中文摘要：

针对经验模型与确定性模型在应用中受到限制问题,采用基于统计学习理论的支持向量机对经验数据进行学习,建立瓦斯含量与其影响因素之间的映射模型,从而实现煤层瓦斯含量预测。支持向量机的惩罚因子和核参数取值不同将会明显影响其预测的精度,支持向量机本身也没给出解决的办法,引入粒子群算法自动搜索支持向量机参数。该方法克服了神经网络过学习问题和支持向量机人为选取参数的盲目性问题。通过对某矿区样本的学习预测研究,表明该方法可取得良好的预测效果,具有较好的适应性。

英文摘要：

In-situ gas content in coal seam is affected by many complicated geological factors. Conventional empirical models and deterministic models have a limited capacity in forecasting coal seam in-situ gas content. This paper proposes a new method to forecast in-situ gas content in coal seam. The proposed method adopts support vector machine （SVM）, which is based on statistical learning theory to map the complex nonlinear relationship between in-situ gas content and its influence factors by learning from empirical data. Therefore, the in-situ gas content can be forecasted. Because the penalty factors and kernel parameters of SVM will affect the forecast accuracy and SVM does not provide any mean to determine these factors and parameters, this paper introduces a particle swarm optimization algorithm to automatically search the parameters for SVM. The method overcomes ANN＇s over-learning problem and the human＇s blindness on parameters selection. The method has been applied to forecast the in-situ gas content in Xinwen Coal Mine. The results demonstrate that the method has high adaptability and forecasting accuracy.