Municipal pipeline detection method and detection system based on non-destructive detection and artificial intelligence

Abstract

The application belongs to the technical field of municipal pipeline detection, and provides a municipal pipeline detection method and detection system based on nondestructive detection and artificial intelligence, which comprises the following steps: AI reconstruction is performed on different pipe section diameter point cloud data to generate a continuous profile graph and identify a pipe diameter reduction pipe section; slope detection data are used to generate a slope curve to identify a slope change pipe section; the proportion of the overlapping mileage of the pipe diameter reduction and slope change pipe sections is calculated, the slope curve of the overlapping pipe section is linearly fitted, and the pipe diameter misjudgment risk pipe section is determined according to the determination coefficient; the coupling relationship between the slope change and the pipe diameter reduction is determined based on the historical data of the same type of pipe section, a coupling model is established, and whether the actual and theoretical pipe diameter reduction amount of the pipe diameter misjudgment risk pipe section conforms to the coupling relationship is determined by comparison; if not, the oblique section diameter is converted into the real internal diameter according to the slope inclination angle by combining the RANSAC algorithm, and whether the pipe diameter reduction occurs is determined by re-comparing the design internal diameter, so that the pipe diameter detection result is ensured to be accurate.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of municipal pipeline detection, and in particular to a municipal pipeline detection method and system based on non-destructive detection and artificial intelligence. BACKGROUND

[0002] At present, when obtaining the inner diameter information of the pipeline, the common pipeline detection means mostly relies on manual measurement or some basic detection equipment. Manual measurement is not only inefficient, but also difficult to ensure the accuracy and comprehensiveness of the measurement in complex and harsh pipeline environments. Some basic detection equipment can obtain part of the data, but it has poor adaptability in the face of complex and variable conditions of different pipe sections, and cannot realize continuous and accurate detection. For example, during the data acquisition process, when some detection methods relying on image acquisition are used, the collected image data may appear to be inclined due to the slope of the pipeline. This inclination will cause errors in the measurement of the pipe diameter, and may further cause misjudgment or deviation of the pipe diameter reduction. Like in the detection of some urban drainage pipelines with large slope fluctuations, the pipe section with normal inner diameter is incorrectly judged as pipe diameter reduction due to the inclination of the collected image, which brings great trouble to the subsequent pipeline maintenance decision.

[0003] At the same time, the traditional method is also difficult to accurately obtain the dynamic change characteristics of the pipeline slope. The change of the pipeline slope will directly affect the flow state of the fluid in the pipeline, and together with the pipe diameter condition, it will have a complex influence on the overall performance of the pipeline. If the internal relationship between the change of the pipeline slope and the pipe diameter reduction cannot be accurately grasped, the actual operating condition of the pipeline cannot be comprehensively and deeply understood.

[0004] The existing detection technology lacks a comprehensive and systematic method that can accurately and visually present the inner diameter state of the pipeline, capture the dynamic change characteristics of the slope, and effectively investigate and verify the possible misjudgment of the pipe diameter reduction. Therefore, it is of great practical significance to develop a new pipeline detection scheme that can solve the above problems and ensure the accuracy of the pipe diameter detection results.

[0005] Therefore, the present application provides a municipal pipeline detection method and system based on non-destructive detection and artificial intelligence. SUMMARY

[0006] In order to make up for the deficiencies of the prior art and solve at least one technical problem proposed in the background art.

[0007] The technical scheme adopted by the present application to solve its technical problems is: a municipal pipeline detection method based on non-destructive detection and artificial intelligence, comprising the following steps:

[0008] Step S10: Use point cloud data of pipe diameters of different sections of municipal pipelines to perform AI reconstruction of municipal pipelines, generate continuous cross-sectional views of the inner diameter of different pipe sections, and identify pipe sections with reduced diameters by comparing the designed inner diameter of the pipe sections.

[0009] Step S20: Using the IMU slope detection data of the municipal pipeline, generate slope curves for different pipe sections, and identify pipe sections with slope changes through the slope curves.

