摘要
针对目前机械制造企业安全预警系统中预警指标难以量化确定状态、预警级别生成算法单一且主观性较强的问题,以我国《机械制造企业安全生产标准化评定标准》为基础,建立了机械制造企业安全预警指标体系,并根据贝叶斯网络推理预警级别,在FullBNT中建立了机械制造企业安全预警系统的贝叶斯网络,提出一种运用BP神经网络量化确定机械制造企业安全预警指标状态的方法。通过实例分析表明:该方法可充分利用专家和以往事故案例经验,且预警效率较高。
The early warning indicators in the safety early warning system of machinery manufacturing enterprises are difficult to quantify, the state is hard to determine, and the early warning level generation algorithm is single and subjective.To solve these problems, this paper establishes the safety early warning index system of machinery manufacturing enterprises based on the Standardization of Safety Production for Machinery Manufacturing Enterprises.According to Bayesian network reasoning and early warning level,the paper establishes the Bayesian network of early warning system in FullBNT,and proposes a method to quantify the state of early warning indicators by BP neural network.The example analysis shows that the method can make full use of the experience of experts and past accident cases and have high early warning efficiency.
作者
李江澜
李欢
LI Jianglan;LI Huan(Wuhan Institute of Digital Engineering,Wuhan 430205,China)
出处
《安全与环境工程》
CAS
北大核心
2020年第1期152-157,165,共7页
Safety and Environmental Engineering
关键词
机械制造企业
安全预警
BP神经网络
贝叶斯网络
machinery manufacturing enterprises
safety early warning
BP neural network
bayesian network