| The method predicting coal seam spontaneous combustion around roadway has not yet been solved successfully, so that it has seriously affected on the development and application of fully-mechanized caving mining technology. On the basis of investigating foreign and domestic literatures and generalizing former research results, the prediction technology is researched by the methods of theoretical analysis, mathematical model and application on the spot. The mathematical models of the coal particle size fuzzy evaluation, neural network predicting for critical parameters of coal spontaneous combustion, dividing of coal spontaneous combustion area and predicting of coal spontaneous combustion are put forward. Firstly, the target system of coal particle size evaluation formed by the geologic structure, the flexibility and hardness of the coal, the pressure indices, the method of coal mining, roadway layout method and support pattern is pointed out and analyzed, and a fuzzy comprehensive evaluation model of coal particle size is established. Secondly, using artificial neural network, the critical parameters are predicted on the basis of analyzing influence factors of the critical parameters of coal spontaneous combustion, and the model of fuzzy dividing of dangerous area for the coal spontaneous combustion is produced by analyzing of the indispensable conditions of the coal spontaneous combustion and the influencing factors of the heat accumulation and scattering in the loose coal of gate, such as temperature of coal, concentration of oxygen, thickness of loose coal, particle size of coal and quantity of supply wind and so on. Finally, the existing mathematical model for coal spontaneous combustion around roadway is complemented and modified. This modified model combined with the forenamed model is successfully applied to the prediction of coal spontaneous combustion around the roadway in No. 4308 workface of DongTan Coal Mine, YanZhou Mining Bureau, which is a fully-mechanized long-wall top-coal caving mining face. |