With the frequent occurrence of fog and haze, particulate matter pollution, especially the concentration of inhalable particulate matter (PM2.5), has attracted wide attention. Epidemiological studies have shown that there is a positive correlation between the morbidity and mortality of human body and the mass concentration of particulate matter in villages [1]. Long-term exposure to low-concentration particulate matter and short-term exposure to high-concentration particulate matter have adverse health effects [2]. Most of the existing air purifiers have good air purification capability for indoor buildings, but the air purification devices for open or semi-open space have not yet appeared. This is due to the lack of data in the early stage, so it is impossible to directly carry out the research and development of equipment. With the development of the city, Metro stands out among all kinds of vehicles for its convenience, safety and fast accuracy. More and more people take the subway as their first choice for daily travel every day. According to the EPA survey, Americans spent about 7.2% of their time on transportation as early as 1993-1994. With the establishment of subway stations, people spend more and more time on the subway. This approach is developed based on time series ARIMA theory to predict the concentration of particulate matter (PM2.5) in Metro stations, which is difficult to measure in Metro stations. It provides theoretical support for the prediction of particulate matter concentration in Metro in the future, and provides reference for future purification equipment and control strategy. The time series is used to fit and predict the PM2.5 concentration of the platform because the time series can predict the increase or decrease of several values under the known PM2.5 concentration of the platform, and the magnitude of the increase or decrease.