基于多源异构信息融合的煤矿冒顶事故预报技术  

Mine roof fall prediction technique based on multi-source information fusion

作　　者：梁亚婷[1] 杨振宏[1,2] 屈利伟[1] 

机构地区：[1]西安科技大学能源学院,陕西西安710054 [2]西安建筑科技大学材料科学与工程学院,陕西西安710055

出　　处：《西安科技大学学报》2013年第1期78-83,共6页Journal of Xi’an University of Science and Technology

基　　金：国家自然科学基金项目(50174045)

摘　　要：基于D-S理论和模糊数学理论,建立了煤矿冒顶监测系统中多源异构信息融合模型,并提取某矿冒顶危险性数值特征,应用隶属度函数概念得出危险征兆的信度函数分配,最后利用判决规则判断其融合结果并预测该矿冒顶危险等级。该方法能简便快速的预测煤矿冒顶危险,准确识别冒顶危险预兆的目标特征,为矿井冒顶事故预报可以提供依据。According to D - S theory and fuzzy mathematical theory, this paper built a model of multisource information in coal mine roof fall prediction. After obtained function assignment in every symptom occurred, the concept of membership function in fuzzy mathematics is generalized to D-S theory to produce basic probability assignment（BPA） of risk when coal mine roof fall occurs. Apply this method to coal mine roof fall prediction can simply and rapidly identify the roof fall risk, which can accurately describe the target feature in identified roof falling risk, this method has the potential significance for coal mine roof fall prediction.

关 键 词：多源异构信息 D—S理论 信息融合 冒顶 

分 类 号：TD771[矿业工程—矿井通风与安全] TD76