Gravity Anomaly And Magnetic Anomaly Inversion Detection Methods Over An Underground Horizontal Pipeline

During the excavation construction,unknown underground pipelines could be at the risk of being damaged,which might cause economic losses and even casualties.Therefore,it is extremely important to detect the buried depth and horizontal location of the underground pipeline.Forward modeling is fundamental to data processing and practical inversion.The synthetic data obtained by forward modeling is supposed to match with the simulation data in the inversion procedure,which shows the applicability and accuracy of inversion algorithm.In this paper,the experiment research object is the underground ferromagnetic pipeline of finite length.From the perspective of gravity and magnetic anomaly,different pipeline forward methods are proposed,and their correctness are verified by comparison.Then genetic algorithm and Euler deconvolution are respectively applied to the pipeline inversion,and their rapidity and accuracy are analyzed.Chapter 1 is an overview of the research motivation and significance.The theme is about the pipeline detection technology development and research status,and a variety of methods are introduced for forward modeling and inversion detection of pipeline gravity anomaly and magnetic anomaly.Chapter 2 provides two forward modeling methods to calculate the gravity anomaly and its gradient tensor of a straight horizontal pipeline of finite length.One is to deduce the forward formulas of the proposed model from the horizontal cylinder.Another is to calculate the pipeline tensor by sphere gravity formulas and superposition principle.To show the validity of this method,the former algorithm is applied for comparison.Chapter 3 provides two boundary extraction functions—HGA and V_zz-HDR to analyze the pipeline boundary recognition.When pipeline azimuth?decreases from 90°to 45°,the gravity gradient tensor and these two functions gradually lose the effectiveness of boundary recognition.When the detecting span ratio varies,the analytic signal and these two functions show different results in pipeline length and outside diameter recognition.Chapter 4 focus on the underground horizontal pipeline inversion detection,the gravity gradient tensor and its analytic signal are respectively applied to Euler deconvolution.Considering the effect of the noise,the influences of the pipeline azimuth and detection range are used to analyze the source location.Chapter 5 focus on the underground horizontal pipeline,the forward models using piece segmentation method and section segmentation method are established for comparison,based on magnetic dipole reconstruction?MDR?,.Results show that the magnetic anomalies calculated by these two methods are almost the same at different heights,inclination and declination.Afterwards,the theoretical formula of magnetic gradient tensor are deduced by the anomaly expression,the influences of the detection range d,inclination I_c and declination D to the pipeline boundary recognition and source location are analyzed.Chapter 6 proposes the total magnetic anomaly inversion procedure by genetic algorithm.The piece segmentation method and section segmentation method are respectively applied to the inversion calculation.The result shows that the section segmentation method has improved the computation efficiency a lot with a good precision.The inversed depth is accurate.Afterwards,the theoretical formula of magnetic gradient tensor are applied to inversion location by Euler deconvolution.The influences of the detection range d,inclination I_c and declination D on the source location are analyzed.Then an expression of the observed magnetic gradient tensor is put forward in this paper.An actual measurement experiment is completed by the magnetometer.The inversion results show that the observed magnetic gradient tensor can roughly identify the pipeline boundary and accurately locate the source coordinate.