Research On Prediction Technology Of Mining Area Pressure Based On Large Data Of Support Working Resistance

During the advancing process of mining face,the analysis and prediction of the strata pressure behavior at the top of the working face has been a difficult problem for the researchers of mine pressure.Based on the huge data of working resistance of working face,according to the general rule of subsection pressure in the top plate section of the working face,the three important parameters,such as the starting position of the roof,the pressure strength of the roof and the step distance,are fully analyzed and predicted in this paper.The nain research contents are as follows.1.Cluster analysis is used to divide the support cycle of the working face,select the support force and the effective resistance of the support,then,based on the working resistance of single frame,the rule of mine pressure is analyzed.2.Define the new concept of the working face area,divide the stents with the close rules to the same area,define the concept of the pressure parameters in the region,and find out the analysis of the law of the regional ore pressure(the starting position of the pressure,the pressure strength,etc.)to press the parameter values.3.According to the results of the analysis of the regularity of regional mining pressure(including the starting position of regional pressure,pressure strength,etc.),a multi linear regional prediction model is designed by machine learning language,and the feasibility of the model is tested by using the hypothesis testing method.4.Using the computer programming C#language,the Windows+.Net Framework4.O+SQL Server 2008 R2express platform is built,and the research model of regional ore pressure analysis and prediction is compiled into software,and it has been successfully applied in the coal mineof Da Liuta.Practice shows that the automatic analysis and prediction software of mine pressure is accurate and has high prediction rate,and has strong practical value and popularization value.