【摘要】 为降低煤矿井下煤与瓦斯突出事故中的人员伤亡和财产损失，提高突出事故中的应急救援能力，提出一种麻雀搜索算法(SSA)优化支持向量机(SVM)的煤与瓦斯突出事故应急救援能力评估模型。首先，依据相关文献与研究报告构建包括应急预防能力、应急准备能力、应急响应能力和恢复善后能力在内的4项一级指标，其中包括18项二级指标，并以各指标的得分数据作为模型训练数据集；然后，利用网络层次分析法(ANP)与熵权法(EWM)分别确定各评估指标在相互影响下的主客观权重，通过拉格朗日函数将各权重融合得到最优权重，运用SSA算法优化SVM的径向基核参数g和惩罚因子C,将最优权重计算得出的结果作为SSA-SVM模型的输入，期望值作为输出进行线性回归预测；最后，以河北省某矿为例，将SSA-SVM模型与传统SVM、粒子群优化算法(PSO)优化SVM、鲸鱼优化算法(WOA)优化SVM 3种不同模型的预测结果分别与期望值作对比分析。结果表明：SSA-SVM模型的预测结果与实际相符，平均绝对误差相较于其他模型分别下降8.04%、5.15%、4.82%,证明所建模型的优越性，可将其应用于矿山企业实际矿井煤与瓦斯突出事故应急救援能力评估。更多还原

【Abstract】 In order to reduce the casualties and property losses and improve the emergency rescue capability in coal and gas outburst accidents, an SSA optimized SVM was proposed to evaluate the emergency rescue capability of coal and gas outburst accidents. First, according to relevant literature and research reports, four first-level indicators, including emergency prevention ability, emergency preparedness ability, emergency response ability and recovery and rehabilitation ability, were constructed. These indicators were further subdivided into 18 second-level indicators, and the score data of each indicator was used as the model training dataset. Then, the network analytic Hierarchy process(ANP) and entropy weight method(EWM) were used to determine the subjective and objective weights of each evaluation indicator under the mutual influence, and the Lagrange function was used to merge the weights to obtain the optimal weights. SSA optimized the radial basis parameters g and penalty factor C of SVM. The result of optimal weight calculation was used as the input of the SSA-SVM model, and the expected value was used as the output for linear regression prediction. Finally, taking a mine in Hebei Province as an example, the prediction results of the SSA-SVM model was compared with the traditional SVM, particle swarm optimization algorithm(PSO) optimization SVM and Whale optimization algorithm(WOA) optimization SVM, and the predicted results and the expected values were analyzed. The results show that the prediction results of the SSA-SVM model are consistent with the reality, and the average absolute error decreases by 8.04%, 5.15% and 4.82%, respectively, compared with other models, which proves the superiority of the proposed model. This model can be applied to the evaluation of the emergency rescue ability of coal and gas outburst accidents in actual mines.更多还原