-
摘要: 以淮南矿区潘一矿13-1煤层为研究对象,在分析勘探钻孔资料的基础上,确定了煤层埋深及厚度、顶底板岩性、地质构造和煤变质程度是影响煤层瓦斯含量的主要因素;使用BP神经网络方法建立了瓦斯含量预测模型;结合实际数据,对预测模型进行训练和检验。预测结果表明:该模型比使用多元线性回归预测能获得更高的精度,说明预测模型可靠。Abstract: The paper presented a BP neural network model to predict content of coalbed gas based on analyzing explored borehole data of No.13-1 coal bed of Panyi mine in Huainan coal mining area.It is determined that the coal burial depth,thickness,lithologic properties of roof and floor,and coal rank are the important factors controlling gas content of coal.And the model was trained and tested with experimental data.Predicting results of the model are more accurate and reliable than that of linear regression prediction model.
-
Keywords:
- gas content /
- BP neural network /
- prediction model
-
-
期刊类型引用(2)
1. 吴波,刘林,郁东旭,王建州,王义江. 激光照射花岗岩抗拉强度及孔隙分布研究. 建井技术. 2024(04): 88-94+74 . 百度学术
2. 杨豫龙,曹卫华,甘超,黎育朋,吴敏. 深部地质钻进过程地层特征参数建模与安全预警研究进展. 煤田地质与勘探. 2024(10): 195-206 . 本站查看
其他类型引用(1)
计量
- 文章访问数: 50
- HTML全文浏览量: 19
- PDF下载量: 4
- 被引次数: 3