论文部分内容阅读
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 causal factors of coal mine roof accidents and prevent coal mine roof accidents, Bayesian network structure learning is carried out by a combination of correlation analysis and K2 algorithm, and the parameter estimation method is used to study network parameters to establish causal analysis of roof accidents Bayesian network model. Apply the established network model to analyze the impact of various factors on the accident. The results show that the roof impact factors are the support problems, roof caving and personnel entering the caverns. The prediction model of causal analysis of roof accidents based on Bayesian network modeling can be used to analyze the impact of accidents Severity factor.