摘要
为加强煤矿安全生产风险预警管理,对煤矿安全生产系统风险因素进行分析,从人-机-环-管理四个方面建立了风险预警指标体系,给出了各指标的定义;提出了指标风险预警等级临界点的设置和指标警度计算方法;应用概率神经网络(PNN)构建了安全生产风险综合预警模型,通过指标风险预警等级临界点构建训练样本,并对预警模型进行了性能检验和工程应用。结果表明,基于PNN的安全生产风险综合预警模型风险识别能力强,运行速度快,计算效率高,可以进行推广应用。
To strengthen the risk early warning management of safety production in coal mine, the early warning in- dex system was established from person quality, production equipment, environmental condition, and safety man- agement through systematic analysis the risk factor in coal mine. This index system include 23 indexes and each in- dex was defined carefully. The early warning level was divided into 5 grades, the critical point of warning grade was put forward for every index, and then, the calculate method was constructed for index warning degree also. The comprehensive model of risk early warning was established for safety production in coal mine based on PNN. The model performance was tested through using the critical point of early warning grades as train sample, and then, the model was applied into 3 coal mines that the recognition accuracy is 100%. The results show that, The PNN model of risk early warning has strong recognition ability and fast running speed and high calculation efficiency, it has good prospect of popularization and application.
出处
《中国安全生产科学技术》
CAS
CSCD
2013年第10期71-77,共7页
Journal of Safety Science and Technology
基金
国家自然科学基金资助(51274100)资助项目
湖南省教育厅科研资助项目(10C0690)
煤矿安全开采技术湖南省重点实验室资助项目(201002)
关键词
风险预警
警度
预警等级临界点
概率神经网络(PNN)
安全
煤矿
risk early warning
warning degree
critical point of warning grade
probabilistic neural network（PPN）
safety
coal mine
