摘要
在海上丛式井井眼防碰监测中,风险邻井套管头检测到的钻头振动信号受到平台噪声干扰,针对钻头振动特征信号难以分离识别的问题,提出了一种基于独立成分分析(ICA)分离同频率范围信号的方法。首先对采集到的信号作限幅滤波处理以滤除部分干扰噪声,然后应用ICA方法分离获得不同源的若干路独立信号(其中包含钻头振动信号),根据独立信号时-频特征分析,结合先验知识实现钻头振动特征信号的识别。通过丛式井防碰现场数据表明该方法可以有效地分离识别钻头振动信号,以提取出信号的能量为标准,能够准确判断钻头趋近风险井套管程度,达到在防碰预警目的。
In process of offshore cluster well borehole anti-collision monitoring,the drill bit vibration signal detection in the casing head of adjacent well is mixed with lots of platform noise. Aiming at the difficulty of the separation and identification of bit vibration signal,a new method based on independent component analysis was proposed to separate the mixed signals with the same frequency range. Firstly,collected signals were made amplitude limiting and smoothing process to filter some noise,and then the ICA method was applied to separate them into several independent signals（ the drill bit vibration signal is included） with different sources,finally the drill bit vibration signal was identified by combining time-frequency analysis and prior knowledge. By field anti-collision data of cluster well,the proposed method was testified effective to separate and identify bit vibration signal,and according to the extracted signal＇s energy,the bit approaches to the risk well casing was successfully alerted.
作者
刘刚
王航
穆文军
LIU Gang;WANG Hang;MU Wen-jun(School of Petroleum Engineering,China University of Petroleum（East China）,Qingdao 266580,China)
出处
《科学技术与工程》
北大核心
2018年第16期33-37,共5页
Science Technology and Engineering
基金
国家"973"计划(2015CB251200)资助
关键词
井眼防碰
振动信号
独立成分分析
时-频分析
信号识别
borehole anti-collision
vibration signal
independent component analysis
time-frequency spectrum signal identification
