Fig 1.
Coordinates of air quality monitoring stations.
Location map of each air quality monitoring station in Beijing.
Table 1.
Name of air quality monitoring station.
Table 2.
Attributes of the correlation factors input when the model is making predictions.
Fig 2.
Structure of the box line diagram.
A boxplot is a statistical graph showing the distribution of one or more sets of continuous quantitative data, the first quartile, the median, and the third quartile.
Fig 3.
The resulting graph of outliers is tested using a boxplot.
Taking the air quality monitoring station of the Aotizhongxin as an example, this research used the boxplot principle to test the outliers of the collected historical data.
Fig 4.
Pre-processed data visualization.
Comparison of meteorological data and air pollutant concentration data before and after data preprocessing at the central station. (a) Visualization of unprocessed data. (b) Visualization of pre-processed data.
Fig 5.
Time autocorrelation analysis.
The autocorrelation coefficient of each station changes with time delay of 0–24 hours.
Fig 6.
Correlation coefficients of meteorological data and pollutant concentration data at each station.
Fig 7.
Informer model maintains encoder-decoder structure, and it adds ProbSpare self-attention mechanism, which effectively reduces time complexity and memory usage.
Fig 8.
Structure diagram of ST-Informer model.
A spatio-temporal embedding layer is added to the original Informer model to fully analyze the spatiotemporal characteristics of the input data.
Table 3.
Setting of ST-Informer model parameters.
Fig 9.
The value of the model evaluation index.
The values of the evaluation indicators obtained by predicting PM2.5 in different models.
Table 4.
Compare the performance of different models.
Fig 10.
The results of four models to predict PM2.5 concentration in different time periods in the future.
(a) The results of four models predicting PM2.5 concentrations over the next 4 hours; (b) The results of four models predicting PM2.5 concentrations over the next 8 hours; (c) The results of four models predicting PM2.5 concentrations over the next 12 hours; (d) The results of four models predicting PM2.5 concentrations over the next 16 hours; (e) The results of four models predicting PM2.5 concentrations over the next 20 hours; (f) The results of four models predicting PM2.5 concentrations over the next 24 hours.
Fig 11.
Predicted results of other air pollutant concentrations.
The ST-Informer model was used to predict 60-hour other air pollutant concentrations. (a) The results of the O3 concentration prediction. (b) The results of the SO2 concentration prediction. (c) The results of the NO2 concentration prediction. (d) The results of the PM10 concentration prediction.