Method and device for buried pipeline integrity comprehensive inspection and maintenance decision

Abstract

The application discloses a buried pipeline integrity comprehensive inspection and maintenance decision method and device, and the method comprises the following steps: determining the pipe diameter and the buried depth of the buried pipeline in a refinery, a gathering pipeline network or a station; determining a signal detection method of the buried pipeline according to the pipe diameter and the buried depth; performing signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result; inputting the signal detection result into a pipeline anomaly grade identification model trained based on a deep learning algorithm to obtain a pipeline anomaly grade identification result; and performing maintenance decision on the buried pipeline according to the pipeline anomaly grade identification result. According to the application, the appropriate signal detection method is determined according to the pipeline characteristics, then the pipeline anomaly grade identification result is determined by using the machine learning mode according to the corresponding signal detection result, and finally the maintenance decision is performed on the buried pipeline according to the pipeline anomaly grade identification result, so that the more reasonable and accurate inspection and maintenance decision can be made.

Description

Technical Field

[0001] This invention relates to the field of pipeline inspection and maintenance technology, specifically to a comprehensive inspection and maintenance decision-making method and device for the integrity of buried pipelines. Background Technology

[0002] Besides long-distance pipelines, refinery and station pipelines also play a vital role in the petrochemical industry, connecting production processes through the transport of media and serving as the lifeblood of the enterprise. Currently, long-distance pipelines can be monitored internally to promptly assess their safety status. However, for pipelines in refineries, gathering and transmission networks, and station pipelines, especially buried pipelines, internal monitoring is not feasible. As these pipelines age, how to proactively assess their safety status, formulate scientific maintenance and repair plans, and avoid blind repairs remains a critical challenge for enterprises. Currently, maintenance and repair decisions for these types of pipelines are limited to using magnetic stress detection to assess stress concentration. However, as these pipelines age, they face increasing problems: 1) Due to the large number of pipelines and significant lack of data on older pipelines, discrepancies often exist between drawings and actual site conditions, as well as information from management personnel. This creates considerable difficulties for construction, maintenance, and the expansion and upgrading of facilities and equipment. 2) Underground conditions are complex, and pipeline corrosion and thinning are widespread. Some older enterprises operate their underground pipeline networks beyond their capacity and lifespan, leading to pipe bursts, leaks, and cross-connections due to corrosion and pressure. Leaks occur frequently, posing significant safety hazards. 3) Internal inspection of underground pipelines within the plant is impossible, making it difficult to promptly assess pipeline safety. 4) Current magnetic stress detection and analysis methods can qualitatively detect pipeline stress concentration but cannot quantify defects, and their accuracy is relatively low. Simple magnetic stress detection has certain limitations. Summary of the Invention

[0003] To address the problems existing in the prior art, this invention proposes a method and device for comprehensive inspection and maintenance decision-making regarding the integrity of buried pipelines.

[0004] Specifically, the embodiments of the present invention provide the following technical solutions:

[0005] In a first aspect, embodiments of the present invention provide a comprehensive maintenance and repair decision-making method for the integrity of buried pipelines, including:

[0006] Determine the diameter and burial depth of buried pipelines within refineries, gathering and transmission networks, or stations;

[0007] The signal detection method for the buried pipeline is determined based on the pipe diameter and the burial depth.

[0008] The buried pipeline is subjected to signal detection according to the aforementioned signal detection method to obtain the signal detection result;

[0009] input the signal detection result into a pipeline anomaly level identification model trained based on a deep learning algorithm to obtain a pipeline anomaly level identification result;

[0010] According to the pipeline anomaly level identification result, a maintenance decision is made for the buried pipeline.

[0011] Further, a signal detection method for the buried pipeline is determined according to the pipe diameter and the burial depth, comprising:

[0012] According to the pipe diameter and the burial depth, it is determined whether the magnetic stress detection condition and the ultrasonic guided wave detection condition are met, if both are met, it is determined that the signal detection method for the buried pipeline comprises magnetic stress detection and ultrasonic guided wave detection; accordingly, the buried pipeline is subjected to magnetic stress detection and ultrasonic guided wave detection to obtain magnetic stress signal detection result and ultrasonic guided wave signal detection result;

[0013] If only the magnetic stress detection condition is met, it is determined that the signal detection method for the buried pipeline comprises magnetic stress detection; accordingly, the buried pipeline is subjected to magnetic stress detection to obtain the magnetic stress signal detection result;

[0014] If only the ultrasonic guided wave detection condition is met, it is determined that the signal detection method for the buried pipeline comprises ultrasonic guided wave detection; accordingly, the buried pipeline is subjected to ultrasonic guided wave detection to obtain the ultrasonic guided wave signal detection result.

[0015] Further, the buried pipeline integrity comprehensive detection and maintenance decision method further comprises: for the detection signal, it is determined whether the detection abnormal point is affected by interference and the signal abnormality affected by the interference is removed.

[0016] Further, the signal detection result is input into a pipeline anomaly level identification model trained based on a deep learning algorithm to obtain a pipeline anomaly level identification result, comprising:

[0017] The magnetic stress signal detection result is input into a first pipeline anomaly level identification model to obtain a defect position on the pipeline and an anomaly level identification result;

[0018] The first pipeline anomaly level identification model is a model obtained by training and testing a neural network model, wherein the magnetic stress signal detection result of a sample pipeline with a pre-determined pipeline anomaly level is used as the input of the model, and the defect position on the sample pipeline and the corresponding anomaly level are used as the output of the model.

[0019] Or,

[0020] inputting the ultrasonic guided wave signal detection result into a second pipeline anomaly level identification model to obtain a defect position on the pipeline and an anomaly level identification result; wherein the second pipeline anomaly level identification model is a model obtained by taking the ultrasonic guided wave signal detection result of a sample pipeline with a pre-determined pipeline anomaly level as an input, taking the defect position on the sample pipeline and the corresponding anomaly level as an output, and training and testing a neural network model;

[0021] or,

[0022] inputting both detection results into the first and second pipeline anomaly level identification models to obtain the defect position on the pipeline and the anomaly level identification result respectively, and then merging the identification results of the two detections.

