Reservoir Pressure of Coal-Bed Methane Prediction Research Based on Analysis Method by Neural Net-work

摘要

In order to achieve accurate quantitative results of parameters for reservoir pressure of coal-bed methane, neural network prediction analytic method is adopted to predict the reservoir pressure of coal-bed methane. The main controlling factors such as conformation Stress, buried depth, in-situ stress and permeability are investigated. Mathematical models of neural network of reservoir pressure of coal-bed methane of mathematical analysis and system architecture are established; taking the Qinshui Basin coal seam as example to forecast and use reservoir pressure of coal-bed methane. Projections show that: the use of neural network prediction of reservoir pressure of coal-bed methane is feasible; neural network method makes up a mathematical point of linear and regularity in order to solve the non-linear complex relationship between the input and output parameter variables.