Research On Fault Diagnosis And Identification Of Drainage Pipe Based On Acoustic Wave Detection

At present,the urbanization of our country is in the stage of rapid development.The underground drainage pipeline is an important part of the urbanization construction,which is closely related to the quality of life of the citizens,the economic development of the city and the industrial production.Due to the pipeline equipment aging,foreign intrusion and external causes of pipe leakage,blocking and deformation and other faults have occurred.The operation failure of the drainage pipe network will directly cause the waste of water resources and the occurrence of the urban waterlogging and other safety accidents.City underground drainage pipeline is an important infrastructure to ensure people's daily life,its running status is directly related to the city's normal life order and the safety of industrial production,therefore,research on the detection method suitable for underground drainage pipeline,and then extract the fault information and fault identification of pipeline,has important practical significance and research value of the city construction and industrial production.The condition monitoring and fault diagnosis of the underground pipeline can optimize the maintenance of the pipeline and ensure the safe operation of the pipeline.Acoustic detection technology is a detection technology based on the relationship between the transmitted acoustic signal and the received echo signal to analyze the state of the object to be detected.This technology is widely applied in the field of nondestructive testing.In the detection of underground drainage pipes,firstly,by analyzing the propagation characteristics of the acoustic signals in the pipeline and then extracting the characteristics that can effectively reflect the internal running status of the pipeline from the acoustic detection signals,a classifier is established to classify and identify the faults of the pipeline.Therefore,how to extract the characteristics of the fault information reflected in the pipeline from the sonic detection signal and identify the type of the fault is very crucial.The main contents of this paper are as follows:(1)Analyze the basic properties of sound waves and the commonly used acoustic parameters,the sound field of sound waves in a finite length pipe is studied,and the propagation characteristics of sound waves in different working conditions such as blocking in the pipe and having branches in the pipeline.(2)Wavelet packet and Locally Linear Embedding(LLE)are combined to extract the fault characteristics of pipeline acoustic detection signal.By wavelet packet decomposition,the original signal is decomposed into different frequency bands.Then,the sample entropy of each frequency band is calculated and the useful information extracted by using the complexity of the component is obtained.Then,the feature space composed by the sample entropy Dimension reduction,remove the redundancy of characteristic data.(3)Considering that the acoustic wave detection technology uses multiple sensors to receive the signals,the Multivariate Empirical Mode Decomposition(MEMD)method can synchronously and adaptively decompose the multi-channel data on the same scale.Therefore,MEMD is used to detect the multi-channel pipeline acoustic wave Signal analysis,a series of component signals are obtained.According to the correlation coefficient and the variance contribution rate,the main components that reflect the characteristics of the pipeline blocking are screened out,and the approximate entropy of the selected components is calculated to form the eigenvectors.The Support Vector Machine(SVM)is established to train and identify the sound wave detection signals in different running states.The experimental results show that the MEMD approximation entropy method is better than the EMD approximate entropy method to extract the effective characteristics of the pipeline blocking signal.