Research On Defect Identification Technology Of Process Pipeline Detection Based On Ultrasonic Guided Wave

As an important part of the gas transmission network.To ensure the safe transportation of natural gas and avoid accidents such as personal casualties,environmental pollution or property losses,it is necessary to carry out risk control and accident prevention for potential safety hazards in gas transmission stations.Ultrasonic guided waves inspection technique shows various advantages such rapid,long-distance detection and relatively low cost.When using ultrasonic guided waves for process pipeline inspection,the amount of excavation can be greatly reduced,the construction period can be shortened,and the cost can be saved.Based on the ultrasonic guided wave detection technology,we proposed a fast and efficient method to identify the defects and other features of process pipelines in gas transmission stations in this study,and the main research contents are as follows:(1)The transmission process was stimulated to analyze the transmission characteristics of ultrasonic guided waves in pipelines on the basis of the finite element method.By the simulation: in terms of energy transformation,there were energy attenuations and transfers in the modes of L(0,2)and T(0,1)in the elbows,among which energy concentrated on the upper and lower sides after guided waves entered the elbow in L(0,2)mode and concentrated on the left and right sides in T(0,1)mode.In terms of mode transformation,the guided wave of the T(0,1)mode always maintains an independent and clear state compared with guided wave of the L(0,2)mode,contributing to defect detection of pipelines with elbows.(2)The influences of frequency and elbow characteristics on the wave signal in defect detection were stimulated under test conditions of the T(0,1)mode.The Wave Maker G4 equipment was applied for experimental verification to determine the optimal detection frequency range.Meanwhile,it was found that elbows could both affect the size of the reflected echo signal of defects and produce large deviations in the axial positioning of defects.(3)A BP neural network which can automatically recognize pipeline features was trained。The pipeline defects were detected by ultrasonic guided wave devices,and data features were classified using the characteristic signal method.A total of 763 groups of data containing different pipeline characteristics were extracted from the result of classification in order to establish and train the BP neural network.The accuracy of the neural network in identifying characteristics and defects of test samples reached 80.9% after training.The defect recognition technology based on the combination of the characteristic signal method and the BP neural network can reduce the subjective influence of the tester and improve the recognition efficiency and accuracy.(4)The system of defect identification and data management for the ultrasonic guided wave was designed and developed.The well-trained network was applied in the system to identify the defects of the station process pipeline and judge it as a defect or other pipeline features.The system also has the function of data management,which can store and modify the detected data.the defects in the process pipelines of the stations automatically and the test data can be stored and managed.