采空区煤自燃危险多场信息叠加判定与火源位置反演识别方法

采空区煤自燃灾害是困扰煤矿安全生产的重大难题之一，研究煤自然发火危险性预测、危险区域判定和火源定位方法，是实现煤自燃灾害预防和处置的重要途径。项目主要研究:（1）采用支持向量机对典型煤样自燃过程的特征信息进行分类和拟合，建立煤自燃温度和气体的非线性关系函数，构建基于混合核函数支持向量机的煤自燃危险性预测模型；（2）采用克里金插值法对采空区煤自燃离散参数进行处理，建立采空区煤自燃多参数连续分布场，构建基于多场信息叠加的采空区煤自燃危险区域判定模型；（3）依据采空区火源温度场和气体浓度场时空演化规律，建立基于最优目标函数的隐蔽火源位置反演模型，研究粒子群优化算法迭代搜索火源的方法；（4）构建基于VC++和Matlab融合编程的采空区煤自燃危险判定与火源位置反演识别系统，能够在采空区煤自燃危险判定和火源位置反演识别过程中主动学习和推理。研究结果为煤自燃灾害针对性预防和应急处置决策提供科学依据。

Coal spontaneous combustion disaster in goaf is one of the major problem that plague coal mine safety production. It is an important way to prevent and deal with coal spontaneous combustion disasters by studying the accurate prediction of coal spontaneous ignition risk, judgment of dangerous areas and positioning of fire source. Research contents: (1) Support Vector Machine（SVM） is used to classify and fit the characteristic information of typical coal sample spontaneous combustion process, the non-linear relationship function between coal spontaneous combustion temperature and gas is established. And a model of coal spontaneous combustion risk prediction based on hybrid kernel function support vector machine will be established; (2) The method of Kriging interpolation is applied to process the discrete parameter of coal spontaneous combustion parameters in gob, establishing the multi-parameter continuous distribution field in gob and a determined model of coal spontaneous combustion dangerous area in gob in terms of multi-field information superposition; (3) According to the spatial and temporal evolution law of simulation temperature field and gas concentration field about fire source in gob, establishing position inversion model of hidden fire source in light of optimistic objective function, studying on the method of iterative search for fire source by particle swarm optimization; (4) A system of the coal spontaneous combustion hazard determination and fire source location inversion identification in goaf will be constructed based on VC++ and Matlab fusion programming, realizing the spontaneous combustion hazard determination and the accurate inversion identification of fire source location in goaf, which can carry out active learning and reasoning in the process of coal spontaneous combustion hazard determination and fire source location identification. The results provide a scientific basis for the targeted prevention and emergency disposal of coal spontaneous combustion disasters.