[0023] Further, inputting both detection results into the first and second pipeline anomaly level identification models to obtain the defect position on the pipeline and the anomaly level identification result respectively, and then merging the identification results of the two detections, including:

[0024] performing a union set processing on the defect position obtained according to the first pipeline anomaly level identification model and the defect position obtained according to the second pipeline anomaly level identification model to determine the defect position on the pipeline, and at the same time, performing data fusion on the anomaly level obtained according to the first pipeline anomaly level identification model and the anomaly level obtained according to the second pipeline anomaly level identification model according to the prediction accuracy weight of the two models to determine the corresponding anomaly level on the pipeline;

[0025] or,

[0026] performing a union set processing on the defect position obtained according to the first pipeline anomaly level identification model and the defect position obtained according to the second pipeline anomaly level identification model to determine the defect position on the pipeline, and at the same time, taking the maximum value of the anomaly level obtained according to the first pipeline anomaly level identification model and the anomaly level obtained according to the second pipeline anomaly level identification model as the corresponding anomaly level on the pipeline.

[0027] Further, the training process of the first pipeline anomaly level identification model includes:

[0028] determining a preset number of sample pipelines, and obtaining a magnetic stress signal detection result of the preset number of sample pipelines; wherein the preset number of sample pipelines need to cover various defect positions and various anomaly levels;

[0029] determining the defect position and the anomaly level of the preset number of sample pipelines according to the experience of a detection signal analysis expert or excavation;

[0030] The magnetic stress signal detection result of a sample pipeline with a previously determined pipeline anomaly level is taken as the input of the model, and the defect position and anomaly level of the sample pipeline are taken as the output of the model, the neural network model is trained and tested, and a first pipeline anomaly level identification model is obtained.

[0031] The training process of the second pipeline anomaly level identification model comprises:

[0032] A preset number of sample pipelines are determined, and the ultrasonic guided wave signal detection result of the preset number of sample pipelines is obtained; wherein the preset number of sample pipelines need to cover various defect positions and various anomaly levels.

[0033] The defect position and anomaly level of the preset number of sample pipelines are determined according to the detection signal analysis expert experience or excavation.

[0034] The ultrasonic guided wave signal detection result of a sample pipeline with a previously determined pipeline anomaly level is taken as the input of the model, and the defect position and anomaly level of the sample pipeline are taken as the output of the model, the neural network model is trained and tested, and a second pipeline anomaly level identification model is obtained.

[0035] Further, according to the pipeline anomaly level identification result, a maintenance decision is made for the buried pipeline, comprising:

[0036] If the pipeline anomaly level identification result is mild or moderate anomaly, the buried pipeline is marked as a pipeline for regular monitoring.

[0037] If the pipeline anomaly level identification result is severe anomaly, on-site investigation is performed on the buried pipeline to determine whether excavation conditions are met, for the pipeline meeting the excavation verification conditions, a point with a serious anomaly of the detection signal is selected for excavation, and further detection is performed by using ultrasonic thickness measurement, laser scanning, TOFD crack detection and manual measurement means, the defect size is further quantified, and the detection result is collected; wherein for the defect after excavation, the defect size information is recorded as a sample for further correcting the model, and according to the defect type, defect size and defect position information, the suitability of the pipeline section with the defect is evaluated, and a defect maintenance decision is made, including repair level, planned repair time and repair method.

[0038] Further, before the magnetic stress detection or ultrasonic guided wave detection of the pipeline, the method further comprises:

[0039] An infrared leakage detector is used to perform infrared leakage screening on the buried pipeline to determine whether there is a leakage pipeline.

[0040] In a second aspect, the embodiments of the present application also provide a buried pipeline integrity comprehensive detection and maintenance decision device, comprising:

[0041] a first determining module configured to determine a pipe diameter and a buried depth of a buried pipeline in a refinery, a gathering pipeline network or a station field;

[0042] a second determining module configured to determine a signal detection method of the buried pipeline according to the pipe diameter and the buried depth;

[0043] a detecting module configured to perform signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result;

[0044] a recognizing module configured to input the signal detection result into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result;

[0045] a maintenance decision module configured to perform maintenance decision on the buried pipeline according to the pipeline anomaly level recognition result.

[0046] Further, the second determining module is specifically configured to:

[0047] determine whether the magnetic stress detection condition and the ultrasonic guided wave detection condition are both satisfied according to the pipe diameter and the buried depth, and if so, determine that the signal detection method of the buried pipeline comprises the magnetic stress detection and the ultrasonic guided wave detection;

[0048] if only the magnetic stress detection condition is satisfied, determine that the signal detection method of the buried pipeline comprises the magnetic stress detection;

[0049] if only the ultrasonic guided wave detection condition is satisfied, determine that the signal detection method of the buried pipeline comprises the ultrasonic guided wave detection.

[0050] As can be seen from the above technical solution, the buried pipeline integrity comprehensive inspection and maintenance decision method and device provided by the embodiments of the present application first determine the pipe diameter and the buried depth of a buried pipeline in a refinery, a gathering pipeline network or a station field, then determine the signal detection method of the buried pipeline according to the pipe diameter and the buried depth, then perform signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result, then input the signal detection result into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result, and finally perform maintenance decision on the buried pipeline according to the pipeline anomaly level recognition result. As can be seen, the embodiments of the present application determine a suitable signal detection method according to the pipeline characteristics, then determine a pipeline anomaly level recognition result in a machine learning manner according to the corresponding signal detection result, and finally perform inspection and maintenance decision on the buried pipeline according to the pipeline anomaly level recognition result, so that a more reasonable and accurate maintenance decision can be made. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to make the technical solutions of the embodiments of the present application or the prior art clearer, the accompanying drawings needed in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description only need to be some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort.

[0052] Figure 1 is a flow chart of the buried pipeline integrity comprehensive inspection and maintenance decision method provided by an embodiment of the present application;

[0053] Figure 2 is the intention of the implementation process of the buried pipeline integrity comprehensive inspection and maintenance decision method provided by an embodiment of the present application;

[0054] Figure 3 is a structural schematic diagram of the buried pipeline integrity comprehensive inspection and maintenance decision device provided by an embodiment of the present application;

[0055] Figure 4 is a structural schematic diagram of the electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0056] The specific embodiments of the present application will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and cannot be used to limit the protection scope of the present application.

