Research On PM_2.5 Concentration Distribution And Influencing Factors Of Subway Platforms

With the development of economic,major cities across China are building subways on a large scale.Studies have shown that the particulate matter generated in the subway contains heavy metal elements,and most of the subway lines are distributed underground,which is prone to accumulation of harmful substances.Long-term exposure to heavy metal environment will seriously endanger health.However,existing studies mainly focus on atmospheric PM2.5,and there are few studies on subway particulate matter.It is urgent to increase the research on subway particulate matter to better understand the distribution characteristics and influencing factors of subway PM2.5 concentration,and to formulate strategies suitable for subway stations to improve air quality.This paper aims to use statistical analysis methods to explore the temporal and spatial distribution of subway PM2.5 concentration,and use machine learning algorithms to explore the importance of factors affecting subway PM2.5 concentration,and forecast PM2.5concentrations in subway platforms.To this end,the project team arranged multiple sampling points for data collection at Shizi Station of Lanzhou Metro.This paper first compared the PM2.5 concentration values at different locations of the subway station,which found that the platform > station hall > carriage,and focused on comparing the PM2.5 concentration values at different sampling points of the platform,which found that the two ends of the platform > the middle of the platform.By analyzing the change trend of various index data monitored on the side of the carriage door along the subway line,it was found that when the train arrived at the station and opened the door,the platform air would pour into the carriage.In addition,comparing the proportions of particulate matter concentrations of different particle sizes at different sampling points in different time periods,it was found that the particulate matter generated by subway stations had a coarser particle size compared with the particulate matter generated by urban street-level traffic emissions.Then,by drawing the PM2.5 concentration value per minute in October into a heat map,it was found that the PM2.5 concentration at the station would show two obvious peaks in the morning and evening peaks during working days,and there were different changes in working days,holidays and weekends.Then,by using regression analysis and correlation analysis,it was found that the PM2.5concentration in the outdoor,mechanical ventilation,the power of ventilation system,and train frequency will affect the platform PM2.5 concentration.By using K-means cluster analysis and variance analysis,it was found that the PM2.5 concentration at the station was higher in rainy weather.Then,based on random forest,this paper fitted the relationship between the PM2.5concentration at the platform and various influencing factors.Based on Shapley’s analysis,the importance of influencing factors of PM2.5 concentration at platform was obtained,and it was found that the outdoor PM2.5 concentration had a great influence on the PM2.5 concentration of the subway platform.Based on the SHAP value,the positive and negative effects of a single influencing factor on the subway PM2.5 concentration were studied,and it was found that holidays,operating periods,and rainy weather would lead to an increase in the PM2.5concentration,and ventilation could reduce the PM2.5 concentration at the subway platform.Outdoor PM2.5 concentration and train frequency were positively correlated with platform PM2.5 concentration.Next,this paper considered a variety of influencing factors,and established models to forecast PM2.5 concentration after 2 hours at the subway platform based on three machine learning algorithms: random forest,XGBoost,and Light GBM.The results of the comparative analysis showed that the PM2.5 concentration forecasting model based on XGBoost has the best effect.Finally,according to the temporal and spatial distribution of subway particulate matter and the importance of influencing factors,this paper presented targeted recommendations on the prevention and control of subway particulate matter,based on the two goals of reducing the particulate matter concentration and energy saving.