煤炭资源是我国主要的基础能源，对国民经济的发展起着至关重要的推动作用，而煤矿安全是保证煤炭行业稳步发展的重中之重。本文研究目的是提出一种科学的煤矿安全风险评级计算模型，为煤矿安全分级监管监察提供理论指导及实施建议，以解决我国目前煤矿安全监管监察需求与监管监察资源不匹配的问题，从而提高我国煤矿安全监管监察效能。 首先，以三类危险源理论为指导思想，分析并确定了煤矿安全风险的主要影响因素，通过专家咨询、现场调研等方法筛选出地质条件、人的因素、设备因素、环境因素、安全管理这5个二级指标及32个三级指标，建立了煤矿安全风险评价指标体系，并对相关指标含义及量化方法做出了详细说明。 然后，构建煤矿安全风险评级计算模型。运用基于博弈论的组合赋权法将层次分析法得到的主观权重和熵权法得到的客观权重结合算出综合权重；运用TOPSIS法计算得出样本贴近度矩阵，与权重向量相结合计算出安全风险评价向量；通过对安全风险评价指标的分级实现煤矿安全风险的分级。经计算所得安全风险评价向量R=[0.1533，0.2653，0.6940，0.8465]，故提出以下煤矿安全分级标准：当r<0.1533时，安全等级为Ⅰ级（不安全）；当0.1533≤r<0.2653时，安全等级为Ⅱ级（较不安全）；当0.2653≤r<0.6940时，安全等级为Ⅲ级（一般安全）；当0.6940≤r<0.8465时，安全等级为Ⅳ级（较安全）；当0.8465≤r时，安全等级为Ⅴ级（安全）。 最后，选取了陕西省内的四家煤矿进行实证研究。经计算得到四家实例煤矿的安全风险评价向量R=[0.4141，0.3502，0.3002，0.7934]，因此，安全等级综合排名为D矿>A矿>B矿>C矿。根据煤矿安全分级标准，A矿、B矿、C矿的安全等级都属于Ⅲ级，即一般安全状态，D矿的安全等级属于Ⅳ级，即较安全状态。通过对比分析提出建设性的改进意见，并制定出煤矿安全分级监管监察策略和分级监管监察保障措施。

Coal resources are an important basic energy source in our country and play a major role in promoting the development of social economy. Coal mine safety is an important task to ensure the steady development of the coal industry. The purpose of this paper is to propose a scientific calculation model of coal mine safety risk grading in order to provide theoretical basis and specific operational recommendations for coal mine safety grading supervision. Coal mine safety supervision resources do not match regulatory monitoring requirements will be solved, and the efficiency of coal mine safety supervision will also be improved. Firstly, based on the theory that three types of dangerous source, factors influencing coal mine safety risk were analyzed, and determined the main influencing factors of coal mine safety risk. In detail, the geological conditions, human factors, equipment factors environmental factors and safety management and 32 third-level indicators are screened by expert analysis and on-site investigation. The five secondary indicators and 32 third-level indicators have established a coal mine safety risk assessment index system. Further more, the meaning and quantitative methods of the indicators are be elaborated. Then, a coal mine safety risk rating calculation model was proposed. Subjective weighting cofficiene by analytic hierarchy process and objective weighting cofficiene by entropy weight method are combined to calculate comprehensive weighting cofficiene using the combination weighting method based on game theory; the sample closeness matrix calculated by the TOPSIS method is combined with the weight vector to calculate the security risk evaluation vector; the coal mine safety classification model is realized by grading the safety risk evaluation indicators. The calculated safety risk assessment vector R=[0.1533,0.2653,0.6940,0.8465], it can be known that when r<0.1533, the safety level is I (unsafe); when 0.1533≤r<0.2653 The safety level is II (less safe); when 0.2653≤r<0.6940, the safety level is III (medium safety); when 0.6940≤r<0.8465, the safety level is IV (safer); when 0.8465 ≤ r, the safety level is V (safe). Finally, four coal mines in Shaanxi Province were selected for empirical research. The safety risk assessment vector R=[0.4141, 0.3502, 0.3002, 0.7934] of the four sample coal mines was calculated, and the comprehensive ranking of safety risks is D > A > B > C. According to coal mine safety grading standards, the mine A, B and C belong to the third grade, that is, the medium safety state. The D mine belongs to the fourth grade, that is, the safer state. Then, put forward constructive improvement suggestions through comparative analysis, and formulate strategies and safeguard measures for coal mine safety grading supervision.