煤矿事故背后都隐藏着大量的险兆事件，且不同的险兆事件都可能会衍化成不同危害程度的事故。若对所有的险兆事件都采取相同的管理措施，不仅会降低险兆事件管理效率，而且会增大事故预防的工作量和难度。同时，目前针对煤矿险兆事件的分级管理尚在探索阶段。因此，为实现煤矿险兆事件的科学分级和精准管理，提高煤矿险兆事件管理水平，达到煤矿事故“超前管控”的目的。本文将针对煤矿险兆事件分级管理模式和综合风险评估开展深入研究。 首先，基于险兆事件的特征界定了煤矿险兆事件定义并分析了其致因机理；依据霍尔三维系统，构建煤矿险兆事件分级管理系统框架；在风险矩阵理论的基础上引入敏感性维度，建立了险兆事件风险的“三维”分级模型，通过对各个维度科学赋值，计算得到煤矿险兆事件风险的分级标准；基于戴明循环（PDCA）模式，设计了策划模块、实施与运行模块、检查模块、改进模块等4个模块的具体内容，以此构建出煤矿险兆事件分级管理模式。 其次，通过文献调研、人员访谈、案例分析和扎根理论等方法确定煤矿险兆事件风险“人因类影响因素”、“机械类影响因素”、“环境类影响因素”、“管理类影响因素”等4个核心范畴包含的15个影响因素；进一步对影响因素进行提炼和德尔菲法修正，构建出包含4个一级指标和21个二级指标的煤矿险兆事件综合风险评估指标体系。 最后，在综合风险评估指标体系的基础上，提出了确定指标权重的直觉模糊层次分析法，结合TOPSIS理论建立了煤矿险兆事件综合风险评估的直觉模糊层次TOPSIS模型。将该模型应用于煤矿A的险兆事件综合风险评估，通过MATLAB编程运算得到煤矿险兆事件综合风险分级标准，评估出煤矿A的险兆事件综合风险值为0.57，风险等级为Ⅱ级（中度风险），并根据评估结果针对性的提出煤矿险兆事件管理措施。 研究结果表明，本文构建的煤矿险兆事件分级管理模式和险兆事件综合风险评估模型具有较好的实用性和可操作性，可用于险兆事件的分级管理和综合风险评估，能为煤矿险兆事件的管理和事故预防提供决策依据和理论指导。

There is a lot of near miss before coal mine accidents, and different near miss may evolve into accidents with different degrees. If we adopt the same management measures for all near miss, it will not only decline the efficiency of near miss management, but also increase the workload and difficulty of accident prevention. At the same time, the classification management of near miss in coal mine is still in the exploratory stage. Therefore, in order to achieve the scientific classification and accurate management of coal mine near miss, improve the management level of coal mine near miss and achieve the goal of "advanced management and control" of coal mine accidents. In this paper, the classification management model and comprehensive risk assessment of coal mine near miss will be studied in depth. Firstly, based on the characteristics of near miss, the definition of coal mine near miss was defined and the cause mechanism was analyzed. Then, based on the Hall three-dimensional system , the system of hierarchical management of coal mine near miss was constructed. The sensitivity dimension was introduced on the basis of risk matrix theory to establish the "three-dimensional" classification model of risk of near miss. Through scientific evaluation of each dimension, the classification standard of the risk about near miss in coal mines was obtained. Based on the PDCA model, the concrete contents of four modules, P-planning module, D-implementation and operation module, C-inspection module and A-improvement module, were designed to construct the grading management model of coal mine near miss. Secondly, by means of literature research, personnel interview, case analysis, grounded theory, the four core categories of mine near miss’s risk were determined: "human factors influencing factors", "mechanical factors influencing factors", "environmental factors influencing factors" and "management factors influencing factors", which contained 15 influencing factors; a comprehensive risk assessment index system of coal mine near miss including 4 first-level indicators and 21 second-level indicators was constructed through preliminary extraction and expert correction of Delphi method. Finally, based on the comprehensive risk assessment-making index system, the intuitionistic fuzzy analytic hierarchy process (AHP) was established to determine the index weights, and combined with the TOPSIS theory, the intuitionistic fuzzy hierarchical TOPSIS assessment model for the comprehensive risk grading of coal mine near miss was established. The model was applied to the comprehensive risk level assessment of the coal mine A's near misss The comprehensive risk grading standard of the mine near miss was obtained through MATLAB programming calculation, and the comprehensive risk value of the risk near miss of mine A was determined to be 0.57. The risk level was Level II (moderate risk). According to the assessment-making results, the management measures of coal mine near miss were proposed from the perspective of man-machine-ring-tube. The results showed that the classification management model and comprehensive risk assessment model of coal mine near miss constructed in this paper have good practicability and operability, can be used for the classification management and comprehensive risk assessment of near miss, and can provide decision-making basis and theoretical guidance for the management of coal mine near miss and accident prevention.