Research On Disaster Early Warning Method Based On Gas Emission Law Of Working Face

The gas accident has threatened the life of underground workers and seriously affects the healthy development of China's coal industry.Full and effective mining of coal mine gas monitoring data will provide a basis for accurate identification and disaster response of gas disasters,thereby greatly improving the prevention and control capabilities of coal mine gas disasters.In this paper,the gas monitoring data of a mine test face is taken as a sample to determine the gas emission law of the test face.Through the dimension analysis method,the dimensionless number describing the gas emission from the working face is obtained,and the gas trend analysis based on the K-line principle is proposed.A gas trend analysis method based on K-line principle is proposed.The short-term gas emission prediction based on time series prediction is used to realize the disaster warning based on the gas emission from the working face.Mainly conducted the following researches:(1)In order to establish a universal index for the variation of various coal thickness,coal seam gas content,coal mining machine operating speed,etc.,the dimensional analysis method is adopted,based on the original gas content,coal seam thickness and real-time equivalent propulsion of the working face.Based on the factors such as speed,a dimensionless index is established to characterize the gas emission velocity of the working face.The dimensionless index is used to characterize the gas emission from the mining face.(2)Established an index model based on the K-line theory to visually analyze the changing laws and trends of gas emission.Firstly,the K-line theory is applied to the dimensionless number of gas concentration and gas emission in the experimental mine.It is concluded that the K-line fluctuation curves of the two data are roughly the same.Secondly,it is proposed to use the MA,the ATR,the MACD and the BOLL indexes in the K-line theory to characterize the gas emission law.The results show that the proposed index can accurately reflect the gas fluctuation range,fluctuation scope and fluctuation trend.It can provide guidance for gas warning methods.(3)Predicting the gas emission concentration and gas emission trend of the test face in the future by ARIMA prediction method and index prediction method,aiming to more clearly quantify the gas emission changes in the future.Using the root mean square error and the mean absolute percentage error and other indicators to measure the accuracy of the gas emission prediction results,a gas emission prediction method suitable for the working surface is preferred.The results show that the prediction accuracy of ARIMA prediction method is relatively high,and it is suitable to use ARIMA prediction method to predict the gas emission concentration of working face in a short time.