Study On Early Warning Technology Of Coal Spontaneous Combustion In Goaf Of Linhuan 9 Coal Seam

Spontaneous combustion of residual coal in goaf is one of the main disasters in coal mine.The occurrence of spontaneous combustion of coal not only brings economic losses,but also endangers the life safety of personnel.There are more residual coal in the goaf of 9136 face of coal 9 in Linhuan mine,and the complicated air leakage environment greatly increases the risk of spontaneous combustion of residual coal in the goaf.In order to understand the spontaneous combustion process of goaf residual coal clearly and make a quick response to the spontaneous combustion,it is necessary to carry out the early warning and forecast for the spontaneous combustion of goaf residual coal.The spontaneous combustion characteristics of coal were studied by the low-temperature oxidation experiment of coal samples,and the output law of gas with temperature under different air flow conditions during the oxidation and heating process of coal was obtained:the oxygen concentration decreased with the increase of temperature,and the oxygen concentration was proportional to the air flow at the same temperature before 240°C.The concentrations of CO and CO₂ show a rapid upward trend with the rise of coal temperature under different air flows,and in most coal temperature ranges,the concentrations of both gases are inversely proportional to the flow rate;CH₄ concentration rises first and then decreases,while the concentrations of C₂H₄,C₂H₆ and C₃H₈ all rise first and then tend to stabilize,and hydrocarbon gases have low air flow conditions and large concentration values.The maximum and minimum exothermic intensity and activation energy of the coal under the three air flow conditions were calculated.According to the characteristics of the change of coal low temperature oxidation experimental gas,exothermic intensity and activation energy with coal temperature,the three characteristic temperatures of the coal spontaneous combustion process（70°C,140°C,220°C）are obtained,and the coal spontaneous combustion process is divided into four stages by the characteristic temperature.The normal temperature oxidation of the original coal,the inlet and return gas of the working surface,the gas and temperature of the goaf area,and the upper corner gas were studied,and the gas types of the above items were obtained,and it was clear that the upper corner CO was produced by the oxidation of the coal left in the goaf area.The upper corner CO critical value is calculated,the primary critical value is 8ppm,and the secondary critical value is 63ppm.The early warning indicators of R₁,R₂,C₂H₆/CH₄,C₂H₄/C₂H₆,C₃H₈/C₂H₆ and other early warning indicators with the oxidation and heating of coal were calculated,and the early warning indicators of different coal spontaneous combustion stages were selected by using the gray correlation theory,and the early warning indicators and their critical values were finally determined by combining temperature,CO value of the upper corner angle,and the initial temperature and concentration value of C₂H₄.Based on the study of coal spontaneous combustion stage and early warning indicators,coal spontaneous combustion early warning is divided into four grades（blue,yellow,orange and red）.The first-level（blue）and second-level（yellow）early warning indicators are the upper corner CO,R₁,C₂H₆/CH₄ and T,the third-level（orange）early warning indicators are C₂H₄,R1,C₂H₄/C₂H₆ and T,and the fourth-level（red）early warning indicators are C₂H₄,R₂,C₃H₈/C₂H₆ and T.Through the study of artificial neural network,on the basis of the BP neural network spontaneous combustion warning of coal,the genetic algorithm（GA）is used to optimize the threshold and weight of the neural network,and the R² is increased from 0.9792 to 0.9997,which greatly improves the accuracy of prediction and early warning.Finally,by analyzing the requirements and functions of the Goaf Coal Spontaneous Combustion Early Warning App,the Matlab App designer was used to develop an app that encapsulates the GA-BP neural network,which realizes the functions of convenient early warning,rapid drawing of early warning indicators and time diagrams,and storage and export of early warning records.Figure 51 Table 13 Reference 91...