摘要
从分析钻进信息与岩石物理力学性质的关系入手 ,利用人工神经网络理论对所钻进的地层进行实时识别。根据钻探生产的特点 ,合理设计了人工神经网络的结构 ,开发了人工神经网络识别地层的软件。选择了结构简单、安装方便和具有较强二次开发功能的无传输信号线钻参仪作为数据采集的硬件系统 ,并开发了相应的数据传输接口软件。并通过室内实钻试验对所开发的人工神经网络识别软件和数据传输接口软件进行了检验。试验结果表明钻参仪选型正确 ,数据接口软件运行稳定可靠 。
Starting from the analyses of the relationship between drilling information and rock characters of physics and mechanics, the drilled strata can be identified using the artificial neural network theory. The structure of the network is designed in reason based on the characteristics of drilling production. The software is developed to identify the drilled strata in method of artificial neural network. The drilling parameter monitor, without transport signal lines, but with simple structure, convenient installation and better function to develop second, is chosen as the sample hardware system. The data traffic interface software is developed. The artificial neural network identification software and data traffic interface software are tested through practical drilling experiment. The result shows that the drilling parameter monitor is chosen correctly, the data traffic interface software runs credibly and the effect of the identification software is ideal.
出处
《探矿工程(岩土钻掘工程)》
2003年第S1期194-196,共3页
Exploration Engineering:Rock & Soil Drilling and Tunneling
基金
国土资源部地质大调查科研项目 ( 2 0 0 0 2 0 170 13 1)
关键词
钻进信息
人工神经网络
识别
钻参仪
drilling information
artificial neural network
identify
drilling parameter monitor