[0010] Step S30: Identify pipe sections at risk of misjudgment of pipe diameter by checking whether the pipe section with reduced diameter overlaps with the pipe section with changed slope and the linear gradient of the slope of the overlapping pipe section.

[0011] Step S40: By using historical slope change data and historical pipe diameter reduction data of the same type of pipe section, determine the coupling relationship between pipe section slope change and pipe diameter reduction, and combine the slope change and pipe diameter reduction of the pipe section with misjudged risk of pipe diameter to determine whether the pipe section slope change and pipe diameter reduction of the pipe section with misjudged risk of pipe diameter conform to the coupling relationship.

[0012] If the condition is not met in step S50, then the RANSAC algorithm is used to convert the diameter of the inclined section into the actual inner diameter based on the slope angle, and finally it is determined whether the pipe section at risk of misjudgment of pipe diameter has a reduced pipe diameter.

[0013] As a further technical solution of the present invention: the process of generating continuous cross-sectional views with different pipe section inner diameters is as follows:

[0014] The U-Net3+ deep learning model is used to reconstruct the inner wall of the pipe using point cloud data. The model input is voxelized point cloud data, and the output is a 3D mesh model of the inner wall of the pipe. Cross sections are cut at fixed intervals along the pipe axis, and the inner diameter data of each cross section is calculated to generate a continuous cross-sectional view with the axis length as the horizontal axis and the inner diameter as the vertical axis.

[0015] As a further technical solution of the present invention: the process of identifying pipe sections with reduced diameter is as follows:

[0016] Retrieve the design inner diameter data of the pipe section from the pipeline design drawings and compare it point by point with the actual inner diameter in the continuous cross-sectional drawings;

[0017] For any pipe segment, when the actual inner diameter of the segment is less than 90% of the design inner diameter in three or more consecutive cross sections, it is determined to be a pipe segment with reduced diameter. The starting mileage, ending mileage, and minimum actual inner diameter value are recorded, and the range of pipe diameter reduction mileage is obtained based on the starting mileage and ending mileage.

[0018] As a further technical solution of the present invention: the process of generating slope curves for different pipe sections and identifying pipe sections with varying slopes through the slope curves is as follows:

[0019] Based on any pipe segment, the pitch angle collected by the IMU is converted into a slope value, and a slope curve is generated with the pipeline axis mileage as the horizontal axis and the slope value as the vertical axis.

[0020] Based on the slope curve, if the slope change of a pipe segment in the slope curve exceeds the slope change threshold for 5 or more consecutive times, or if the difference between the maximum and minimum slope in the pipe segment exceeds the slope difference threshold, then the pipe segment is determined to be a slope change pipe segment, and the starting mileage, ending mileage, and slope change range are recorded. Based on the starting mileage and ending mileage, the slope change mileage range is obtained.

[0021] As a further technical solution of the present invention: the process of identifying pipe sections at risk of misjudgment of pipe diameter is as follows:

[0022] Obtain the overlapping mileage ratio. If the pipe section with reduced diameter and the pipe section with changed slope are the same pipe section and the overlapping mileage ratio meets the requirements, then mark the pipe section with reduced diameter as an overlapping pipe section.

[0023] Obtain the slope variation curve within the overlapping pipe section, fit it using the least squares method, and calculate the coefficient of determination of the fitted curve after fitting.

[0024] Overlapping pipe sections that meet the determination coefficient requirements are marked as pipe sections at risk of misjudgment of pipe diameter.

[0025] As a further technical solution of the present invention: the method for obtaining the overlapping mileage ratio is as follows:

[0026] Calculate the overlapping mileage between the mileage range of pipe diameter reduction and the mileage range of slope change, and calculate the ratio of the overlapping mileage to the total mileage of pipe diameter reduction to obtain the overlapping mileage ratio.

[0027] The total mileage for pipe diameter reduction is the difference between the maximum and minimum mileage within the pipe diameter reduction mileage range.