[0057] The present application provides a buried pipeline integrity comprehensive inspection and maintenance decision method, and particularly relates to a comprehensive inspection and maintenance decision method for buried pipelines in a refinery, a gathering pipeline network or a station field which cannot be internally detected. The buried pipeline integrity comprehensive inspection and maintenance decision method provided by the present application will be explained and described below through specific embodiments. Figure 1 shows a flow chart of the buried pipeline integrity comprehensive inspection and maintenance decision method provided by an embodiment of the present application, as shown in Figure 1 The buried pipeline integrity comprehensive inspection and maintenance decision method provided by the embodiment of the present application specifically includes the following contents:

[0058] Step 101: determining the pipe diameter and the buried depth of the buried pipeline in the refinery, the gathering pipeline network or the station field;

[0059] In this step, the pipe diameter and the buried depth of the buried pipeline in the refinery, the gathering pipeline network or the station field need to be determined first, and then the signal detection method of the buried pipeline is determined according to the pipe diameter and the buried depth of the buried pipeline. For example, the detection conditions corresponding to the magnetic stress detection and the detection conditions corresponding to the ultrasonic guided wave detection are different. Specifically, the detection conditions corresponding to the magnetic stress detection generally require that the pipe diameter D of the detected pipeline should be greater than 0.1 times the buried depth. For the pipeline with a buried depth of 1.5 m, the pipe diameter of the detection object should be greater than 150 mm; the detection conditions corresponding to the ultrasonic guided wave detection are different according to different detection equipment. For example, the ultrasonic guided wave equipment with a clamp band requires that the pipe diameter D is greater than 51 mm and has a position for installing a sensor.

[0060] Step 102: determining the signal detection method of the buried pipeline according to the pipe diameter and the buried depth;

[0061] Step 103: performing signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result;

[0062] In this step, the signal detection method suitable for the buried pipeline is determined according to the pipe diameter and the buried depth. After the signal detection method suitable for the buried pipeline is determined, signal detection is performed on the buried pipeline according to the suitable signal detection method to obtain a signal detection result. For example, for the pipeline on which the magnetic stress detection has been carried out and which has the conditions for ultrasonic guided wave detection, further ultrasonic guided wave depth detection is carried out. For the pipeline which does not have the conditions for magnetic stress detection but has the conditions for ultrasonic guided wave detection (such as the pipeline with a pipe diameter greater than 51 m and less than 150 mm), ultrasonic guided wave detection is directly carried out to collect the signal detection result data.

[0063] Step 104: inputting the signal detection result into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result;

[0064] In this step, after the signal detection result is obtained, the signal detection result is input into a pipeline anomaly level recognition model trained based on a deep learning algorithm, and a pipeline anomaly level recognition result is obtained through machine intelligent learning, so as to be not affected by human intervention, thereby effectively improving the accuracy and reliability of the recognition, and also improving the efficiency and saving the labor cost.

[0065] Step 105: making a maintenance decision for the buried pipeline according to the pipeline anomaly level recognition result.

[0066] In this step, after the anomaly level recognition result is obtained, a maintenance decision for the buried pipeline is made according to the pipeline anomaly level recognition result. For example, if the pipeline anomaly level recognition result is a mild or moderate anomaly, the buried pipeline is marked as a pipeline to be periodically monitored.

[0067] If the pipe anomaly level identification result is severe anomaly, on-site investigation is performed on the buried pipeline to determine whether excavation conditions are met, for the pipeline meeting the excavation verification conditions, a point with a serious abnormal detection signal is selected for excavation, and further detection is performed by using ultrasonic thickness measurement, laser scanning, TOFD crack detection and manual measurement means to further quantify the defect size, and the detection results are collected; wherein for the defects after excavation, the defect size information is recorded as a sample for further correcting the model for model learning, and according to the defect type, defect size and defect position information, the suitability evaluation of the pipe section with defects is performed to make a defect repair decision, including repair level, planned repair time and repair method.

[0068] According to the technical solution, the buried pipeline integrity comprehensive inspection and repair decision method provided by the embodiment of the present application first determines the pipe diameter and the buried depth of the buried pipeline in the refinery, gathering pipeline network or station, then determines the signal detection method of the buried pipeline according to the pipe diameter and the buried depth, then performs signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result, then inputs the signal detection result into a pipe anomaly level identification model trained based on a deep learning algorithm to obtain a pipe anomaly level identification result, and finally makes a repair decision for the buried pipeline according to the pipe anomaly level identification result. As can be seen, the embodiment of the present application determines a suitable signal detection method according to the characteristics of the pipeline, then determines a pipe anomaly level identification result by using a machine learning method according to the corresponding signal detection result, and finally makes a repair decision for the buried pipeline according to the pipe anomaly level identification result, so that a more reasonable and accurate inspection and repair decision can be made.

[0069] Based on the content of the above embodiment, in the present embodiment, the signal detection method of the buried pipeline is determined according to the pipe diameter and the buried depth, including:

[0070] According to the pipe diameter and the buried depth, it is judged whether the magnetic stress detection condition and the ultrasonic guided wave detection condition are met, if both are met, it is determined that the signal detection method of the buried pipeline includes magnetic stress detection and ultrasonic guided wave detection;

[0071] If only the magnetic stress detection condition is met, it is determined that the signal detection method of the buried pipeline includes magnetic stress detection;

[0072] If only the ultrasonic guided wave detection condition is met, it is determined that the signal detection method of the buried pipeline includes ultrasonic guided wave detection.

[0073] The embodiment determines the signal detection method suitable for the pipeline according to the pipe diameter and the buried depth of the pipeline, and then performs pipeline detection according to the signal detection method suitable for the pipeline.

[0074] Further, it needs to be explained that before the magnetic stress detection or the ultrasonic guided wave detection is performed on the pipeline, the method further comprises: using an infrared leakage detector to perform infrared leakage screening on the buried pipeline to determine whether there is a leaking pipeline.

[0075] Therefore, it can be seen that by using the non-contact detection methods such as infrared leakage, magnetic stress detection, and ultrasonic guided wave detection, the embodiment solves the problem that the detection accuracy of the buried pipeline, especially the buried pipeline in a refinery, gathering and transportation pipeline network or station, is low and maintenance decision cannot be made.

