Research On Flaw Detection And Identification Of Pipeline Based On Wideband Guided Wave

As an important infrastructure,pipeline is widely used in the construction of national economy and national defense.Due to the influence of the environment and transportation medium,cracks,corrosion and fatigue and other forms of damage are easy to appear in the pipeline in use,which poses potential hazards to the safe transportation.The new ultrasonic guided wave nondestructive testing technology can realize long distance online scanning and has a good application prospect in pipeline testing.The key of pipeline guided wave detection is to identify and extract the characteristic information of pipeline defects from ultrasonic guided wave response signals.In practical pipeline guided wave applications,single-mode excitation is used to simplify the received signals,so it is often difficult to realize the rapid and comprehensive detection and evaluation of complex pipeline defects.The choice of excitation signal of guided wave detection system has great influence on the detection result.The wideband guided wave response can provide more abundant information of pipeline defects.In this paper,an overall scheme based on wideband guided wave detection and identification of pipeline defects is proposed by referring to the idea of sweep-frequency capacitance sensing and identification technology.On this basis,research is carried out on the problem of wideband guided wave detection and identification of pipeline defects.The main contents include:Research and analysis of wideband excitation signal,modeling and simulation of guided wave detection of complex defective pipeline,construction and extraction of pipeline defect characteristic index and identification of pipeline defect severity.The traditional guided wave detection of pipeline mostly uses Toneburst excitation.This narrow-band excitation method mainly uses the single mode or a small number of guided wave modes of ultrasonic guided wave,which cannot guarantee the accurate and reliable detection of various defects of pipeline.In this paper,different guided wave modes have different detection sensitivities to different defect features.Further,different forms and optimization parameters of wideband guided wave excitation signals are discussed through comparative study.In order to effectively carry out the research on the detection and evaluation of wideband guided wave pipeline defects,this paper proposes a data-driven modeling method based on the finite element theory,aiming at the pipeline problems involved in this paper,and realizes the universal modeling of different pipeline defects and the automation of pipeline ultrasonic guided wave modeling and simulation process.In order to extract the defect related information from the defect reflection echo,a variety of waveform time domain statistical indexes were introduced to characterize the defect characteristics after envelope and noise reduction processing of the echo signal.The interaction process between wideband guided waves and pipeline defects was further studied,and a support vector machine classifier was constructed to train and test the defect characteristics,so as to realize the classification and evaluation of the damage level of pipeline defects.The work in this paper can provide a theoretical basis for further comprehensive detection and evaluation of pipeline defects based on guided wave technology.