摘要：钻柱振动非常复杂，简单的计算已经远不能够满足目前现场应用的需求。钻柱振动信号是一种十分复杂的振动波，能够实时地反映出井下的工况信息。开展模式滤波法分离钻柱顶部振动信号的研究，实现各种振动成分的合理分离，对分离的信号进行归类与识别处理，并实现各种典型信号的提取，通过将分离后的信号转换成数字化音频进行耳听测试，可迅速判断分离效果和分离信号的具体成分，从而极大地提高了处理的效率与针对性。能够实时识别井下工况，进行钻井参数优化，防止钻井事故的发生，对提高钻井速度，降低钻井成本有着十分重要的意义。本论文首先讨论了钻柱振动信号的采集装置，实现了动信振号的地面实时测量。并讨论了信号处理的方法，其中包括：最优频率匹配法、小波变换、频谱分析等分析方法。最后用模式滤波法对钻柱振动信号进行分析处理，并应用到井下工况的识别应用中去。

关键词：钻柱振动，模式滤波法，井下工况，信号分析，信号识别

Abstract: Drill string vibration is very complex, the simple calculation is far from being able to meet the needs of the field application. Drilling string vibration signal is a very complex wave, the vibration waves can reveal the real—time information for downhole drilling． To carry out the mode filtering method to separate the top of the drill string vibration signal, and achieve a reasonable separation of the various vibration components. Classification and recognition processing on the signal separation，and a variety of typical signal extraction，converted into a digital audio signal will be separated ears test，it can quickly determine the specific composition of separation and separation of signals, thereby greatly enhancing the efficiency of the treatment and targeted. It can identify the downhole working status and optimize the well drilling parameters online．To prevent the occurrence of drilling accident，increase drilling speed，reduce the cost of drilling has great significance.This paper first discusses the drill string vibration signal acquisition device，real-time measurement of the action letter vibration ground.And discussed the method of signal processing，it Including：the optimal frequency matching, wavelet transform, spectral analysis and other analytical methods. Mode filtering method of drill string vibration signal analysis and processing, and application recognition applications to downhole working conditions.

KEY WORDS：Drill-string vibration；Pattern filter method； Downhole working condition；Signal analysis；Signal identification；