我国煤矿生产安全事故统计分析及预测

发布时间：2018-03-27 13:57

本文选题：煤矿安全　切入点：统计分析　出处：《西南科技大学》2017年硕士论文

【摘要】：基于国家煤矿安全监察局事故查询系统所公布的煤矿生产安全事故信息,汇总、统计了2001~2015年我国煤矿生产安全事故数据。结合该数据离散型强、数据量少以及非线性等特点,从煤矿事故发生的年度、季度、月份、种类、地域和死亡事故率方面,分别运用统计图表法、排列图法、管理图法、显著性检验法、主成分分析法、系统聚类法分析了事故发生的特征。根据事故发生的不同特征,寻找事故发生的规律,为我国煤矿安全生产制定科学的预防和控制措施提供参考。根据我国煤矿生产安全条件复杂多变的特点,分别选用了时间序列预测法中的ARIMA模型和指数平滑预测法预测我国煤矿生产安全事故死亡人数和发生次数;选用了支持向量回归机预测法预测我国某大型国有煤矿生产安全事故总损失。这些方法在预测我国煤矿未来安全形势上,体现了较高的准确性,为预防事故决策提供了支持。
[Abstract]:Based on the information of coal mine production safety accident published by the accident inquiry system of the State Bureau of Coal Mine Safety Supervision, the data of coal mine safety accidents in China from 2001 to 2015 are collected and analyzed. In terms of year, quarter, month, type, region and death rate of coal mine accidents, statistical chart method, arrangement chart method, management chart method, significant test method, principal component analysis method were used respectively. The systematic clustering method is used to analyze the characteristics of the accident. According to the different characteristics of the accident, the regularity of the accident is found. It provides a reference for the establishment of scientific prevention and control measures for coal mine safety production in China. According to the characteristics of complex and changeable production safety conditions in coal mines in China, The ARIMA model and exponential smoothing method are used to predict the number and frequency of coal mine safety accidents in China. The prediction method of support vector regression machine is used to forecast the total loss of production safety accidents in a large state-owned coal mine in our country. These methods reflect higher accuracy in predicting the future safety situation of coal mines in our country and provide support for the decision-making of accident prevention.
【学位授予单位】：西南科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TD771

【参考文献】