论文部分内容阅读
针对煤矿开采中煤与瓦斯突出强度的预测问题,利用免疫遗传算法和最小二乘支持向量机相结合的方法,选取最大主应力、瓦斯压力、瓦斯含量、顶板岩性、距断裂距离、煤层厚度、开采垂直深度、绝对瓦斯涌出量、相对瓦斯涌出量等9个主要影响因素.对相关程度较高的因素进行因子分析,提取公共因子作为IGA-LSSVM模型的输入,建立基于因子分析和IGA-LSSVM的煤与瓦斯突出强度的预测模型.利用实测的14组数据作为学习样本,训练预测模型.另外5组数据作为测试样本,使用所得模型进行预测.研究结果表明:经过免疫遗传算法优化最小二乘支持向量机的参数后,所得模型可有效预测煤与瓦斯突出的强度,检验结果的误判率为0.
According to the prediction of coal and gas outburst intensity in coal mining, the combination of immune genetic algorithm and least square support vector machine is adopted to select the maximum principal stress, gas pressure, gas content, roof lithology, distance to fracture, thickness of coal seam , Vertical depth of mining, absolute gas emission, relative gas emission, etc. Factor analysis is carried out on the factors with higher correlation degree, common factor is extracted as input of IGA-LSSVM model, and based on factor analysis and IGA-LSSVM prediction model of coal and gas outburst intensity.Using 14 groups of measured data as training samples to train the predictive model.Another five groups of data as a test sample, the use of the resulting model to predict.The results show that: After the optimization of immune genetic algorithm The least square support vector machine parameters, the model can effectively predict the intensity of coal and gas outburst, the test results of the false positive rate of 0.