基于Bayes判别方法的矿井通风系统评价研究

矿井通风技术是保证井下安全生产的前提条件,高精度的通风系统可靠性评判可以改善通风环境和预防事故的发生。为了实现对矿井通风系统的快速和准确的评价,考虑矿井通风系统的技术、监测、能力等几个方面,选取16项评价指标。借鉴一种多元统计方法,建立了矿井通风系统Bayes判别分析评价模型,选取18组矿井通风数据作为学习样本进行训练和检验。结果表明:在各总体协方差矩阵不全相等的情况下,5种不同的训练和测试样本个数下的Bayes判别分析模型仍具有较好的评价效果。其中,回判法中模型的正判率均为100%,交叉确认法中模型的正判率分别为83.33%、94.44%、88.89%、94.44%和88.89%,该模型可为矿井通风系统可靠性评价提供借鉴。

Study of Mine Ventilation System Assessment Based on Bayes Discriminant Methods

Mine ventilation technology is a prerequisite for ensuring the underground production safety, High-accuracy mine ventilation reliability evaluation can improve ventilation environment and prevent accidents from happening. In order to realize rapid and accurate evaluation of mine ventilation system, the article selects 16 evaluation indices in terms of the technologies, monitoring and abilities. The author refers to a multiple statistic method, sets up a Bayes discriminant analysis and evaluation model, and selects 18 groups of mine ventilation data as learning samples for training and testing. The results show that: when the total covariance matrixs are not equal, Bayes discriminant analysis model of five different training and tested samples still has good evaluation effect. The correct judgment rate in rejudgement method is 100%, while the correct judgment rate of model with cross-confirmation method is 83.33%, 94.44%, 88.89%, 94.44%, and 88.89% respectively. The model provides reference for the reliability evaluation of mine ventilation system.