论文部分内容阅读
利用电测深法能够获得半衰时St/2、衰减度D和视极化率ηs这些与煤系地层含水量相关的指标,通过分析,再引入含水层相对因子参数T"这一指标,选取以上4个物探观测指标作为预测煤系地层含水量的输入参数。通过实例,分别采用多元线性回归模型、人工神经网络模型和最优组合预测模型来预测煤系地层的含水量,研究各个模型的预测精度。结果表明:最优组合预测模型的预测精度最高,证明采用最优组合预测模型预测煤系地层含水量的准确性和实用性。
By using electrical sounding method, we can get the indexes related to the water content in the coal measures, such as the half-decay time St / 2, the decay degree D and the apparent polarizability ηs. By analyzing and re-introducing the indicator of relative factor T , The above 4 geophysical parameters are selected as the input parameters for predicting the water content of the coal measures.Multiple linear regression models, artificial neural network models and the optimal combined forecasting model are respectively used to predict the water content of the coal measures strata, The results show that the optimal combination prediction model has the highest prediction accuracy and the accuracy and practicability of the coalition formation water content prediction using the optimal combined forecasting model are proved.