Research On Nondestructive Technology Of Oil And Gas Pipeline Based On EMAT

Pipeline is the main tool of transport oil and gas.At present,most domestic longdistance pipelines transport media that is high temperature,high pressure,flammable,explosive,highly toxic and corrosive.When pipelines have defects,if not handled in time,leakage or explosion accidents will occur.So pipeline integrity testing is necessary.Because of the uneven surface of pipeline,crack defects are easy to appear in circumferential weld and other problems,which makes the detection technology of oil and gas pipeline difficult.Nondestructive testing is very important.Electromagnetic acoustic transducers(EMAT)is EMAT is a new non-destructive testing method.Because of its high temperature resistance,no coupling agent,wide range,no need to clean the surface of pipeline,especially allowing a certain lift-off distance between the probe and the measured object,the electromagnetic ultrasonic testing technology of oil and gas pipeline has become a new pipeline detection technology.Firstly,the principle of excitation and acceptance of electromagnetic ultrasound in different materials is analyzed,the physical model and mathematical equation are established,the propagation of ultrasonic echo is simulated,the formation process of electromagnetic ultrasound is visually displayed,and it is proved that electromagnetic ultrasound technology can be applied to oil and gas pipelines.Secondly,a new model of electromagnetic echo is proposed for the high-lift off electromagnetic echo signal.The new model points out that in the high-lift off electromagnetic echo signal,the echo signal is composed of useful ultrasonic signal,high-frequency narrowband background noise signal and white noise.Thirdly,an electromagnetic ultrasonic denoising algorithm for high lift-off distance is proposed.This algorithm successfully removes the noise in the electromagnetic ultrasonic signal of oil and gas pipeline,and achieves the extraction of weak signal.Moreover,this algorithm can retain the feature of defect information and extract defect echo,and different types of defects are verified.Finally,the echo signals of defects in pipeline girth weld and Non-defects in pipeline girth weld are collected,and the time-domain and time-frequency-domain features of signals are extracted.The dimension reduction processing is carried out by principal component analysis method.The obtained eigenvectors are trained in support vector machine.After that,the trained support vector machine can automatically distinguish whether the ultrasonic signals are defect signals or not.