Prediction of Gas Concentration Based on ARIMA and GARCH Model

In order to accurately predict the dynamic gushing process of gas in fully mechanized mining face, based on the historical monitoring data of underground gas concentration, with the help of R language, the ARIMA model is first established and fitted to determine the prediction equation of the ARIMA (p, d, q). The results of data fitting show that the model has a high degree of fitting to the gas concentration time series. Then the GARCH (u, v) is applied to the residual sequence of ARIMA (p, d, q), and the predicted value of the noise term in the ARIMA model is simulated, and the prediction result of the gas emission concentration is optimized. Finally, the 1001 fully mechanized mining face of Huangling No. 1 Mine in Shaanxi Province is taken as an application example. The results show that the combined model of ARIMA (p, d, q) and GARCH (u, v) can not only reflect the change trend of gas emission concentration but also has a high fitting effect and prediction accuracy. Keywords—gas emission concentration; time series; ARIMA model; GARCH model; prediction; fitting; R language