Research On Wavelet De-noising Method For Downhole Received Signal

In a rotary steerable drilling system,the ground command monitoring device sends a control command to the downhole guiding tool via a downlink channel using a negative pulse of drilling fluid.The downlink signal forms a control command in the form of a pulse width code of "three drops and three liters",and a pulse width signal is formed by the change of the drilling fluid displacement.The downlink communication receiving device located in the downhole guiding tool obtain s an accurate downlink command word by detecting and collecting the change of the drilling fluid output voltage of the downhole mud motor and calculating the displacement of the drilling fluid through the pulse width time and sequence of the five command codes.In order to eliminate the noise generated during the underground transmission process,the improved layered threshold denoising method and the singular value decomposition method are used to denoise the received signal to achieve accurate decoding and recognition.Based on the analysis of the received signal and noise characteristics in the downhole,this paper proposes an improved layered threshold denoising method.By improving the threshold function,the method ensures a smooth transition of the wavelet coefficients after the threshold action and has strong flexibility.Firstly,different wavelets were selected for multi-scale decomposition and reconstruction experiments,and the noise reduction results of Sym6 wavelet base were obtained.Different noise reduction algorithms such as global threshold,layered threshold and improved layered threshold are used to reduce the noise of the downlink signal.At the same time,the influence of the wavelet decomposition layer on the noise reduction effect is studied.The experimental data is the best decomposition.With a scale of 4,the wavelet denoising method with improved layered threshold also achieved the expected effect.For the singular points contained in the signal,the WT-SVD decomposition method is used for double-layer filtering,that is,the signal structure matrix after the layered threshold denoising is improved by wavelet,and then the singular value decomposition and reconstruction are performed on the matrix.Compare and analyse the signal-to-noise ratio(SNR),root mean square error(RMSE)and correlation coefficient(p)after noise reduction,observe the signal waveform after noise reduction,the experimental results show that the WT-SVD method has obvious noise reduction effect.