Research On Defect Recognition Of Oil And Gas Pipeline Based On Laser Ultrasonic Technology

The safety of oil and gas pipelines is closely related to the transportation of fossil energy such as oil and natural gas.With the steady development of fossil energy at home and abroad,the transportation safety of oil and gas pipelines in service becomes more and more important.As a new nondestructive testing technology in recent years,laser ultrasonic testing technology can realize fast and efficient damage detection for oil and gas pipelines in service.Based on the research and analysis of laser ultrasonic testing technology theory,laser ultrasonic testing system and laser ultrasonic defect recognition technology at home and abroad,this paper uses laser ultrasonic testing system to collect oil and gas pipeline defect time domain signal and oil and gas pipeline damage image to realize defect recognition of oil and gas pipeline in service.The specific research contents and innovations of this paper are as follows:(1)By discussing the selection of wavelet parameters for time domain signal of oil and gas pipeline defects,the denoising effects of wavelet denoising and multi-resolution analysis denoising in oil and gas pipeline ultrasonic simulation signal are analyzed.The results show that coif4 wavelet,three-layer decomposition and soft threshold denoising under multi-resolution analysis are the optimal denoising methods for stimulated light ultrasonic simulation signal,which can maintain the smoothness and stability of ultrasonic signal On the basis of continuity,the average signal-to-noise ratio is increased by 13.75 db,compared with the wavelet denoising,the average signal-to-noise ratio is increased by 1.393 db.On the basis of the optimal denoising of laser ultrasonic analog signal,the optimal denoising method is adopted to denoise the laser ultrasonic measured signal.(2)Taking the time domain signal of oil and gas pipeline defect after de-noising as the research object,the feature energy and entropy of oil and gas pipeline ultrasonic defect signal are extracted by using wavelet packet energy decomposition method.The extracted feature energy and entropy are used as the input of BP neural network and the output of oil and gas pipeline damage size to realize the preliminary quantitative analysis of laser ultrasonic oil and gas pipeline defect The BP neural network model for quantitative analysis of trace defects is stable and has good performance.The average prediction accuracy is 0.997645,and the prediction error is close to 2%.(3)The oil and gas pipeline damage image collected by laser ultrasonic testing system is pretreated by gray-scale method.The gray-scale laser ultrasonic pipeline damage image is segmented and matched by mask sad image segmentation and matching algorithm.The mask defect image is processed by Retinex algorithm and improved Retinex algorithm of image enhancement and clutter filtering.The clutter and interference of oil and gas pipeline damage grayscale image are analyzed For the problem of wave filtering,the classical filtering algorithm is used to filter;the processed defect mask image and the oil and gas pipeline damage image filtered by clutter and forward wave are processed by sad matching algorithm and Canny algorithm to realize the qualitative identification of oil and gas pipeline flaw detection.(4)In order to realize the rapid and efficient identification of oil and gas pipeline damage by laser ultrasonic,MATLAB software is used to make the software,through which different types and sizes of oil and gas pipeline damage can be displayed intuitively.The realization of the software design provides the feasibility for the rapid detection and rapid loss determination problems faced by laser ultrasonic field detection.Finally,through the detection and research of laser ultrasonic oil and gas pipeline defect identification,it provides theoretical and research basis for laser ultrasonic detection of oil and gas pipeline in service,and provides reference value for damage detection of oil and gas pipeline in service in the future.