Regional forecast of coal and gas burst based on fuzzy cluster analysis and fuzzy pattern recognition

For complex mechanics of coal and gas outburst,the uncertainty or fuzziness between outburst-factors and outburst-hazard,makes predicting methods both based on experiences and on math-models be limited.A new method in combining the fuzzy cluster analysis and fuzzy pattern recognition is presented to forecast coal and gas burst.Firstly,the simple integration is classified for the coal and gas outburst by adopting clustering analysis method.The fuzzy model is set up for different outburst degree.Afterwards,the danger degree of the undetermined forecasting sample is predicted by applying fuzzy pattern recognition.The accuracy of forecast is improved.Through actual validation,the reliability of the prediction is tested and verified.