Relationship Between Indoor And Outdoor Of Fine Particulate Matter In Residential Buildings

At the present time, the air quality, which affects human health, is anxious because of the exhaust gas emitted from industries, transports and residential. It is improved by a large amount of papers that the morbidity, hospital admissions, and mortality rate highly relate to the ambient particulate matter (PM) concentration. Especially within the rooms, where individuals spent their 70%～90% life time, the indoor air quality is influential for the public health. Hence, the indoor air quality and the relationship between indoor and outdoor particle mass concentration is firstly focused on.In this study, based on the datasets collected from typical residential buildings in Changsha city, the relationships between indoor and outdoor PM2.5/PM10 mass concentration and the factors which effect indoor concentrations were evaluated by the different statistics methods (Compare Means Analysis, Pearson Correlation Analysis and Multiple Regression Analysis). The indoor PM prediction model (Mixed Model) was established which based on the Multiple Linear Regression and the system identification theory. The main achievements and conclusions were as follows:(1) PM10 and PM2.5 are the most important pollutants in the ambient and indoor environment, respectively. And the PM2.5/PM10 pollution levels of both indoor and outdoor in Changsha are 2～8 times higher than such levels in the European countries and North America, which may be construed by that the habits and customs of Chinese residents are different from such areas.(2) PM2.5 and PM10 within the rooms mainly come from the outdoor when there are absences of PM sources of the room inside.(3) Ventilation conditions and indoor sources are the important factors for the indoor PM pollution level. Temperature and relative humidity, which are the key factors of the coagulation process, affect the PM concentrations to the extent.(4) A new indoor PM concentration prediction model (Mixed Model) is established. It combined the merits of the multiple linear regression models which can estimate the weight factor of the variable parameters and the ARIMA model which can reflect the time series prediction. Compared with the monitor data, the predicted data calculated by Mixed Model are inosculated.