Study On Prediction Model Of Residual Gas Content In Protected Coal Seam

Mining protective layer is a common measure to prevent gas accidents in coal mine production,the mining protective layer can make the coal seam in the pressure relief area expand, the fracture develops, the permeability coefficient increases, the gas in the coal bed is analyzed in a large amount, which is beneficial to the gas extraction.The residual gas content of coal seam is one of the important indicators of the protection layer mining effect test, the main method of coal seam gas content detection is the direct measurement or underground coal samples were collected in the laboratory of indirect measurement,But it can not get the data of coal seam gas content and there are some errors in the measurement.Based on this, the SPH method is used to simulate the gas in the protective layer of the through hole drilling in this paper, the gas migration of the protective layer and the factors affecting the residual gas content of the protective layer are studied, the main controlling factors of residual gas content in coal seam are obtained by grey correlation analysis,then through the BP neural network process the access to data and numerical simulation of data,the BP neural network based on grey relational analysis is used to predict the residual gas content of the protective layer,the residual gas content can be obtained in real time.conducive to the safety of coal mining.The main work of this paper is as follows:(1)According to the theory of gas seepage, the SPH model of coal seam gas is established, analytical solution and SPH solution for gas pressure and emission in homogeneous coal seam, this paper analyzes the accuracy and feasibility of the numerical simulation of gas seepage in coal sam by SPH method, and studies the general law of gas migration in the seam of the protected seam.(2)The gas flow pattern of the protected coal seam was extracted by the SPH method, and the gas flow law in the protected coal seam was simulated under different initial gas pressure, permeability coefficient and negative pressure.studied the relationship between the gas flow rate and the residual gas content,and get the fitting parameters m and n with the coal seam gas pressure becomes larger, its value will increase, and vice versa; when the permeability coefficient of coal seam increases, the value of m will increase, and the absolute value of n will be reduced.And the feasibility of using the law of gas emission to inverse the residual gas content in coal seam is analyzed.(3)The operation principle and algorithm of BP neural network are analyzed, then the sample data needed to construct the neural network is collected,finally, the input layer of the neural network is selected by grey correlation analysis.Determine the input layer BP neural network to predict the protected layer of residual gas content,the number of nodes in hidden layer and output layer are respectively 5,7,3, transfer function. training function and learning function respectively, logsig function,trainlm function and leangm function.(4)The realization and analysis of the model for predicting the residual gas content in the protective layer, through a series of preparatory work, the neural network is trained by MATLAB software,the obtained data into the model is tested,through the analysis of the error rate, the error rate of the model is within the acceptable range, thus, the prediction model of the residual gas content in the protective layer is obtained.So.the residual gas content of the protective layer can be detected in time to prevent the occurrence of gas accidents.