Research On Noise Removal And Recognition Algorithm Of LWD Signal Based On Second Generation Wavelet Transform

In petroleum exploration engineering,the logging while drilling system transmits drilling engineering parameters such as downhole well angle,tool face angle,and torques,physical parameters such as natural gamma and compensated neutron density to the surface in real time,and It has a crucial guiding role in the process of drilling steering and geological characteristics detection.The method of wireless mud pulse transmission is widely used for signal transmissionIn practice,due to the large drilling depth,harsh mining environment,and complicated geological composition,the effective signals of mud positive pulses are completely submerged in noise.The acquisition of downhole parameter information is gradually increasing,and the rate of data transmission is getting higher and higher.Therefore,high-rate filtering algorithm is the key to LWD signal processing technology.Because the amplitude of effective signal is weak,the signal transmission rate is high and the noise strength is large,this thesis proposes a noise reduction algorithm based on the second generation wavelet transform and a variable amplitude-maximized correlation coefficient pulse recognition algorithm for this type of signal.The algorithms are applied to simulated positive pulse mud signal of deep well and the positive pulse measured in an oil field.The algorithms are experimentally verified with wavelet transform,complete ensemble empirical mode decomposition,and least mean square adaptive noise removal algorithm,The experimental results show that the algorithm proposed in this thesis has a higher computing rate for deep well mud pulse signals than the above three algorithms in terms of noise removal,and achieves optimal results in pulse recognition for simulated deep well mud pulse signals.The research topic of this thesis comes from the national key research and development plan:the project of "Development of high-speed mud pulse system".The main research excerpts of wireless mud pulse transmission in logging while drilling in this thesis are as followsFirstly,this thesis studies the transmission characteristics of mud pulse signals in logging while drilling technology,summarizes and analyzes the signal encoding format and the source of noise during transmission,and designs a pulse interval encoding method that meets mud pulse data transmission,and lays the foundation for establishing a mathematical model for simulating mud pulse signals of deep wellSecondly,this thesis studies the hardware and software of signal acquisition and processing in LWD.According to the channel transmission characteristics of the mud pulse signal,the signal acquisition module,numerical simulation software,signal processing flow and five metrics for evaluating algorithms are analyzed in detailThirdly,this thesis uses the NI cDAQ-9132 signal acquisition module and Lab View to set up an experimental platform for simulating mud pulse signal acquisition and describes the front panel,hardware block diagram,and working principle of the signal acquisition module in detail.This thesis also simulates the noise interference of mud positive pulse pressure signals after transmission through complex channels,and completes the collection,storage and output of mud pulse signal.Lastly,this thesis studies the processing algorithms for the simulated positive mud pulse signal of deep well and the positive pulse signal measured in an oil field.According to the characteristics of weak mud pulse signal amplitude and strong noise signal strength,the second generation wavelet transform has the characteristics of multi-resolution analysis,does not rely on Fourier transform,and completes the wavelet transform directly in the time domain.The transform algorithm is simple and the calculation speed is fast,it is suitable for parallel processing,and it requires less hardware memory,which is easy to implement in the DSP chip.The experimental results show that noise removal algorithm proposed in this study based on the second generation wavelet transform has the advantages of fast operation speed,does not need additional expected signal.The algorithm proposed in this thesis has less calculation amount,and outstanding filtering effect on mud pulsed noise signals of deep well.It has the advantages of simple,fast and direct in signal reconstruction,and does not require additional auxiliary storage.This algorithm is suitable for analyzing signals collected in the condition of large oil well depth,harsh mining environment and complicated geological composition in the logging while drilling system.It can improve the signal processing efficiency and has practical application value.