[0076] Based on the content of the above embodiment, in the embodiment, the buried pipeline is signal detected according to the signal detection method to obtain a signal detection result, which comprises:

[0077] If the signal detection method of the buried pipeline comprises the magnetic stress detection and the ultrasonic guided wave detection, the buried pipeline is subjected to the magnetic stress detection and the ultrasonic guided wave detection to obtain a magnetic stress signal detection result and an ultrasonic guided wave signal detection result;

[0078] If the signal detection method of the buried pipeline comprises the magnetic stress detection, the buried pipeline is subjected to the magnetic stress detection to obtain a magnetic stress signal detection result;

[0079] If the signal detection method of the buried pipeline comprises the ultrasonic guided wave detection, the buried pipeline is subjected to the ultrasonic guided wave detection to obtain an ultrasonic guided wave signal detection result.

[0080] Based on the content of the above embodiment, in the embodiment, the signal detection result is input into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result, which comprises:

[0081] The magnetic stress signal detection result is input into a first pipeline anomaly level recognition model to obtain a defect position on the pipeline and an anomaly level recognition result;

[0082] The first pipeline anomaly level recognition model is a model obtained by training and testing a neural network model, in which the magnetic stress signal detection result of a sample pipeline with a previously determined pipeline anomaly level is used as the input of the model, and the defect position on the sample pipeline and the corresponding anomaly level are used as the output of the model.

[0083] Or,

[0084] The ultrasonic guided wave signal detection result is input into a second pipeline anomaly level identification model to obtain a defect position on the pipeline and an anomaly level identification result; the second pipeline anomaly level identification model is a model in which ultrasonic guided wave signal detection results of sample pipelines for which pipeline anomaly levels have been determined in advance are input, defect positions on the sample pipelines and corresponding anomaly levels are output, and a neural network model is trained and tested to obtain the model;

[0085] or,

[0086] Both detection results are input into the first and second pipeline anomaly level identification models to obtain defect positions on the pipeline and anomaly level identification results, and the identification results of the two detections are combined.

[0087] In this embodiment, when the signal detection result is input into a pipeline anomaly level identification model trained based on a deep learning algorithm to obtain a pipeline anomaly level identification result, different pipeline anomaly level identification models are used to identify the magnetic stress signal detection result and the ultrasonic guided wave signal detection result, respectively, to obtain more accurate identification results. When magnetic stress signal detection and ultrasonic wave guide signal detection are performed simultaneously, the defect positions and corresponding anomaly levels detected by the magnetic stress signal detection result and the defect positions and corresponding anomaly levels obtained by the ultrasonic wave guide signal detection result are combined to determine the defect positions and corresponding anomaly levels on the pipeline, thereby making the identification results more comprehensive and reliable.

[0088] Based on the content of the above embodiment, in this embodiment, both detection results are input into the first and second pipeline anomaly level identification models to obtain defect positions on the pipeline and anomaly level identification results, and the identification results of the two detections are combined, including:

[0089] The defect positions obtained according to the first pipeline anomaly level identification model and the defect positions obtained according to the second pipeline anomaly level identification model are processed in a union set to determine the defect positions on the pipeline, and the anomaly levels obtained according to the first pipeline anomaly level identification model and the anomaly levels obtained according to the second pipeline anomaly level identification model are data fused according to the prediction accuracy weights of the two models to determine the corresponding anomaly levels on the pipeline;

[0090] or

[0091] The defect positions obtained according to the first pipeline anomaly level identification model and the defect positions obtained according to the second pipeline anomaly level identification model are processed in a union set, and the defect positions on the pipeline are determined, and the maximum value of the anomaly levels obtained according to the first pipeline anomaly level identification model and the anomaly levels obtained according to the second pipeline anomaly level identification model is taken as the corresponding anomaly level on the pipeline.

[0092] In the embodiment, it should be noted that when the magnetic stress signal detection and the ultrasonic waveguide signal detection are simultaneously performed, the defect positions and the corresponding anomaly levels obtained by the magnetic stress signal detection result can be comprehensively determined, and the defect positions and the corresponding anomaly levels obtained by the ultrasonic waveguide signal detection result can be comprehensively determined to determine the defect positions and the corresponding anomaly levels on the pipeline, so that the identification result is more comprehensive and reliable. For example, in one implementation, the defect positions obtained according to the first pipeline anomaly level identification model and the defect positions obtained according to the second pipeline anomaly level identification model are processed in a union set to determine the defect positions on the pipeline, and the anomaly levels obtained according to the first pipeline anomaly level identification model and the anomaly levels obtained according to the second pipeline anomaly level identification model are fused according to the model prediction accuracy weight to determine the corresponding anomaly level on the pipeline, so that the defect positions are more complete and there is no omission, and the anomaly level is more reasonable and accurate. For another example, in another implementation, the defect positions obtained according to the first pipeline anomaly level identification model and the defect positions obtained according to the second pipeline anomaly level identification model are processed in a union set to determine the defect positions on the pipeline, and the maximum value of the anomaly levels obtained according to the first pipeline anomaly level identification model and the anomaly levels obtained according to the second pipeline anomaly level identification model is taken as the corresponding anomaly level on the pipeline, so that the defect positions are more complete and there is no omission, and the maximum value of the two is selected as the anomaly level, which can maximize the safety of the pipeline.

[0093] Based on the content of the above embodiment, in the embodiment, the training process of the first pipeline anomaly level identification model comprises:

[0094] A preset number of sample pipelines are determined, and the magnetic stress signal detection results of the preset number of sample pipelines are obtained; wherein the preset number of sample pipelines need to cover various defect positions and various anomaly levels;

[0095] The defect positions and the anomaly levels of the preset number of sample pipelines are determined according to the experience or excavation of a detection signal analysis expert;

[0096] The magnetic stress signal detection result of the sample pipeline with a previously determined pipeline anomaly level is taken as the input of the model, and the defect position and anomaly level of the sample pipeline are taken as the output of the model, the neural network model is trained and tested, and a first pipeline anomaly level identification model is obtained.

