Study On Damage Identification Of Long And Large Pipeline Underground Based On Guided Waves

With the development of economy and society,the level of domestic urbanization is increasing day by day.Both industrial and civil gas and liquid transport require a more efficient and safe pipeline transportation network.However,the working environment of pipeline is corrosive and humid,longterm work makes pipelines vulnerable to different degrees of damage and leakage,which leads to waste of resources and casualties.The research on damage identification of pipeline is an inevitable way to establish early warning mechanism scientifically,and it is of great significance to ensure the normal operation of pipeline system.However,the traditional nondestructive testing methods need point-by-point detection along the pipeline,which is inefficient and prone to missed detection.The traditional testing methods need to expose the part of the pipeline to be tested.When the pipeline is buried underground,it cannot be effectively detected.As a new non-destructive testing technology,guided wave detection technology has good application prospects because of its advantages of longdistance,full cross-section and high-speed detection.At present,many scholars at home and abroad have done a lot of research on pipeline damage identification based on guided wave,and have made some achievements.However,there are still some problems to be solved in the aspects of rapid drawing of guided wave dispersion curve,quantitative analysis of pipeline damage degree and damage location of heterogeneous pipelines.Therefore,the research of pipeline damage identification technology based on guided wave is of great significance.In this dissertation,according to the characteristics of guided wave dispersion,a program for the analysis of guided wave dispersion and guided wave structure is developed by means of the combination of theoretical derivation and finite element analysis.Based on the artificial neural network(Ann)technology,the quantitative analysis model of the damage degree of pipeline is established,and the damage identification of heterogeneous pipeline is studied in this paper.The damage location of the pipeline is identified by the principle of echo location,and the quantitative analysis model of the damage degree is established based on the artificial neural network.This paper deals with elastic dynamics,guided wave theory,discrete mathematics and programming,neural network and other disciplines.Through theoretical and finite element analysis,some conclusions for the practical application of guided wave technology are obtained.The research work in this paper is summarized as follows:1)Based on the Navier’s equation in elastic dynamics,the dispersion equations for describing the frequency-wave velocity relationship and the guided wave structure in any mode of guided wave are established by theoretical derivation.The numerical solution of the dispersion equation of guided wave is obtained by programming and the dispersion curve of guided wave is drawn.The numerical value of the velocity of guided wave in any mode can be obtained and the damage on the pipeline can be located by using the principle of echo location.2)The modeling process of the finite element model used in the pipeline guided wave detection is described in detail,including the selection of the analysis module,the damage simulation method,the excitation setting,and so on.By comparing the guided wave velocity obtained by the numerical method and the guided wave velocity obtained in the finite element model,the rationality of the related modeling process is verified,which provides a basis for further study on the damage location method and the quantitative analysis of the damage degree.3)Based on the validated modeling method,a pipeline model with circumferential and axial cracks is established,and the identification effects of longitudinal and torsional guided waves under two damage conditions are studied respectively.The advantages and disadvantages of displacement,acceleration and strain echo signals in the damage identification performance are compared and analyzed.A total of 18 work conditions including different damage locations and geometry sizes are established.Based on these models,the effects of damage locations and geometry size on damage detection are studied..4)The finite element model of pipeline with different damage degree is established.On the basis of guided wave displacement echo signal,the damage index displacement echo envelope is constructed to identify the damage degree of pipeline.The neural network based on the BP training algorithm is established and the damage degree of the damaged pipeline is identified.The results show that the displacement echo envelope as the input parameter of the neural network can effectively indicate the damage degree of the pipeline.5)The echo signals of longitudinal guided wave and torsional guided wave are collected and noise is added to the echo signal,and the influence of different noise level on the ability of guided wave damage location is compared and analyzed.Based on wavelet theory,an effective de-noising method is established and the de-noising effects of hard valve method and soft valve method are compared.6)The heterogeneous pipeline model of curved pipe is established and the influence of bending angle and crack size on the intensity of guided wave echo is studied.The bending pipe with different bending angles and cracks in different geometric dimensions is set up under 16 kinds of working conditions.The heterogeneous pipeline model with interface is established and the influence of interface on the ability of guided wave damage location is studied.By simplifying the adjunct component,the heterogeneous pipeline model with adjunct component is established,and the identification strategy of pipeline damage is discussed in the case of historical detection data and no historical detection data.The innovations of this paper are as follows:1)Based on the Navier’s equation,the dispersion equation describing the relationship between the guided wave frequency and the wave velocity is established and the corresponding program is worked out.Through this program,the arbitrary specified mode guided wave velocity can be determined.Based on the singular value decomposition(SVD)method,the solution method of guided wave structure under each mode is established,which provides a basis for the selection of guided wave mode.2)On the basis of the traditional single point displacement echo signal,the damage index based on the multi-point displacement echo signal is proposed,which can be used as the input parameter of the neural network to identify the damage degree of the pipeline.And the result has higher precision and stability.