泥浆MWD遥传信号泵冲干扰的消除方法研究

发布时间：2018-06-21 21:30

本文选题：随钻测量 + 泵冲干扰　；参考：《电子科技大学》2017年硕士论文

【摘要】：钻井作业的同时,需要实时地从井下向地面传输数据,目前国际上普遍采用的是随钻测量(Measurement While Drilling,MWD)技术。其中,以钻柱管道中的循环泥浆液作为传输载体,通过泥浆脉冲传输信号的方式,以其易于安装维护、低成本的优点,成为了目前应用最为广泛的MWD传输方式。但由于泥浆信道衰减效应严重,脉冲信号的有明显的畸变失真,另外其工作现场环境复杂,钻井平台的各种机械噪声,尤其是泵冲干扰噪声,其频谱能量分布在脉冲信号带内,对脉冲信号的识别和处理造成了极为不利的影响。因此在处理泥浆MWD遥传信号的问题中,对泵冲干扰的消除效果会直接影响系统的传输性能。本论文针对泥浆MWD遥传信号的泵冲干扰消除问题,详细分析了泥浆信道、信号和噪声的特性,建立了信号仿真模型。并以此为依据,分别从单通道盲源分离和基于稀疏成分分析这两个角度研究了泵冲干扰的消除方法,并通过仿真实验验证了其有效性。本文的主要工作内容如下:1.简述了泥浆MWD遥传信号的产生和传输过程,分析研究了信道的传输特性、泥浆脉冲信号和背景噪声的特性,并结合实验数据建立特征脉冲信号波形和泵冲干扰模型,用于生成不同参数的仿真信号。2.研究了基于单通道盲源分离的方法,从常规的适定多通道观测信号的盲源分离理论入手,详细描述了基于最大峭度的Fast ICA的盲源分离算法。并针对单通道的盲源分离问题,分别设计了基于EEMD分解和基于稀疏表示的构建虚拟多通道的方法盲源分离方法,对它们的性能进行了仿真实验,对于特征脉冲和泥浆泵冲噪声的混合信号,得到了基于DCT稀疏表示分解的Fast ICA盲源分离性能优于基于EEMD分解的Fast ICA盲源分离的结论。3.研究了基于稀疏成分分析的信号分析方法,详细阐述了形态成分分析的推广得到的稀疏成分分析方法的理论模型和求解步骤,讨论和设计了小波字典基的优选方法,并针对泥浆MWD遥传信号的特征进行了算法优化。仿真结果验证了基于稀疏成分分析方法对泥浆MWD遥传信号处理的有效性。
[Abstract]:At the same time, it is necessary to transmit data from downhole to ground in real time. At present, measurement while drilling is widely used in the world. The circulating mud in the drilling string pipeline is used as the transmission carrier and the signal is transmitted by the mud pulse. Because of its advantages of easy installation and maintenance and low cost, it has become the most widely used MWD transmission mode. However, due to the serious attenuation effect of mud channel, obvious distortion of pulse signal, complicated working environment, various mechanical noise of drilling platform, especially pump noise, the spectrum energy is distributed in the pulse signal zone. The recognition and processing of the pulse signal have a very adverse effect. Therefore, in the process of mud MWD remote transmission signal, the effect of eliminating pump impulse interference will directly affect the transmission performance of the system. In this paper, the characteristics of mud channel, signal and noise are analyzed in detail, and the signal simulation model is established. On the basis of this, the methods of eliminating pump impulse interference are studied from single channel blind source separation and sparse component analysis, respectively, and their effectiveness is verified by simulation experiments. The main work of this paper is as follows: 1. The generation and transmission process of mud MWD remote transmission signal are briefly described. The characteristics of channel transmission, mud pulse signal and background noise are analyzed and studied. The characteristic pulse signal waveform and pump impulse interference model are established based on the experimental data. Used to generate different parameters of the simulation signal. 2. The method of blind source separation based on single channel is studied. The blind source separation algorithm based on maximum kurtosis based on Fast ICA is described in detail from the conventional blind source separation theory of multi-channel observation signals. Aiming at the problem of single channel blind source separation, the methods of constructing virtual multi-channel based on EEMD decomposition and sparse representation are designed, and their performance is simulated. For the mixed signals of characteristic pulse and mud pump impulse noise, the conclusion that fast ICA blind source separation based on DCT sparse representation decomposition is better than fast ICA blind source separation based on EEMD decomposition is obtained. The signal analysis method based on sparse component analysis (SAA) is studied. The theoretical model and solution steps of the sparse component analysis (SAA) method, which is generalized by morphological component analysis, is described in detail. The optimal selection method based on wavelet dictionary is discussed and designed. The algorithm of mud MWD remote transmission signal is optimized. The simulation results verify the effectiveness of sparse component analysis (SAA) method for mud MWD remote signal processing.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TE92

【相似文献】