[0097] The training process of the second pipeline anomaly level identification model comprises:

[0098] A preset number of sample pipelines are determined, and an ultrasonic guided wave signal detection result of the preset number of sample pipelines is obtained; wherein the preset number of sample pipelines need to cover various defect positions and various anomaly levels.

[0099] The defect position and anomaly level of the preset number of sample pipelines are determined according to the detection signal analysis expert experience or excavation;

[0100] The ultrasonic guided wave signal detection result of the sample pipeline with a previously determined pipeline anomaly level is taken as the input of the model, and the defect position and anomaly level of the sample pipeline are taken as the output of the model, the neural network model is trained and tested, and a second pipeline anomaly level identification model is obtained.

[0101] In the embodiment, the training process of the first pipeline anomaly level identification model and the second pipeline anomaly level identification model is given, and through the training of the model, the trained model can accurately identify the defect position and anomaly level. It can be understood that when the model is trained, CNN or RNN model can be used for deep learning.

[0102] Based on the content of the above embodiment, in the embodiment, according to the pipeline anomaly level identification result, a maintenance decision is made for the buried pipeline, comprising:

[0103] If the pipeline anomaly level identification result is mild or moderate anomaly, the buried pipeline is marked as a pipeline for regular monitoring;

[0104] If the pipeline anomaly level identification result is severe anomaly, on-site investigation is performed on the buried pipeline to determine whether the excavation condition is met, for the pipeline meeting the excavation verification condition, a point with a serious detection signal anomaly is selected for excavation, and further detection is performed by using ultrasonic thickness measurement, laser scanning, TOFD crack detection and manual measurement means, the defect size is further quantified, and the detection result is collected; wherein for the defect after excavation, the defect size information is recorded as a sample for model learning to further correct the model, and according to the defect type, defect size and defect position information, the suitability evaluation of the pipeline section with defects is performed, and the defect maintenance decision is made, including repair level, planned repair time and repair method.

[0105] According to the above description, the buried pipeline integrity comprehensive inspection and maintenance decision method provided in the embodiment specifically includes pipeline design data and field data collection, determination of pipeline diameter, connecting equipment, buried depth, trend and other data; classified development of magnetic stress and ultrasonic guided wave detection; interference material is investigated according to the detection result, and the detection result is classified according to the severity; for severe abnormalities, for those with excavation conditions, excavation verification and detection by contact detection technology are carried out, finally, after investigating signal abnormalities, integrity evaluation and maintenance decision are made according to the signal severity and quantitative detection results. For those without excavation conditions, further analysis by experts is needed to determine whether immediate repair or repair within one year is needed, if needed, timely excavation repair is needed; the following will be combined with Figure 2 The buried pipeline integrity comprehensive inspection and maintenance decision method provided in the embodiment is described in detail as follows:

[0106] First step, data research. Through consulting design data, field investigation, collecting data such as pipeline diameter, buried depth, connecting equipment, whether having magnetic stress or ultrasonic guided wave detection conditions of buried pipeline in refinery, gathering pipeline network or station, the pipeline is divided into four categories through pipeline diameter and buried depth: ① magnetic stress and ultrasonic guided wave detection can be carried out simultaneously, ② ultrasonic guided wave detection can be carried out but magnetic stress detection cannot be carried out, ③ magnetic stress detection can be carried out but ultrasonic guided wave detection cannot be carried out, ④ neither of the two detections has conditions. The basis for the division is to judge according to the magnetic stress detection conditions and the ultrasonic guided wave detection conditions. For detection conditions, different detection equipment requires different, specifically, such as magnetic stress detection generally requires that the diameter D of the pipeline to be detected should be greater than 0.1 times the buried depth, for the pipeline with buried depth of 1.5 m, the diameter of the detection object should be greater than 150 mm; ultrasonic guided wave detection also varies according to different equipment, such as the clamp band type ultrasonic guided wave equipment requires that the pipeline diameter D should be greater than 51 mm, and has a position for installing sensor.

[0107] Second step, field detection. The pipeline is detected and positioned, and the GPS probe can be installed on the gyro or ground detection equipment to accurately detect the pipeline positioning and collect the detection results. The infrared leakage detector is used to screen the buried pipeline near the device and the device connected to the buried pipeline in the refinery, gathering pipeline or station, to determine whether there is a leakage pipeline and collect the detection results. For the pipelines with magnetic stress detection, non-contact stress concentration detection equipment is used to detect the pipeline on the ground along the pipeline direction, and the detection signal results and surrounding interference such as trees, power poles and metal objects are collected. For the detection signal, it is checked whether the abnormal point is affected by interference, and the signal abnormality affected by interference is removed. Then, for the pipelines that have carried out magnetic stress detection and have ultrasonic guided wave detection conditions, further ultrasonic guided wave detection is carried out, and for the pipelines that do not have magnetic stress detection but have ultrasonic guided wave detection conditions (such as pipelines with D>51m and D<150mm), ultrasonic guided wave detection is directly carried out, and the detection result data is collected. According to the detection results, the magnetic stress and ultrasonic guided wave detection signals are evaluated by machine learning method. On the one hand, the two kinds of detection signals can be evaluated by the model, and on the other hand, the intelligent evaluation model based on weight data fusion of the two detection methods can be used for evaluation. The repair decision level is divided into three categories: light, medium and heavy. For the buried pipelines with light and medium signal abnormalities, they can be temporarily marked as periodic monitoring pipelines, and for the pipelines with severe signal abnormalities, further verification or analysis of the detection results is required.

[0108] Third step, further verification. For the severe abnormal signals in the second step, further field investigation is carried out to determine whether the excavation condition is met. For the pipelines with excavation verification conditions, the points with severe abnormal detection signals are selected for excavation, and further detection is carried out by ultrasonic thickness measurement, laser scanning, TOFD crack detection and manual measurement, etc. to further quantify the defect size and collect the detection results.

[0109] Fourth step, repair decision. For the pipelines without excavation verification conditions in the second step, further detailed investigation of the surrounding environmental interference is carried out. After the interference is determined to be removed, the expert intervention method is further used to evaluate the abnormal signal. If it is still severe, the repair decision level is determined as "repair immediately or plan to repair within one year". For the defects after excavation in the third step, the defect size information is recorded as a sample for further correction of the model by machine learning method. According to the defect type, defect size, defect position and other information, the suitability of the pipeline with defects is evaluated, and the defect repair decision is made, including repair level (immediate repair, planned repair, monitoring use), planned repair time and repair method.

