基于证据融合理论的煤矿顶板安全评价模型

Model for safety evaluation of coal mine roof based on evidence fusion theory

针对顶板监测数据来源的复杂性及不确定性,由各监测系统的传感器获得不同信息,根据设定的阈值产生对顶板安全状态的度量,通过构造的基本概率分布函数,对所有的命题赋予一个可信度,应用基于Dempster-Shafer证据理论的合成法则进行多证据融合后,即可凸显顶板安全状态。通过与历史数据对比分析,发现该模型能显著降低监测系统的虚警率,提高顶板安全状态判定的准确性。

Aiming at the complexity and uncertainty of mine roof monitoring data source,the sensors of monitoring system got different information and the measure of roof safety according to the pre-set threshold values,assigned a reliability for all propositions by constructing the basic probability distribution function.After multi-evidence amalgamation that applied combination rule of Dempster-Shafer evidence theory,it showed the mine roof safety status.Historical data analysis proves that the model can significantly reduce the roof safety monitoring system probability of false alarm,and increase the accuracy of determining the roof safety status.