摘要
在钻井过程中,由于岩石的不均质性、井下工况的不可预测性,采集到的信号通常被背景噪声所淹没。文章针对钻井过程中钻头与岩石啮合产生钻柱轴向调幅振动信号中强烈的背景噪声,采用小波包多层分解、域值滤波、通过能量系数值选取最优小波系数向量,并重构信号。对重构信号做FFT分析,有效的提取出了被噪声淹没的调幅信号特征,检测出了中心频率及其边频率带,将其作为钻头正常工作的图谱。采集并重构牙轮旷动的轴向振动信号,FFT及对比分析结果明显区别于正常工作钻头图谱,从而检测出了三牙轮钻头轴承故障。文章提出了新的钻柱轴向振动监测方法,可以及时发现钻头事故的前兆,实现安全、高效钻井。
Owing to the heterogeneity of rocks and the u npredictability of downhole behaviors, the signal acquired was often masked by the striking background noise in the process of drilling. The wavelet pocket mul ti-decomposition and threshold filtering were adopted in light of the striking background noise in the drilling string axial amplitude modulation vibration sig nal caused by the bit-rock interfit during drilling, and the signal was recons tructed by optimizing wavelet coefficient vector. The reconstructed signal was a nalyzed by FFT and the amplitude modulation signal features masked by noise were effectively extracted. The detected central frequency and its sideband were tak en as the bit normal working pattern. Based on this, the tricone bit bearing fa ult may be detected, if the result of FFT analysis of the acquired and reconstr ucted axial vibration signal of cone shake is evidently different from the norma l pattern through correlation. Therefore, the forewarning of bit accident can b e discovered in time so as to realize a safe and high-effective drilling by use of the method introduced in the paper
出处
《天然气工业》
EI
CAS
CSCD
北大核心
2005年第5期56-58,共3页
Natural Gas Industry
基金
国家自然科学基金(No. 50074024)
江钻股份公司资助项目。~~
