Visual Management Of Safety In Coal Mines Based On Data Mining

The coal mines’ information is striding towards the smart mine period after the digital mine and sensor mine stage. Currently a comprehensive perception of the underground individuals, equipments and underground environment has been achieved in coal mines which makes it convenient for the safety managers to get and use safety data. Smart mine with its typical characteristics of thorough perception, deep interconnection and intelligent application puts forward new requirements for the safety management in coal mines with more emphasis on the value of safety information resource. The extraction of useful information and discovery of valuable knowledge from the mass data are the foundation to achieve intelligent safety management in coal mines.The traditional safety management theory is of little help to process the huge amount of safety data. The visual safety management attempt has been carried out not only by the academic researchers but also by the coal enterprises’ managers to solve the predicament with the application of both visualization and data mining techniques. The attempt has got some achievement but is not satisfactory on the whole due to its locality, casualness and one-sidedness. The locality manifests as some of the safety information resource is shown with visualization while the other is not. The casualness manifests as the selection of the visualization does not consider its matching with the information content. The onesidedness manifests as the neglect of safety knowledge visualization. The research issue of visual safety management in coal mines is put forward considering the smart mine’s new requirements and the existing predicament. Firstly, the paper builds up the visual safety management theory system with the specific study on its connotation, mechanism and key models according to the logical thinking process. Then the safety knowledge visualization as the core issue in the visual safety management is studied specifically with the study on the “7W1H” quantitative conversion model and “RPCIA” realization model. The theoretical analysis and empirical study are fully integrated in the research process. The “7W1H” and “RPCIA” model are applied into TF Coal Mines and achieves the expected goals. The paper is implemented by the idea “putting forward the general theory system first and studying the realization later” with seven chapters included as follow:Chapter one: Introduction. In this chapter the stage of coal mines’ information is summarized. The promotion of digital mine and sensor mine to safety management and the new requirements of smart mine for coal mines’ safety management are emphasized. The theoretical exploration and enterprises’ practice on visual safety management in coal mines are reviewed with the conclusion that problems including locality, casualness and one-sidedness exist. So the research issue is put forward with its purpose and significance analyzed. Then the content and the corresponding research method to be adopted are discussed with the technical route given in the end.Chapter two: Theoretical basis research. This chapter gives a review on the previous research including visual management, data mining and safety management. For the visual management area, the theoretical development and its application in safety management are summarized. For the data mining area, the evolution of its process model and the application in coal mines’ safety issues are reviewed. For the safety management area, the theoretical development and its trends are concluded. Through the literature review in related research area at home and abroad, the paper obtains the research achievements and shortcomings. Moreover the paper studies the cognitive science theory specifically on the cognitive information process model, multi-stage memory and the cognitive load theory. The research on cognitive science provides a theoretical foundation to discuss the mechanism of visual safety management.Chapter three: Coal mines’ visual safety management theoretical system research. The visual safety management is put forward with its connotation, characteristics and content discussed to make it clear what it is. Also the mechanism is studied by the theoretical analysis and the experimental design with the guidance of information processing theory and cognitive load theory to make it clear how it works. And the corresponding models are given to make it clear how to realize it. At the end the theoretical framework including the connotation layer, mechanism layer, model layer, application layer and the supporting layer is established.Chapter four: The structured expression model for coal mines’ safety data. The procedure including the structured representation stage, interactive modes stage, model realization stage and the knowledge application stage is analyzed to realize safety knowledge visualization according to the typical knowledge discovery process, the actual safety management requirements and the characteristics of the safety data. The “7W1H” structured expression model is put forward based on the extraction of dimensions for describling coal mines’ safety problems and the analysis on the attributes of each dimension. The “7W1H” model is used to convert the text-type safety problem records into quantitative stytle. An investigation is carried out in TF Coal Mine to collect its safety problem data. With the data preprocessing the quantitative safety problem sets is derived using the “7W1H” model.Chapter five: Interactive analysis within the coal mines’ safety issues. For the univariate analysis of the safety problem object the frequency analysis method is used. The distribution law of the safety problems on different attributes of each dimension is derived. For the multivariate interactive analysis, the interactive content and the corresponding meaning is discussed. The corresponding analysis method and log-linear model are adopted to realize the interactive analysis. The empirical analysis is also done on TF’s safety problem sets with the valuable conclusion drawn.Chapter six: Interactive analysis between coal mines’ safety issue and other issues. Some of the dimensions of safety problem issue are treated as the separate issue. Through the interactive analysis between different issues the deep law of coal mines’ safety problems can be got. For the interactive analysis between the safety and production, the correlation degree between different kinds of safety problems and the production links is drawn with the grey relational analysis method adopted. For the interactive analysis between the safety and the dangerous sources, the paper makes a metaphor that the coal mines’ workplace can be regarded as the dangerous source supermarket. The association rules is mined by the data mining method which reflect that the dangerous sources are prone to generate safety problems together. At the end of this chapter, the safety knowledge visualization RPCIA model is established by summarizing the above research.Chapter seven: Conclusion and prospect. The research conclusions and main innovative points in this paper are summarized. Meanwhile the prospect for the future research on the visual safety management is discussed.The main innovative points are as follows. 1) Reveal of the mechanism of coal mines’ visual safety management with the application of cognitive information process theory and cognitive load theory. 2) The “7W1H” structured expression model for coal mine safety data is put forward. 3) Establishment of the corresponding analysis model and loglinear model for the knowledge visualization with coal mines’ safety issue. 4) Establishment of the grey correlation model to study the relationship between coal mines’ safety and production and the association rule mining model for coal mines’ dangerours sources.