| With the continuous growth of economic and building scale in China,the comfort of life and proportion of energy consumption in buildings people were paied more and more attention.The energy consumption in buildings has huge energy-saving potential.As the opening part and one of the important links for energy exchange between inside and outside of residential building,window is the most common and effective means for people to adjust the indoor environment.At present,resident usually operate indoor window for natural ventilation in order to create comfortable indoor environment as much as possible in the transitional seasons.In addition to relying on environmental control equipment to adjust indoor temperature and humidity,the indoor environment can also be adjusted by opening and closing window to meet human comfort requirements in winter and summer.In summary,studying and analyzing the factors that affect residential window opening behavior in different seasons can provide references for establishing appropriate window opening behavior prediction models and studying the impact of residential window opening behavior on building energy consumption.In this study,we choosed different residences in different locations in Xi’an province as the experimental objects.In addition,we have monitored the opened or closed states of the windows and indoor and outdoor environmental parameters in six residences for two years.Then dividied the collected data in different seasons,and used the SPASS software to analyze the environmental factors that affect the state of the windows in the four seasons.（air conditioning season,heating season,transition season 1（spring）and transition season 2（autumn））.Established the Naive Bayes classification window forecasting models of the four seasons and verified their results.Analyzing and discussing the influence of resident windowing behavior on building energy consumption in heating and air conditioning based on TRNSYS energy consumption simulation software to provide more theoretical guidance for the energy-saving transformation and evaluation of residential buildings in Xi’an city.We can get the following conclusions through research:（1）Transition season 1（spring）lasts about 18 days longer than Transition season 2（autumn）in Xi’an,and the transition time from air conditioning season to the transition season 2（autumn）is shortest only half a month in Xi’an.（2）The influence of indoor and outdoor environmental parameters on the state of indoor window is not identical in different seasons.The main influencing factor in heating season and air-conditioning season are indoor CO2 concentration.（3）The accuracy of Naive Bayesian classification window prediction model is bigger than 82%in the air conditioning season and transition seasons.The accuracy is about 77%in the heating season.The accuracy of the model was the highest and the performance was stable in the air conditioning season.（4）The energy consumption of urban resident building in Xi’an is obviously different under three different windowing behavior modes.The energy consumption of the model based on the probability of people indoor in heating season is the largest,which is about 6%higher than the energy consumption based on the real window opening data,and smallest in the fully closed window mode.In the air conditioning season,The energy consumption of the model established based on the probability of people indoor and the real window opening datas are about 13%less than the energy consumption obtained in the fully closed window mode. |
