Pipeline Defect Detection Based On Ultrasonic Guided Wave And Machine Vision

As an efficient and economical mode of transportation,pipeline transportation plays an increasingly important role in economic development.With the continuous extension of pipeline running time,due to corrosion,aging and deformation and other factors caused by frequent pipeline leakage accidents,which on the one hand will cause waste of raw materials,on the other hand may also lead to environmental pollution and even casualties,so it is particularly important to regularly conduct non-destructive testing of pipelines and evaluate the operating status of pipelines,which is of great significance to ensure the normal operation of pipelines and reduce losses.Based on the theory of ultrasonic guided wave and machine vision inspection,this thesis adopts a combination of theory and experiment to study the pipeline defect detection method based on ultrasonic guided wave and machine vision.Ultrasonic guided wave has the advantages of quickly realizing defect location and giving total loss rate,but it cannot intuitively reflect the details of pipeline defects,while machine vision can intuitively reflect pipeline defect information,but the system is more complex and the detection efficiency is relatively low,and it is not possible to quickly achieve pipeline defect location.The combination of ultrasonic guided wave and machine vision inspection method,the use of complementary fusion methods can quickly locate pipeline defects and identify the size of defects and other information,which can provide a reliable basis for pipeline maintenance.Based on the theory of ultrasonic guided wave detection,this paper selects L(0,1)mode ultrasonic guided wave as the excitation mode for the detection of pipeline defects.Firstly,the ABAQUS software is used to establish the pipeline model,perform numerical simulation calculations,determine the defect location of the pipeline,use the collected pipeline defect echo signal to locate the pipeline defect axially,use the energy amplitude method and the circular trajectory curve method to achieve the circumferential positioning of the pipeline defect,and analyze and compare the two methods.The results show that the amplitude energy method can only roughly determine the circumferential position of the defect,and the positioning accuracy is not high,while the circular trajectory curve method can be used to better position the pipe defect circumferentially,and the smaller the angle,the higher the defect positioning accuracy,and the overall defect positioning accuracy is within 8.5%.Then,the single-defect and multi-defect imaging of the pipeline is realized by the full focus imaging algorithm,and the signal processing methods such as median filtering,Hilbert transform extraction signal envelope,and signal sharpening are used to improve the imaging accuracy of pipeline defects.The Hilbert transform and EMD method are used to achieve feature extraction of pipeline defect echo signal,and finally a three-input and two-output model is established by using BP neural network to identify the circumferential defect size of the pipeline.Based on the machine vision inspection theory,the image acquisition of the inner surface of the pipeline,the analysis of the inner surface defect of the pipeline,and the three-dimensional visualization of the inner surface defect of the pipeline are realized.The ultrasonic guided wave and machine vision are complementarily integrated,giving full play to their respective advantages,realizing the all-round and multi-angle analysis and judgment of pipeline defect detection data,laying the foundation for the subsequent development of pipeline defect detection system.