The Research On Measurement While Drilling Mud Pulse Signal Denoising

Measurement While Drilling(MWD)is a new well logging technology developed in recent years,it can measure the parameters under the well in real time while drilling the well,and transmits the signal to the ground receiving system.So it plays an important role in improving the efficiency and rate of drilling.In MWD system,the most widely used signal transmission medium is mud,MWD is constantly improving while being used in application,but it is still not so perfect in signal transport process.The transmission of mud pulse signal is affected by kinds of factors: firstly,signal attenuation is inevitable through long distance,and secondly,useful signal is affected by much noise.As a consequence,the signal detected by mouth of the well is so weak and include much noise.How to extract useful signal from the strong background noise whose frequency is not fixed,become a key issue in the technology of MWD.For the denoising effect is not ideal in MWD,on the basis of analyzing the domestic and foreign research status about MWD denoising,the algorithm of wavelet packet transform combined with LMS adaptive filter is used in mud pulse signal denoising.And this algorithm is promoted according to the characteristics of signal modulation,a new variable-step size Least Mean Square(LMS)adaptive filter algorithm is proposed.The relative components after wavelet packet decomposition are chosen to rebuilt as the input of LMS adaptive filter.With the re-treatment of the filter,the useful mud pulse signal is obtained.The algorithm can increase the signal-to-noise ratio(SNR)compared with other methods,especially in low SNR environment,laying a foundation for the following signal identification.