Bo Li

In this study, the indoor air quality (IAQ) was investigated in a subway station with fully enclosed platform screen doors in Beijing, China. Eight indoor air pollutants, including PM_2.5, PM₁₀, SO₂ (sulfur dioxide), NO₂ (nitrogen dioxide), NH₃ (ammonia), CO (carbon monoxide), CH₂O (formaldehyde) and TVOC (total volatile organic compound), were measured for six consecutive days in October 2019. The results indicated that the IAQ in the subway station was basically stable at good levels for most times during the whole measurement period. All eight indoor air pollutants were far below their corresponding maximum allowable concentrations, except for the PM_2.5 concentrations, which occasionally exceeded the concentration limits. The concentrations of indoor air pollutants in the subway station were basically within the corresponding standards. The correlation analyses showed that outdoor air pollutants have important influences on indoor air pollutants. The concentrations of PM₁₀, PM_2.5, SO₂, NO₂ and CO in the subway station were positively correlated with their corresponding outdoor concentrations. PM₁₀ was statistically significantly correlated with the passenger flow and train frequency, but the other air pollutants were less impacted by the passenger flow and train frequency.

With the increasing number of metros, the comfort and safety of crew and passengers in metro stations have been paid great attention. The environment forecasting has become very important for decision-making. The outputs of the traditional point prediction methods are some exact values in the future. However, it might be closer to the real conditions that the predicted variables are given a probability range with a different confidence rather than exact values. This paper proposes a probabilistic forecasting method of metro station environment based on autoregressive Long Short Term Memory (LSTM) network. It has a good performance to quantify the uncertainty of environment trend in a metro station. Seven-day field tests were carried out to obtain the measured data of 7 internal environmental parameters in a metro station and 8 external environment parameters. In order to ensure the prediction performance, the random forest algorithm is used to select the input variables for the proposed probabilistic forecasting method. The selected input variables and the previous predicted values are as the input variables to build the probabilistic forecasting model. The proposed method can realize to predict the probabilistic distribution of internal environmental parameters in a metro station. This work may contribute to prevent emergency events and regulate environment control system reasonably.