Research On Signal Noise Mechanism And Noise Reduction Method Of Pipe Pulse Magnetic Eddy Current Detection Based On EEMD

China's petroleum and petrochemical industry has developed rapidly,and the demand for oil products in all walks of life is huge.As one of the most reliable,fast and convenient ways of long-distance oil transportation,pipeline has been widely used in China.At present,oil and gas pipeline network is all over the country,and the use of pipeline is increasing year by year.However,with the increasing service life of pipelines,the life and reliability of pipelines affected by self-corrosion and stress are declining,and with the change of surrounding complex environment and man-made damage,pipelines are prone to cracks,pits,wall thinning and other risk factors in the process of transportation,resulting in environmental pollution,economic losses and life safety.Ferromagnetic coated pipes have been widely used.The insulation layer and aluminium coating on the outer surface of pipes provide better protection for pipes,but it also increases the difficulty of defect detection for pipes.Compared with other conventional detection methods,pulsed magnetic eddy current detection method has many advantages,such as rich information,non-disassembly,good security and low cost.In view of the complex working conditions,such as medium temperature,pressure,electromagnetic interference,environmental impact,etc.,the detected data can not be directly used to analyze pipeline defects.It is necessary to do relevant noise reduction processing to the pulse eddy current signal collected in the field,so as to make the signal more stable and easy to identify the defects.There are many kinds of noise reduction methods.In view of the non-stationary and non-linear characteristics of the obtained pulsed magnetic eddy current signal,considering the advantages and disadvantages of various algorithms,a signal denoising method based on improved EEMD and singular value decomposition is proposed.Firstly,the collected signals are decomposed by EEMD,and the IMF components of each mode are obtained.According to the correlation coefficient method,the noise-dominated IMF components and the signal-dominated IMF components are selected.Singular value decomposition(SVD)is used to denoise the high frequency IMF components.Finally,the denoised IMF and the signal-dominated IMF components are reconstructed to obtain the denoised signal.Firstly,the test signal processing is analyzed and compared with the signal-to-noise ratio(SNR)and root mean square error(RMSE),and the obvious effect of noise reduction is obtained.Secondly,the pulse eddy current signals with high SNR and low RMSE are obtained by analyzing and processing the experimental signals,which proves the feasibility and superiority of this method.