| In recent ten years,the number of coal mine safety accidents and the million-ton mortality have achieved a "double decline" year by year.Nevertheless,coal mine safety accidents still happen frequently,so coal mine safety needs to be paid attention to.The realization of a long-term mechanism for the investigation and control of coal mine hidden dangers and the construction of an information system for the detection and control of coal mine accident hidden dangers are conducive to the prevention of accidents caused by the failure of the investigation and control of coal mine hidden dangers.In this paper,through the analysis of the theoretical study of coal mine accident hidden dangers,combined with the actual situation of Hongyuan coal industry,in-depth study the framework of the coal mine accident hidden dangers investigation and management system and text mining text data analysis and application.The main research content is as follows.(1)this paper studies the theory of work safety accident hidden dangers,which can effectively reduce the dangerous state of accident factors and control the occurrence of accidents.This paper expounds the classification and identification methods of accident hidden dangers,and introduces the PDCA closed-loop management theory.A framework of the coal mine accident hidden dangers investigation and control system was established,which also paved the way for the research and exploitation of coal mine hidden dangers investigation information system.(2)at present,the function of the information system of coal mine accident hidden dangers is mainly focused on the control level of the process of hidden dangers detection and control,ignoring the value research of mining hidden danger text data to reveal the weak link of coal mine hidden danger management.Pretreatment of coal mine accidents,using R language jiebaR package hidden text participle,builds the coal mine accidents user dictionary,expands the hidden danger stop list,and by using the vector space model to represent hidden text information.(3)by means of text mining,the weak links in the mine accident hidden dangers investigation and management are studied by using topic mining and association rule analysis..Through K-Means text clustering and LDA theme model algorithm to study the general situation of hidden danger topics,the advantages and disadvantages of the two algorithms are analyzed and compared.In order to enhance the specific form and visual effect under the hidden danger themes,Apriori algorithm is used to mine the hidden dangers information association rules.The results show that the LDA thematic model and Apriori association rule algorithm can fully mine and utilize the data of historical hidden dangers and effectively expose the shortcomings in the mine hidden dangers investigation and management.(4)the software of information system for mine accident hidden danger detection based on text mining is developed,and the functions of the system are introduced.This paper also carries out hidden danger topic analysis for the latest 1000 hidden danger texts in Hongyuan coal industry Fenghuangtai coal mine accident hidden dangers investigation information system,and the data show that the mine concerns mainly concentrated in the " working surface"," hydraulic support" and " safety protection" and " damper waterfall" four weak link,accounted for 18%,15%,16% and 11% respectively,and the hidden theme under specific forms of visualization,and according to the result of text mining to site link specific guidance.Since the popularization of the information system for mine accident hidden danger detection in Hongyuan coal industry,it has significantly improved the ability and efficiency of mine accident hidden danger detection and management,and effectively guaranteed the safety of life and property of coal mine workers and the safe and efficient production of coal mine.There are 58 graphs,16 tables and 83 references. |
PDF Full Text Request