Table 1.
Status of dust data acquisition.
Fig 1.
Schematic diagram of PM2.5 monitoring system.
Fig 2.
APRUS adapter.
Fig 3.
Data interaction rule code.
Fig 4.
Protocol adaptation code.
Fig 5.
MIXIOT platform.
Fig 6.
Customize visualization scheme.
Fig 7.
Visualization of monitoring data.
Fig 8.
Monitoring device.
Fig 9.
Overall architecture.
Fig 10.
No. 5 monitoring point of transportation road of opencast coal mine.
Fig 11.
Time series cross-validation.
Fig 12.
PM2.5 concentration change from 10:40 to 11:40 on October 18, 2021.
Fig 13.
Time series of PM2.5 concentrations.
Fig 14.
PM2.5 time series decomposition.
Table 2.
Iterative results.
Fig 15.
Fitting results of DES model with different parameters.
Fig 16.
Cross-validation of DES model.
Fig 17.
Autocorrelation diagram of differential sequence.
Table 3.
Original sequence of ADF test results.
Table 4.
Results of 1st order differencing.
Fig 18.
ACF and PACF.
Fig 19.
AIC thermal diagram.
Fig 20.
BIC thermal diagram.
Fig 21.
ARIMA (0,1,0) model and ARIMA (4,1,0) model.
Fig 22.
Residual autocorrelation diagram and partial autocorrelation diagram of ARIMA (0,1,1) (a) and ARIMA (4,1,0) (b).
Table 5.
Characteristics of the best Arima fitting model for PM2.5 concentration in the training set.
Fig 23.
ARIMA (0,1,1) model cross-validation.
Fig 24.
ARIMA (4,1,0) model cross-validation.
Fig 25.
Residual analysis.
Table 6.
Characteristic evaluation of DES model and ARIMA (4,1,0,).
Fig 26.
Prediction process.