[0028] As a further technical solution of the present invention: the process of determining the coupling relationship between the change in pipe section slope and the reduction in pipe diameter is as follows:

[0029] Based on historical slope change data and historical pipe diameter reduction data, the slope change and pipe diameter reduction at different time points are obtained and integrated to obtain the slope change sequence and pipe diameter reduction sequence.

[0030] The Pearson correlation coefficient between the slope change sequence and the pipe diameter reduction sequence was calculated and then converted to absolute value to obtain the coupled correlation value.

[0031] If the coupling correlation value is greater than or equal to the coupling correlation threshold, it indicates that there is a linear coupling relationship between the change in pipe slope and the reduction in pipe diameter; otherwise, there is a non-linear coupling relationship.

[0032] As a further technical solution of the present invention: the process of determining whether the change in the slope of the pipe segment at risk of misjudgment of pipe diameter conforms to the coupling relationship with the reduction in pipe diameter is as follows:

[0033] If there is a linear coupling relationship, the least squares method is used to linearly fit the slope change sequence and the pipe diameter reduction sequence to obtain the coupling model between the pipe section slope and the pipe diameter reduction.

[0034] If there is a nonlinear coupling relationship, then the slope change sequence and the pipe diameter reduction sequence are nonlinearly fitted, and the fitting model with the highest goodness of fit is selected as the coupling model between the pipe slope and the pipe diameter reduction.

[0035] Obtain the minimum actual inner diameter value of the pipe section at risk of misjudgment, and perform absolute difference processing with the design inner diameter of the pipe section to obtain the current pipe diameter reduction. Based on the slope change range of the pipe section at risk of misjudgment, perform absolute difference processing with the maximum slope value and the minimum slope value within the slope change range to obtain the current slope change.

[0036] Substituting the current slope change of the pipe section at risk of misjudged diameter into the coupled model, we obtain the theoretical reduction in pipe diameter of the pipe section at risk of misjudged diameter.

[0037] Calculate the pipe diameter reduction deviation between the current pipe diameter reduction and the theoretical pipe diameter reduction for pipe sections at risk of misjudgment.

[0038] If the deviation in pipe diameter reduction is greater than the allowable error, then the slope change of the pipe section at risk of misjudgment of pipe diameter does not have a coupling relationship with the pipe diameter reduction; otherwise, it does.

[0039] As a further technical solution of the present invention: the process of finally determining whether the pipe section at risk of misjudgment of pipe diameter has experienced pipe diameter reduction is as follows:

[0040] Based on the risk of misjudged pipe diameter, the RANSAC algorithm is used to fit the angle between the pipe axis and the cross section, and the true inner diameter of each cross section is calculated according to the geometric relationship.

[0041] If there are still three or more consecutive cross sections that are smaller than 90% of the designed inner diameter, then the pipe diameter is considered to have been reduced; otherwise, it has not been reduced.

[0042] A municipal pipeline inspection system based on non-destructive testing and artificial intelligence, comprising:

[0043] The initial identification module for pipe diameter reduction uses point cloud data of pipe diameters of different sections of municipal pipelines to perform AI reconstruction of municipal pipelines, generate continuous cross-sectional views of the inner diameter of different pipe sections, and identify pipe diameter reduction sections by comparing with the designed inner diameter of the pipe sections.

[0044] The slope change pipe segment identification module generates slope curves for different pipe segments using IMU slope detection data of municipal pipelines, and identifies pipe segments with slope changes through the slope curves.

[0045] The pipe diameter misjudgment analysis module identifies pipe sections at risk of misjudgment by checking whether pipe sections with reduced diameters overlap with pipe sections with changing slopes and by assessing the linear gradient of the slope in overlapping pipe sections.

[0046] The pipe diameter-slope coupling analysis module determines the coupling relationship between pipe slope changes and pipe diameter reduction by using historical slope change data and historical pipe diameter reduction data of the same type of pipe section. It also combines the slope changes and pipe diameter reduction of pipe sections with misjudged risk of pipe diameter to determine whether the slope changes and pipe diameter reduction of pipe sections with misjudged risk of pipe diameter conform to the coupling relationship.

