煤矿瓦斯爆炸事故是井下最严重的灾害之一，具有隐蔽性、突发性和巨大破坏性的特点，严重威胁井下作业人员的生命安全。本文针对煤矿瓦斯爆炸风险耦合关系定性定量分析不全面和风险评估过程的主观性及模糊性问题，开展了煤矿瓦斯爆炸耦合风险评估和风险解耦策略研究。主要研究工作如下：

首先通过分析我国2010年至今的98起瓦斯爆炸事故案例进行煤矿瓦斯爆炸风险致因分析和致因因素识别。将瓦斯爆炸事故案例进行扎根理论三级编码，共得出98个初始范畴、20个主范畴、4个核心范畴，根据编码结果得出煤矿瓦斯爆炸风险致因因素。

其次建立5层解释结构模型对风险因素间的耦合关系进行定性分析，模型拓扑结构表明底层管理因素是导致人、机、环中间层风险因素出现的根本原因，也是导致煤矿瓦斯爆炸事故发生的最深层次的原因；通过N-K模型对煤矿瓦斯爆炸风险耦合关系进行定量计算，得出双因素耦合中机-管耦合程度最高，三因素耦合中人-机-环耦合程度最高。

然后提出一种基于贝叶斯网络的煤矿瓦斯爆炸耦合风险评估方法，并以山西某矿为例进行煤矿瓦斯爆炸耦合风险评估，根据解释结构模型建立贝叶斯网络结构，利用云模型逆向发生器计算各网络节点的先验概率和条件概率，将先验概率和条件概率导入Genie软件中进行贝叶斯网络参数学习，结果得出煤矿瓦斯爆炸发生概率为2%。通过贝叶斯网络的诊断推理和敏感性分析功能，分析影响该矿瓦斯爆炸的关键致因。

最后引入解耦思想，提出煤矿瓦斯爆炸耦合风险解耦策略。基于瓦斯爆炸风险耦合关系定性定量分析结果及贝叶斯网络评估结果，结合解耦思想，建立煤矿瓦斯爆炸风险解耦模型。根据风险解耦模型提出了煤矿瓦斯爆炸风险耦合前、中、后三个阶段的解耦策略，为煤矿瓦斯爆炸耦合风险管理提供对策支持。

Coal mine gas explosion accident is one of the most serious disasters in the underground, with the characteristics of concealment, suddenness, and huge destruction, which seriously threatens the life safety of underground workers. Aiming at the incomplete qualitative and quantitative analysis in the process of coal mine gas explosion risk coupling analysis and the subjectivity and fuzziness in the risk assessment process, this paper conducted research on coal mine gas explosion coupling risk assessment and risk control strategies. The main research work and conclusions are as follows:

Firstly, by analyzing 98 gas explosion accident cases in China from 2010 to present, the causes of coal mine gas explosion risk are analyzed and the causal factors are identified. Using gas explosion accident cases as raw data, a three-level coding based on grounded theory was conducted, resulting in a total of 98 initial categories, 20 main categories, and 4 core categories. Based on the coding results, the risk factors of coal mine gas explosion are obtained.

Secondly, a 5-layer interpretative structure model is established to qualitatively analyze the coupling relationship between risk factors. The model topology shows that the underlying management factors are the root cause of the occurrence of risk factors in the middle layers of human, machine, and environment, as well as the deepest cause of coal mine gas explosion accidents; The N-K model is used to quantitatively calculate the risk coupling relationship of coal mine gas explosion. The calculation results show that the highest degree of machine pipe coupling is found in the dual factor coupling, and the highest degree of human machine ring coupling is found in the three factor coupling.

Then, a Bayesian network based coupling risk assessment method for coal mine gas explosion is proposed. Taking a coal mine in Shanxi Province as an example, a Bayesian network model is established based on the interpretative structural model. The prior probability and conditional probability of each node are calculated using a cloud model inverse generator, and the prior probability and conditional probability are imported into Genie software for Bayesian network parameter learning, The result shows that the probability of coal mine gas explosion is 2%. Through the diagnostic reasoning and sensitivity analysis functions of Bayesian networks, the key causes of gas explosions affecting the mine are analyzed.

Finally, the decoupling idea is introduced and a decoupling strategy for the coupling risk of coal mine gas explosion is proposed. Based on the qualitative and quantitative analysis results of gas explosion risk coupling relationship and Bayesian network evaluation results, combined with the decoupling idea, a decoupling model of coal mine gas explosion risk is established. Based on the risk decoupling model, a decoupling strategy for three stages of coal mine gas explosion risk coupling, namely before, during, and after coupling, was proposed to provide policy support for coal mine gas explosion coupling risk management.