The Process Of MWD Mud Pulse Signals

With the reform and opening up, China’s demand for energy has put forward higherrequirements. Thus the increasing China’s oil and gas production in order to get rid of thedependence of foreign energy can not only promote the development of the national economydeveloping well and quickly, but also ensure the energy security of China. MWD(measurement while drilling) technology can effectively monitor and control the downholesituation, which has great signiifcance to improve the oil production. Signals collected byMWD surfacc system arc disturbed by many kinds of noises emerged during the transmission,so it can not be directly used to decode to monitor the underground situtations real-time. Thusthe process (denoising and decoding operations) of MWD mud pulse signals is helpful toobtain real-time downhole information to monitor and control the trajectory of the borehole, itis a key technology in MWD.This paper researched on MWD mud pulse signals, introduced the generation andtransmission of MWD mud pulse signals briefly, aimed at the noise problems of mud pulsesignals, analysised the noise sources and noise characteristics of the mud pulse signals,proposed signal de-noise scheme of mud pulse signals based on wavelet transform andempirical mode decomposition (EMD)‘Duirng wavelet de-noising, statred from the basictheory of wavelet transform, researched on wavelet de-noising methods and theories, made adeeply analysis and discussion on the selection of wavelet bases, wavelet decomposition leveland threshold function. For the insuiffcient of hard and sotf threshold function, proposed asignal processing method of MWD mud pulse signals based on the maximum likelihoodestimation threshold, and used MATTAB program to veirfy the corresponding algoirthm.EMD method has a good adaptive in dealing with non-linear and non-stationary signals, itcould decompose the original signals into a seires of Intrinsic Mode Function (IMF). Thispaper implemented corresponding decomposition program in MATLAB to use thedecomposed IMFs to realize the de-noising of MWD mud pulse signals.At last, analysised and processed the ‘M-crest’，true and false pulse problem of mud pulsesignal after denoise, decoded the MWD mud pulse signals according to the coding protocol to obtain real-time down-hole conditions and proposed the future direction of research.