基于GA-SVM的煤层瓦斯涌出里量预测技术研究

发布时间：2018-06-14 17:29

本文选题：安全工程 + 支持向量机　；参考：《辽宁工程技术大学》2015年硕士论文

【摘要】：为了对煤层瓦斯涌出量进行预测,提出将支持向量机(SVM)与遗传算法(GA)相耦合。利用GA寻找SVM最优的惩罚参数c和核函数参数g,并结合SVM具有训练速度快且具有良好泛华性能的特点,建立基于GA-SVM的煤层瓦斯涌出量预测模型。煤层深度、煤层厚度、煤层倾角、开采层原始瓦斯含量、煤层间距、采高、临近层瓦斯含量、临近层厚度、层间岩性、工作面长度、推进速度、采出率、日产量对瓦斯涌出量的影响是复杂的、非线性的,因而将其作为预测的影响参数。将瓦斯涌出量作为目标参数。分别用影响参数和目标参数作为GA-SVM的输入值和输出值进行训练,训练后的预测输出和期望输出之间的误差绝对值倒数作为GA的适应度函数值进行参数优化。研究结果表明：该预测模型预测的最大相对误差为5.91%,最小相对误差为0.92%,平均相对误差为2.2%,相比耦合前及其他预测模型有更强的泛化能力和更高的预测精度,并在铁法集团大平煤矿进行了实际应用,得出预测的最大相对误差为6.80%,最小相对误差为0.47%,平均相对误差为2.89%。证明了该模型具有实际的应用价值。
[Abstract]:In order to predict coal seam gas emission, the support vector machine (SVM) coupled with genetic algorithm (GA) is proposed. The GA is used to find the optimal penalty parameter c and kernel function parameter g of SVM, and combining the characteristics of SVM with fast training speed and good flooding performance, a prediction model of coal seam gas emission based on GA-SVM is established. Coal seam depth, coal seam thickness, seam dip angle, original gas content in mining layer, coal seam spacing, mining height, adjacent gas content, adjacent layer thickness, interlayer lithology, working face length, advance speed, extraction rate, The influence of daily production on gas emission is complex and nonlinear, so it is regarded as the influence parameter of prediction. The quantity of gas emission is taken as the target parameter. The influence parameters and target parameters are used as the input and output values of GA-SVM, and the inverse of the absolute error between the predicted and expected outputs is optimized as the fitness function of GA. The results show that the maximum relative error is 5.91, the minimum relative error is 0.92 and the average relative error is 2.2. The prediction model has a stronger generalization ability and higher prediction accuracy than before coupling and other prediction models. The results show that the maximum relative error of prediction is 6.80, the minimum relative error is 0.47 and the average relative error is 2.89. It is proved that the model has practical application value.
【学位授予单位】：辽宁工程技术大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD712.5

【参考文献】