随钻测量系统脉冲信号特征提取的多尺度方法  

FEATURE EXTRACTION FOR MUD PULSE OF MWD SYSTEM BASED ON MULTI-SCALE METHOD

作　　者：李建勋[1] 王睿[1] 柯熙政[1] 李昕[2] 耶亚林[3] 

机构地区：[1]西安理工大学自动化与信息工程学院 [2]长庆石油勘探局工程技术研究院钻井所 [3]西安通信学院

出　　处：《钻采工艺》2008年第1期25-27,30,共4页Drilling & Production Technology

基　　金：陕西省教育厅科技专项基金(04JK247)

摘　　要：在钻井液随钻测量系统中,从接收信号中识别有效脉冲是地面系统正确解码的关键。文章引入多尺度小波分解方法分析钻井液信道噪声特性和信号特性的基础上,分别使用小波熵法和小波阈值滤波方法对含有多频泵噪声的脉冲信号进行基线波动修正及有用脉冲提取。结果表明,小波熵能够刻画钻井液脉冲信号能量分布的真实特征,而小波阈值滤波也是有效可行的MWD信号特征提取方法。通过工程实践的检验,文中的两种方法达到了检测率高、虚警率低的效果。Detecting true pulses from received mud signals is the key for decoding pulse train sequences in Measurement While Drilling system.The characteristics of downhole signals and noises were analyzed through multi-scale analysis.Two new methods,that is,wavelet entropy and threshold de-noising were used respectively to eliminate the large amplitude,multi-frequency,frequency-shifting pump noise,especially the low-frequency baseline fluctuation,and to extract the legal pulse.Results showed that wavelet entropy can reflect the real feature of pulse signal based on energy distribution,and,threshold de-noising was effective for extracting the characteristics of MWD signal.The two methods were tested by engineering practice,showing its high detection rate.

关 键 词：随钻测量 小波阈值滤波 小波熵 特征提取 

分 类 号：TE927[石油与天然气工程—石油机械设备]