基于支持向量分类机的煤与瓦斯突出危险性预测

Coal and Gas Outburst Prediction Based on Support Vector Classification Machine

为了提高煤与瓦斯突出预测的准确性,提出了基于支持向量分类机的煤与瓦斯突出危险性分类的预测方法。该方法以地应力、瓦斯压力、煤的厚度和煤样筛分模数作为特征值,建立了煤与瓦斯突出危险性预测模型,分别选用单项式、多项式、RBF以及Sigmoid核函数对所选样本训练学习,确定了RBF做为预测突出危险性模型的核函数。

In order to improve the prediction accuracy of outburst hazard, the method of outburst prediction based on support vector classification machine is proposed. Stress, gas pressure, coal seam thickness and screening modulus are used as characteristic value, and prediction model of outburst hazard is established,the Kernel function of monomial,polynomial,RBF and Sigmoid are applied for training sample, RBF is determined as Kernel function for the established predicted model.