摘要
为探索煤矿顶板事故致因因素并预防煤矿顶板事故,利用相关性分析和K2算法相结合的方法进行贝叶斯网络结构学习,并利用参数估计法进行网络参数学习,建立顶板事故致因分析的贝叶斯网络模型。应用建立的网络模型,分析各因素对事故的影响程度。结果表明:顶板事故的直接影响因素为支护问题、顶板冒落和人员是否进入冒落区;基于贝叶斯网络建模的顶板事故致因分析模型预测精度较高,能用来分析影响事故严重程度的因素。
In order to explore causation factors of coal mine roof accident and offer effective measures for controlling mine roof accident, a Bayesian network model was built for coal mine roof accident analysis after structure and parameter learning were carried out, using correlation analysis, K2 algorithm and Bayesian method. Based on Bayesian network model, effects of causing factors on casualties of mine roof accident were analyzed. Results show that factors affecting directly roof accident are supporting problem, roof fall, personnel in caving area, and that Bayesian network model can express the complicated relation- ship between mine roof accident and causes more accurately.
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2014年第7期10-14,共5页
China Safety Science Journal
基金
国家自然科学基金资助(71271206
71173216)
关键词
顶板事故
贝叶斯网络
安全管理
致因分析
K2算法
roof accident
Bayesian network
safety management
causation analysis
K2 algorithm
