COAL AND GAS OUTBRST DYNAMIC PREDICTION IN MINING FACE BY PATTERN RECOGNITION

摘要

Coal and gas outburst risk prediction is the guarantee of coal mine safety production.First, we expounded the principles of coal and gas outburst prediction in working face, and established multivariate pattern recognition criteria and model.Fully consider the influence of human engineering activities on coal and gas outburst, we select gas content, gas pressure, faults, mining stress etc.factors as the main factors of coal and gas outburst, Using the multivariate pattern recognition method to predict coal and gas outburst sub-unit probability.Then, we developed the coal and gas outburst of working face probability prediction system which is based on AutoCAD and VBA.As the working face advancing, the system can be easily and dynamically to predict and manage the risk probability values about each units.Combined with the actual situation in Pingdingshan No.12 coal mine, we predicted the risk probability about working face Ji15-17200.It shows that: The working face has a large outburst danger in the natural state, the risk probabilities are from 0.6 to 0.92; After taken the gas drainage measures, most the risk probabilities are from 0.32 to 0.52, but a small part is still high, from 0.75 to 0.91; The risk probabilities around the stop line are high, from 0.35 to 0.74, some measures should be taken.Compared with the scene, the overall trend of prediction results are good consistent with the actual, which is a good guide for the safety in production.