中文摘要：

为有效预防瓦斯灾害,以预测矿井瓦斯涌出量为研究目的,提出经改进的蚁群（ACO）粒子群（PSO）混合算法优化的最小二乘支持向量机（LS-SVM）,并用其预测非线性动态瓦斯涌出量。算法通过对LS-SVM的正则化参数C和高斯核参数σ进行寻优,建立了基于蚁群粒子群混合算法优化的瓦斯涌出量预测模型,并根据赵各庄矿矿井监测到的各项历史数据进行实例分析。实验结果表明：该预测模型预测的最大相对误差为1.05%,最小相对误差为0.28%,平均相对误差为0.75%。较其他预测模型拥有更强的泛化能力和更高的预测精度。

英文摘要：

In order to prevent gas disasters effectively and predict mine gas emission, an improved LS-SVM model based on ant colony optimization mixing with particle swarm optimization was presented, which was used to predict nonlinear dynamic gas emission. The regularization C and the Gaussian kernel parameter σ of LS-SVM were optimized by the prediction model of gas emission based on hybrid algorithm of ant colony particle swarm optimization. The model was validated by using the historical data from Zhaogezhuang coal mine in China. The results show that both the maximum and minimum relative errors predicted by the model are 1.05% and 0.28% respectively, and the average is 0.75%. Compared with others, the model has higher generalization ability and predicting precision.