| China’s urban water supply and drainage pipe network system was built in the 1970 s and1980s,which is an important infrastructure for urban residents’ life.However,with years of disrepair,the aging of pipe network is becoming increasingly serious,and pipeline damage often occurs,which not only causes serious waste of water resources,but also has a great impact on residents’ life and urban development,and even causes pipeline blockage and traffic stagnation,ground collapse and other accidents.There are many kinds of current pipeline leakage detection methods,but the accuracy of conventional detection methods for leakage disease location is low,and the ground penetrating radar detection method has the advantages of high detection efficiency and accurate location,which can realize the disease data collection and accurate location of pipe network all over the city.However,in the current research,the two-dimensional detection technology is widely used by GPR,which is difficult to accurately describe the size,scope and location of the leakage disease.Moreover,a large number of disease detection data need to be manually processed by professionals,and the radar image feature analysis also has great subjectivity.In order to identify the GPR image of pipeline leakage more efficiently,summarize the rule of 3D radar Atlas of pipeline leakage,and realize the automatic identification of leakage disease,this paper starts from two kinds of water supply and drainage pipeline leakage disease models,carries out the GPR 3D model detection test,studies the 2D,3D radar image and 3D forward simulation of the disease.The main research contents are as follows:(1)According to the theory of electromagnetic wave and the method and technology of GPR detection,the detection mode of GPR and the setting method of technical parameters are analyzed.According to the specifications,the leakage models of single pipeline and cross pipeline are designed.The GPR 3D detection test is carried out by arranging reasonable grid lines.The radar data are processed by Radan7 software,and the characteristics of 2D image and3 D image are analyzed.(2)According to the theory of finite element method(FEM)ground penetrating radar forward modeling and scattering boundary condition,the applicability of scattering boundary condition in forward modeling is analyzed,which meets the condition of simulating electromagnetic wave propagation in infinite space in finite memory space.(3)Through Comsol finite element simulation software,the 3D forward modeling of single pipe and cross pipe leakage disease model is carried out,and the 2D and 3D image rules of pipe leakage in the incident process are analyzed.Then,the reflected signal received by the radar probe group is imaged by Voxler4 software.Finally,the results are compared with the model test results to verify the correctness of the numerical simulation results.(4)Using MATLAB programming software,combined with morphological processing method and edge detection,the 3D coordinate set of the target area of the radar image in the model test is obtained.Through point cloud imaging,the 3D slice image in the model test is transformed into 3D point cloud image,and the GUI operation interface is compiled,which realizes the automatic identification of the leakage disease,and improves the accuracy of the radar image of the water supply and drainage pipeline leakage disease Identify work efficiency and accuracy. |
PDF Full Text Request