[0110] According to the above description, the embodiment provides a buried pipeline integrity comprehensive inspection and maintenance decision method. The embodiment adopts infrared leakage, magnetic stress detection, ultrasonic guided wave detection and other non-contact detection and ultrasonic thickness measurement, laser scanning, TOFD (Time of Flight Diffraction) detection and other contact detection, solves the problem that the detection precision of the buried pipeline, especially the buried pipeline in a refinery, gathering pipeline network or station field is low and maintenance decision cannot be implemented.

[0111] Figure 3 A structure diagram of a buried pipeline integrity comprehensive inspection and maintenance decision device provided by an embodiment of the application is shown, as shown in Figure 3 The buried pipeline integrity comprehensive inspection and maintenance decision device provided by the embodiment of the application comprises:

[0112] A first determination module 21 is configured to determine the pipe diameter and the buried depth of the buried pipeline in the refinery, gathering pipeline network or station field.

[0113] A second determination module 22 is configured to determine the signal detection method of the buried pipeline according to the pipe diameter and the buried depth.

[0114] A detection module 23 is configured to perform signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result.

[0115] An identification module 24 is configured to input the signal detection result into a pipeline anomaly level identification model trained based on a deep learning algorithm to obtain a pipeline anomaly level identification result.

[0116] A maintenance decision module 25 is configured to perform maintenance decision on the buried pipeline according to the pipeline anomaly level identification result.

[0117] Based on the content of the above embodiment, in the embodiment, the second determination module 22 is specifically configured to:

[0118] determine whether the magnetic stress detection condition and the ultrasonic guided wave detection condition are met according to the pipe diameter and the buried depth, and if both conditions are met, determine that the signal detection method of the buried pipeline comprises magnetic stress detection and ultrasonic guided wave detection.

[0119] if only the magnetic stress detection condition is met, determine that the signal detection method of the buried pipeline comprises magnetic stress detection.

[0120] if only the ultrasonic guided wave detection condition is met, determine that the signal detection method of the buried pipeline comprises ultrasonic guided wave detection.

[0121] The buried pipeline integrity comprehensive inspection and maintenance decision device provided by the embodiment can be used to implement the buried pipeline integrity comprehensive inspection and maintenance decision method provided by the above embodiment, and has similar working principles and beneficial effects, which will not be described in detail here.

[0122] Based on the same inventive concept, another embodiment of the present application provides an electronic device, referring to Figure 4 , which specifically comprises the following contents: a processor 301, a memory 302, a communication interface 303 and a communication bus 304.

[0123] The processor 301, the memory 302 and the communication interface 303 can communicate with each other through the communication bus 304; the communication interface 303 is used to realize information transmission between devices.

[0124] The processor 301 is used to call a computer program in the memory 302, and the processor realizes all steps of the above-mentioned buried pipeline integrity comprehensive inspection and maintenance decision method when executing the computer program, for example, the processor realizes the following steps when executing the computer program: determining the pipe diameter and the buried depth of the buried pipeline in the refinery, the gathering pipeline network or the station field; determining the signal detection method of the buried pipeline according to the pipe diameter and the buried depth; performing signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result; inputting the signal detection result into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result; and making a maintenance decision for the buried pipeline according to the pipeline anomaly level recognition result.

[0125] Based on the same inventive concept, another embodiment of the present application provides a non-transitory computer readable storage medium, which stores a computer program, and the computer program realizes all steps of the above-mentioned buried pipeline integrity comprehensive inspection and maintenance decision method when executed by a processor, for example, the processor realizes the following steps when executing the computer program: determining the pipe diameter and the buried depth of the buried pipeline in the refinery, the gathering pipeline network or the station field; determining the signal detection method of the buried pipeline according to the pipe diameter and the buried depth; performing signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result; inputting the signal detection result into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result; and making a maintenance decision for the buried pipeline according to the pipeline anomaly level recognition result.

[0126] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of a software function unit and sold or used as an independent product, and can be stored in a computer-readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0127] The device embodiments described above are only schematic, wherein the units illustrated as separate components can or can not be physically separated, and the components illustrated as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiments of the present application. Those skilled in the art can understand and implement without creative labor.

[0128] From the above description of the embodiments, those skilled in the art can clearly understand that the embodiments can be implemented by means of software and the necessary general hardware platform, and of course can also be implemented by hardware. Based on such understanding, the above technical solutions essentially or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the buried pipeline integrity comprehensive inspection and maintenance decision method described in each embodiment or some part of the embodiment.

[0129] In addition, in the present application, such as "first", "second" is only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise specifically limited.

[0130] Moreover, in the subject specification, the term "engagement" simply denotes the relationship between or among multiple entities or operations, and does not necessarily require any such actual relationship or order between or among such entities or operations. Also, the terminology "includes," "has," "holds," "contains" or any other variant thereof is intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements is not necessarily limited to those elements, but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0131] Furthermore, in the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the particular feature, structure, material or characteristic being described in connection with the embodiment or example contains in at least one embodiment or example of the present application. The illustrative representation of the above terms in the specification does not necessarily refer to the same embodiment or example. Moreover, the particular feature, structure, material, or characteristic being described can be combined in any suitable manner in one or more embodiments or examples. Furthermore, different embodiments or examples described in the specification can be combined and combined with features of other embodiments or examples, if such combination does not result in a contradiction.

