庆深油田神经网络法优选钻头研究  

Research on optimizing bit by neural network in Qingshen Oilfield

作　　者：毕雪亮[1] 闫铁[1] 陶丽杰[1] 

机构地区：[1]大庆石油学院提高油气采收率教育部重点实验室,黑龙江大庆163318

出　　处：《哈尔滨工程大学学报》2006年第B07期111-114,共4页Journal of Harbin Engineering University

基　　金：黑龙江省自然科学基金（E200502）.

摘　　要：利用人工神经网络法中的自适应共振理论进行石油工业中优选钻头类型的研究、为了改善钻头工作性能，在提高钻速的同时，还要注意钻头在井下的可靠性，认真进行钻头优选研究非常必要，这将会大幅度降低石油钻井成本．人工神经网络理论为钻头优选提供了一种有效的方法、文章应用自适应共振人工神经理论来提高石油钻井中钻头类型选择的可靠性，分析了自适应共振神经网络的结构并给出了计算步骤，建立了钻头优选模型．该理论在庆深油田徐深10井进行了实验，结果表明：实验井的平均机械钻速比以前提高20％，单只钻头进尺提高37．4％．提高深井钻井速度的效果明显，该优选钻头方法切实可行．Adaptive resonance theory is applied to optimize bit type in petroleum industry. It is greatly demanded to increase drilling speed in petroleum exploration and development. To increase drilling speed with good reliability, bit type needs to be chosen carefully for drilling performance. Good bit optimization is important for decreasing petroleum industry cost. In bit optimization, there are many related factors, which need to be analyzed comprehensively. Neural networks are helpful tools to improve bit optimization. Bit optimization method is developed by adaptive resonance theory to improve the reliability. The network structure is analyzed and the steps of the model are given. The model of optimizing bit type is used in deep well Xushen 10 of Qingshen Oilfield. It shows that the average drilling rate is 20% faster than former wells at the same depth, and the drilling footage of the single bit is 37.4% higher than former wells. The drilling rate is obviously increased, and this method is confirmed to be practicable.

关 键 词：钻井 ART神经网络 钻头优选 庆深油田 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]