Abstract:The study of coal and gas outburst with multi-factor pattern recognition method is related to coal mining,safety engineering,geology engineering,geographic information system(GIS),probability theory,information science,and artificial intelligence(AI). Based on theoretical analysis and the relation among the factors that affect coal and gas outburst,the norm and model of pattern recognition are established. Then the design and algorithm of pattern recognition system are completed,on which the probability prediction norms are certain and the development of coal and gas outburst prediction system is completed. With eight factors including active fault,maximal stress,gas pressure,and gas content acting as the main discriminant,the pattern recognition method was used to perform possibility prediction of coal outburst;and the mining area was then divided into coal and gas outburst dangerous area,threaten area and safe area,respectively,to assess and predict the danger of coal and gas outburst. Thus the accuracy of coal and gas outburst prediction is improved. By the method,a comparatively scientific region prediction of coal and gas outburst are built;and it is possible to judge and precontrol the coal and gas outburst.