[0132] Finally, it should be noted that the above embodiments are merely used to illustrate the technical solutions of the present application, rather than limiting them; even though the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some technical features; and such modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A buried pipeline integrity comprehensive inspection and maintenance decision method, characterized by, The method comprises the following steps: determining the pipe diameter and the buried depth of a buried pipeline in a refinery, a gathering pipeline network or a station field; determining a signal detection method of the buried pipeline according to the pipe diameter and the buried depth; performing signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result; inputting the signal detection result into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result; making a maintenance decision for the buried pipeline according to the pipeline anomaly level recognition result; determining the signal detection method of the buried pipeline according to the pipe diameter and the buried depth, which comprises the following steps: determining whether the magnetic stress detection condition and the ultrasonic guided wave detection condition are met according to the pipe diameter and the buried depth, and if both conditions are met, determining that the signal detection method of the buried pipeline comprises magnetic stress detection and ultrasonic guided wave detection; accordingly, performing magnetic stress detection and ultrasonic guided wave detection on the buried pipeline to obtain a magnetic stress signal detection result and an ultrasonic guided wave signal detection result; if only the magnetic stress detection condition is met, determining that the signal detection method of the buried pipeline comprises magnetic stress detection; accordingly, performing magnetic stress detection on the buried pipeline to obtain a magnetic stress signal detection result; if only the ultrasonic guided wave detection condition is met, determining that the signal detection method of the buried pipeline comprises ultrasonic guided wave detection; accordingly, performing ultrasonic guided wave detection on the buried pipeline to obtain an ultrasonic guided wave signal detection result; before performing magnetic stress detection or ultrasonic guided wave detection on the pipeline, the method further comprises the following steps: performing positioning detection on the pipeline, which can be performed by a gyroscope or a ground detection device installed with a GPS probe to accurately detect the positioning of the pipeline and collect the detection result; using an infrared leakage detector to perform infrared leakage screening on the buried pipeline to determine whether there is a leaking pipeline and collect the detection result; the magnetic stress detection on the pipeline comprises the following steps: for the pipeline that is suitable for magnetic stress detection, using a non-contact stress concentration detection device to perform segmented detection on the pipeline along the pipeline direction on the ground, while collecting the detection signal result and the surrounding interference objects such as trees, power poles and metal objects; for the detection signal, checking whether the abnormal points are affected by the interference objects, and removing the signal abnormalities affected by the interference objects; inputting the signal detection result into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result, which comprises the following steps: inputting the magnetic stress signal detection result into a first pipeline anomaly level recognition model to obtain a defect position on the pipeline and an anomaly level recognition result; wherein the first pipeline anomaly level recognition model is obtained by training and testing a neural network model, in which the magnetic stress signal detection result of a sample pipeline with a pre-determined pipeline anomaly level is used as the input of the model, and the defect position on the sample pipeline and the corresponding anomaly level are used as the output of the model; or inputting the ultrasonic guided wave signal detection result into a second pipeline anomaly level identification model to obtain a defect position on the pipeline and an anomaly level identification result; the second pipeline anomaly level identification model is obtained by taking the ultrasonic guided wave signal detection result of a sample pipeline with a pre-determined pipeline anomaly level as the input of the model, taking the defect position on the sample pipeline and the corresponding anomaly level as the output of the model, and training and testing a neural network model; or, inputting both detection results into the first and second pipeline anomaly level identification models to obtain the defect position on the pipeline and the anomaly level identification result, and then merging the identification results of the two detections; inputting both detection results into the first and second pipeline anomaly level identification models to obtain the defect position on the pipeline and the anomaly level identification result, and then merging the identification results of the two detections, including: performing a union set process on the defect position obtained according to the first pipeline anomaly level identification model and the defect position obtained according to the second pipeline anomaly level identification model to determine the defect position on the pipeline, and performing data fusion on the anomaly level obtained according to the first pipeline anomaly level identification model and the anomaly level obtained according to the second pipeline anomaly level identification model according to the prediction accuracy weight of the two models to determine the corresponding anomaly level on the pipeline; or, performing a union set process on the defect position obtained according to the first pipeline anomaly level identification model and the defect position obtained according to the second pipeline anomaly level identification model to determine the defect position on the pipeline, and taking the maximum value of the anomaly level obtained according to the first pipeline anomaly level identification model and the anomaly level obtained according to the second pipeline anomaly level identification model as the corresponding anomaly level on the pipeline; based on the pipeline anomaly level identification result, making a maintenance decision for the buried pipeline, including: if the pipeline anomaly level identification result is mild or moderate anomaly, marking the buried pipeline as a pipeline to be monitored regularly; if the pipeline anomaly level identification result is severe anomaly, conducting a site investigation on the buried pipeline to determine whether the pipeline has excavation conditions, for the pipeline that has the excavation verification conditions, selecting a point with a severely abnormal detection signal, excavating, and further detecting by using ultrasonic thickness measurement, laser scanning, TOFD crack detection, and manual measurement to further quantify the defect size and collect the detection result; wherein for the defect after excavation, the defect size information is recorded as a sample for model learning to further correct the model, and based on the defect type, defect size, and defect position information, the suitability of the pipeline segment with the defect is evaluated to make a defect maintenance decision, including repair level, planned repair time, and repair method; for the pipeline that does not have the excavation verification conditions, the surrounding environmental interference is investigated in detail, and after determining that the interference investigation is completed, the abnormal signal is further evaluated by using expert intervention, and if it is still severe anomaly, the maintenance decision level is set to immediate or planned repair within one year; the training process of the first pipeline anomaly level identification model includes: Determine a preset number of sample pipelines, and acquire a magnetic stress signal detection result of the preset number of sample pipelines; wherein the preset number of sample pipelines need to cover various defect positions and various anomaly levels; Determine defect positions and anomaly levels of the preset number of sample pipelines according to detection signal analysis by experts or excavation; Acquire the magnetic stress signal detection result of the sample pipeline whose anomaly level has been determined in advance as input of a model, acquire the defect positions and anomaly levels of the sample pipeline as output of the model, train and test a neural network model, and obtain a first pipeline anomaly level identification model; The training process of the second pipeline anomaly level identification model includes: Determine a preset number of sample pipelines, and acquire an ultrasonic guided wave signal detection result of the preset number of sample pipelines; wherein the preset number of sample pipelines need to cover various defect positions and various anomaly levels; Determine defect positions and anomaly levels of the preset number of sample pipelines according to detection signal analysis by experts or excavation; Acquire the ultrasonic guided wave signal detection result of the sample pipeline whose anomaly level has been determined in advance as input of a model, acquire the defect positions and anomaly levels of the sample pipeline as output of the model, train and test a neural network model, and obtain a second pipeline anomaly level identification model.

