基于FTA的矿井监控预警诊断知识表示及推理机制  

Knowledge representation and reasoning mechanism of mine monitoring warning diagnostic based on FTA

作　　者：赵丹[1,2] 陈帅[1] 潘竞涛[1,3] 

机构地区：[1]辽宁工程技术大学安全科学与工程学院,辽宁阜新123000 [2]矿山热动力灾害与防治教育部重点实验室,辽宁葫芦岛125105 [3]煤科集团沈阳研究院有限公司,辽宁沈阳110016

出　　处：《中国安全生产科学技术》2015年第6期38-43,共6页Journal of Safety Science and Technology

基　　金：国家自然科学基金项目(51204088)

摘　　要：当前的矿井监测监控系统给出的预警结果只能反映传感器所在位置的参数变化,并不能依此判断传感器所在位置就是故障源。研究矿井监控预警诊断专家系统,通过建立矿井通风系统故障树设计面向故障树的基于框架与规则的混合知识表示方法,利用框架中诊断规则的推理确定故障传播关系,找到故障源。设计了矿井监控预警诊断推理机制,利用监控系统对灾变时期的火灾和瓦斯大量涌出以及正常生产时期矿井中的活塞效应、巷道风阻变化、通风构筑物故障、爆破、通风动力故障以及自然风压变化进行故障诊断推理分析,给出监控预警诊断故障源分析界面,为管理人员对矿井正常生产和灾变时期的故障处理提供依据。The warning result given by current mine monitoring system can only reflect the parameter changes where the sensor is located, and the location of sensor cant be determined as the fault source. The expert system of mine monitoring warning diagnosis was studied, the hybrid knowledge representation method based on framework and rule for fault tree was designed through the establishment of fault tree for mine ventilation system, and the fault propaga- tion relationship was determined utilizing the reasoning of diagnosis rules in the framework, then the fault source was found. The reasoning mechanism of warning diagnostic in mine monitoring was designed, and the fault diagnosis reasoning analysis was completed for fires and gas emission in the disaster period, as well as the mine piston effect, drag changes, ventilation structures fault, explosion, ventilation power fault and natural air pressure changes in normal production period by using monitoring system, the analysis interface of fault source for monitoring warning diagnostic was given to provide evidence for managers to deal with fault in mine normal production and disaster period.

关 键 词：矿山安全 监控系统 预警诊断 专家系统 故障树 

分 类 号：X936[环境科学与工程—安全科学]