摘要
风动潜孔锤钻进在钻效方面有着传统钻进方法无可比拟的优势。以往的钻速预测主要是针对回转钻进,而准确预测风动潜孔锤钻速则更具有现实意义,更有利于合理制定钻井计划。利用有多隐层的BP神经网络原理,综合考虑钻压、转速、风压、风量、钻井深度、钻头工作时间等各方面参数对钻速的影响,研究了贯通式潜孔锤冲击回转钻井钻速的预测方法,并对现场实际钻井作业进行了预测。预测结果与现场实际情况能很好地吻合。同时又以预测结果为基础,优选钻进参数以指导施工。
Percussive drilling has an incomparably priority to traditional rotary drilling methods in the aspect of penetration efficiency and the previous prediction of the drilling rate is only refered to rotary drilling. Practically, accurate prediction of the rate of penetration （ROP） of percussive drilling is very important in that it can help make the planning of the rock excavation projects in high efficiency. The multilayer neural network with back propagation algorithm （BPNN） has been employed to analyze the influence of the different parameters （i. e. thrust, rotational speed, pressure of compressed air, volume of compressed air, drilling depth and drilling bit operation time etc. ） to ROP and a method of predicting ROP for the hollow- through type DTH drilling is also studied and applicated in the field drilling work with a very good agreement to actual values in the fields. At the same time, drilling parameters have been optimized to guide the penetration operation based on the prediction results.
出处
《吉林大学学报（地球科学版）》
EI
CAS
CSCD
北大核心
2009年第5期882-886,共5页
Journal of Jilin University：Earth Science Edition
基金
中国地质调查局区域地质调查项目(1212010660804)
吉林大学'985'工程研究生创新基金项目(20080245)
关键词
神经网络
钻速
冲击回转钻进
风动潜孔锤
neural network
the rate of penetration（ROP）
percussive rotary drilling
DTH hammer