2. An apparatus for buried pipeline integrity comprehensive inspection and maintenance decision making, characterized in that, Comprise: The first determination module is used for determining the pipe diameter and the burial depth of the buried pipeline in the refinery, the gathering pipeline network or the station field; The second determination module is used for determining the signal detection method of the buried pipeline according to the pipe diameter and the burial depth; The detection module is used for performing signal detection on the buried pipeline according to the signal detection method to obtain a signal detection result; The identification module is used for inputting the signal detection result into a pipeline anomaly level identification model trained based on a deep learning algorithm to obtain a pipeline anomaly level identification result; The repair decision module is used for making a repair decision for the buried pipeline according to the pipeline anomaly level identification result; The second determination module is specifically used for: According to the pipe diameter and the burial depth, it is judged whether the magnetic stress detection condition and the ultrasonic guided wave detection condition are met, if both are met, it is determined that the signal detection method of the buried pipeline comprises magnetic stress detection and ultrasonic guided wave detection; If only the magnetic stress detection condition is met, it is determined that the signal detection method of the buried pipeline comprises magnetic stress detection; If only the ultrasonic guided wave detection condition is met, it is determined that the signal detection method of the buried pipeline comprises ultrasonic guided wave detection; Before the magnetic stress detection or the ultrasonic guided wave detection is performed on the pipeline, the method further comprises: Perform positioning detection on the pipeline, which can be accurate detection of pipeline positioning by a gyroscope or a ground detection device installed with a GPS probe to collect detection results; adopt an infrared leakage detector to perform infrared leakage screening on the buried pipeline to determine whether there is a leakage pipeline and collect detection results; The magnetic stress detection on the pipeline comprises: for the pipeline with magnetic stress detection, using a non-contact stress concentration detection device, segmentally detecting the pipeline on the ground along the pipeline direction, and collecting detection signal results and surrounding interference objects such as trees, power poles and metal objects; for the detection signal, checking whether the detection abnormal point is affected by the interference object, and removing the signal abnormality affected by the interference object; The signal detection result is input into a pipeline anomaly level recognition model trained based on a deep learning algorithm to obtain a pipeline anomaly level recognition result, which comprises: The magnetic stress signal detection result is input into a first pipeline anomaly level recognition model to obtain a defect position on the pipeline and an anomaly level recognition result; The first pipeline anomaly level recognition model is obtained by taking the magnetic stress signal detection result of a sample pipeline with a pre-determined pipeline anomaly level as the input of the model, taking the defect position on the sample pipeline and the corresponding anomaly level as the output of the model, and training and testing a neural network model; Or, The ultrasonic guided wave signal detection result is input into a second pipeline anomaly level recognition model to obtain a defect position on the pipeline and an anomaly level recognition result; the second pipeline anomaly level recognition model is obtained by taking the ultrasonic guided wave signal detection result of a sample pipeline with a pre-determined pipeline anomaly level as the input of the model, taking the defect position on the sample pipeline and the corresponding anomaly level as the output of the model, and training and testing a neural network model; Or, Both the above two detection results are input into the first and second pipeline anomaly level recognition models to obtain a defect position on the pipeline and an anomaly level recognition result, respectively, and the recognition results of the two detections are merged; Both the above two detection results are input into the first and second pipeline anomaly level recognition models to obtain a defect position on the pipeline and an anomaly level recognition result, respectively, and the recognition results of the two detections are merged, which comprises: The defect positions obtained according to the first pipeline anomaly level recognition model and the second pipeline anomaly level recognition model are processed in a union set to determine the defect position on the pipeline, and the anomaly levels obtained according to the first pipeline anomaly level recognition model and the second pipeline anomaly level recognition model are data fused according to the prediction accuracy weights of the two models to determine the corresponding anomaly level on the pipeline; Or, The defect positions obtained according to the first pipeline anomaly level recognition model and the second pipeline anomaly level recognition model are processed in a union set to determine the defect position on the pipeline, and the maximum value of the anomaly levels obtained according to the first pipeline anomaly level recognition model and the second pipeline anomaly level recognition model is taken as the corresponding anomaly level on the pipeline; According to the pipeline anomaly level recognition result, a maintenance decision is made for the buried pipeline, which comprises: If the pipeline anomaly level recognition result is a mild or moderate anomaly, the buried pipeline is marked as a pipeline for regular monitoring; If the pipeline anomaly level identification result is severe anomaly, the buried pipeline is investigated on site to determine whether it has excavation conditions. For the pipeline with excavation verification conditions, the point with severe abnormal detection signal is selected for excavation, and further detection is performed by using ultrasonic thickness measurement, laser scanning, TOFD crack detection and manual measurement means to further quantify the defect size and collect the detection results. For the defect after excavation, the defect size information is recorded as a sample for further correcting the model, and according to the defect type, defect size and defect position information, the suitability of the pipeline segment with defects is evaluated, and the defect repair decision is made, including repair level, planned repair time and repair method; For the pipeline without excavation verification conditions, the surrounding environmental interference is investigated in detail. After determining that the interference investigation is completed, the abnormal signal is further evaluated by expert intervention. If it is still severe anomaly, the repair decision level is set to immediate or planned repair within one year; The training process of the first pipeline anomaly level identification model comprises: A predetermined number of sample pipelines are determined, and the magnetic stress signal detection results of the predetermined number of sample pipelines are obtained; wherein the predetermined number of sample pipelines need to cover various defect positions and various anomaly levels; The defect positions and anomaly levels of the predetermined number of sample pipelines are determined by experts according to experience or excavation; The magnetic stress signal detection results of the sample pipelines with pre-determined pipeline anomaly levels are taken as the input of the model, and the defect positions and anomaly levels of the sample pipelines are taken as the output of the model. The neural network model is trained and tested to obtain the first pipeline anomaly level identification model; The training process of the second pipeline anomaly level identification model comprises: A predetermined number of sample pipelines are determined, and the ultrasonic guided wave signal detection results of the predetermined number of sample pipelines are obtained; wherein the predetermined number of sample pipelines need to cover various defect positions and various anomaly levels; The defect positions and anomaly levels of the predetermined number of sample pipelines are determined by experts according to experience or excavation; The ultrasonic guided wave signal detection results of the sample pipelines with pre-determined pipeline anomaly levels are taken as the input of the model, and the defect positions and anomaly levels of the sample pipelines are taken as the output of the model. The neural network model is trained and tested to obtain the second pipeline anomaly level identification model.

Images