residential air-conditioning energy consumption present situation and characteristics is an important foundation to do residential building energy conservation. Many researchers found that due to the environmental needs of different tenants, cause there is great difference between air conditioning usage of different households, and then there is a huge difference between the air conditioning energy consumption in different households. In the field of building simulation, to improve the effectiveness of the simulation results, researchers have been committed to building air conditioning behavior model, in order to ensure a reasonable description of air conditioning behavior. However, most of researchers focus on what the model describes the air conditioning behavior, while ignoring the drivers of air conditioning behavior actions taking place, but that is the basic premise of the description model. Therefore, discuss the main factors driving the air conditioning behavior, it is key to ensuring model can truly reflect the behavior of the air conditioning.In this paper, the research mainly includes field investigation and questionnaire,and the method of statistical analysis. First, continuous testing of 3 residential indoor temperature and humidity was carried out in Beijing area in 2014 and 2015 in the summer(June and August). Second,Questionnaire is used to understand human actual daily schedule and air conditioning operation. Final, To ensure an accurate grasp of the households of air conditioning operation, considering the results of the test data analysis and questionnaire investigation.Based on the above data, This study focuses on the three methods: logistic regression analysis、Grey Relational Analysis(GRA)、questionnaire. The above three methods are used to look into the drivers for the actions taken by occupants(air conditioning opening and closing). Further, based on the motion model of air conditioning behavior, with regard to different air conditioning operating conditions(turning on when feeling hot, turning off when feeling cold、turning off before sleep、turning off after sleep etc.), when assuming several different driving force condition, that is independent variable of the motion model probability function, comparative analysis the reflection status of air conditioning behavioral characteristics in different independent variables, In order to determine the appropriate drivers of air conditioning behavior.The results show, Logistic regression analysis cannot reflect behavioral characteristics in different operating conditions, Due to blend together the data of different operating conditions for calculation and analysis. For a reasonable description of air conditioning behavior, action-based quantitative stochastic model establish the corresponding conditional probability function for different trigger conditions. This method is suitable methods for drivers analysis.For air conditioning behavior description in different usage patterns, When determined the probability function form, Determining the drivers of air conditioning behavior has become the key of air conditioning behavior description. This study discussed and determined the drivers of air conditioning behavior in every mode,based on the probability function form of action-based quantitative stochastic model.