基于数据挖掘技术的事故隐患预警方法研究

发布时间：2018-03-17 21:00

本文选题：事故隐患　切入点：隐患辨识　出处：《中国安全生产科学技术》2017年07期 　论文类型：期刊论文

【摘要】：为改良企业既有隐患排查治理方法,提高辨识效率,提升整改效果,应用并优化数据流滑动窗口模型、等价类变化算法、变化挖掘算法等数据挖掘技术,构建1种关联隐患预警方法,该方法以企业历史隐患数据为基础,通过挖掘实时隐患数据的关联规则,发布关联隐患的类型、存在可能性和变化模式等预警信息,指导企业进行针对性排查和整改。最后,以72家安全管理性质相似企业的28 957项隐患数据为例,验证该方法可行性。研究结果表明:该方法不但能挖掘关联隐患类型及其变化模式,而且预警准确率随置信度水平提升而不断提高。
[Abstract]:In order to improve the existing hidden trouble detection and management method, improve the identification efficiency, improve the effect of rectification, apply and optimize data flow sliding window model, equivalent class change algorithm, change mining algorithm and other data mining technology. Based on the historical hidden trouble data of the enterprise, this method is based on mining the association rules of the real-time hidden danger data, promulgating the type of the association hidden danger, the possibility of existence and the change pattern and so on. Guiding enterprises to carry out targeted investigation and rectification. Finally, taking 28,957 hidden danger data from 72 companies with similar safety management nature as an example, The feasibility of the method is verified. The results show that the method can not only mine the types of association hidden dangers and their changing patterns, but also improve the accuracy of early warning with the increase of confidence level.
【作者单位】：国网吉林省电力有限公司;中国地质大学(北京)工程技术学院;国网吉林省电力有限公司延边供电公司;
【基金】：国家重点研发计划项目(2016YCF0801906)
【分类号】：F273;F426.61;TP311.13

【参考文献】