[0047] If the pipe diameter reduction determination module does not meet the requirements, it combines the RANSAC algorithm to convert the diameter of the inclined section into the actual inner diameter based on the slope inclination angle, and finally determines whether the pipe section at risk of misjudgment of pipe diameter has experienced pipe diameter reduction.

[0048] The beneficial effects of this invention are as follows: First, point cloud data of different pipe diameters are acquired using a dedicated pipeline laser scanning device. A U-Net3+ deep learning model is used for AI reconstruction to generate continuous cross-sectional images, which are then compared with the designed inner diameter to identify pipe sections with reduced diameter. Next, a high-precision IMU sensor mounted on a pipeline inspection robot is used to acquire slope detection data, generating slope curves to identify pipe sections with slope changes. Then, the proportion of overlapping mileage between pipe sections with reduced diameter and those with slope changes is calculated. The slope curves of the overlapping sections are linearly fitted, and the risk of misjudged pipe diameter is determined based on the coefficient of determination R². Subsequently, the coupling relationship between slope changes and pipe diameter reduction is determined based on historical data of similar pipe sections. A coupling model is established, and the actual and theoretical pipe diameter reduction amounts of the risk-of-misjudged pipe sections are compared to determine if the coupling relationship is met. Finally, if not, the RANSAC algorithm is used to convert the inclined section diameter into the true inner diameter based on the slope inclination angle, and the comparison is repeated with the designed inner diameter to determine if pipe diameter reduction has occurred. This solution can accurately visualize the pipe's inner diameter status, capture dynamic changes in slope, initially identify misjudgments of pipe diameter reduction, further verify the rationality of the pipe diameter reduction judgment, and ultimately correct measurement deviations to ensure accurate pipe diameter detection results. Attached Figure Description

[0049] The invention will now be further described with reference to the accompanying drawings.

[0050] Figure 1 This is a flowchart of the steps in Embodiment 1 of the present invention;

[0051] Figure 2 This is the logic judgment diagram of Embodiment 1 of the present invention;

[0052] Figure 3This is a system module diagram of Embodiment 2 of the present invention. Detailed Implementation

[0053] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0054] Example 1

[0055] Please see Figures 1-2 As shown in the figure, the municipal pipeline inspection method based on non-destructive testing and artificial intelligence described in this embodiment of the invention includes the following steps:

[0056] Step S10: Use point cloud data of pipe diameters of different sections of municipal pipelines to perform AI reconstruction of municipal pipelines, generate continuous cross-sectional views of the inner diameter of different pipe sections, and identify pipe sections with reduced diameters by comparing the designed inner diameter of the pipe sections.

[0057] In step S10, the point cloud data of the pipe diameter of different sections of the municipal pipeline is obtained in the following way:

[0058] The pipeline-specific laser scanning equipment (equipped with a laser ranging module, high-definition camera, etc.) moves along the pipeline axis to capture the three-dimensional coordinates (X,Y,Z) of the inner wall of the pipeline in real time and generate pipe diameter point cloud data.

[0059] In step S10, the process of performing AI reconstruction of municipal pipelines to generate continuous cross-sectional views of different pipe segment inner diameters is as follows:

[0060] The U-Net3+ deep learning model is used to reconstruct the inner wall of the pipe using point cloud data. The model input is voxelized point cloud data, and the output is a 3D mesh model of the inner wall of the pipe. A cross-section is cut at fixed intervals (e.g., 0.5m) along the pipe axis, and the inner diameter data of each cross-section is calculated to generate a continuous cross-sectional view with the axis length as the horizontal axis and the inner diameter as the vertical axis.

[0061] In step S10, the process of identifying pipe sections with reduced diameters by comparing them with the designed inner diameter of the pipe section is as follows:

[0062] Retrieve the design inner diameter data of the pipe section from the pipeline design drawings and compare it point by point with the actual inner diameter in the continuous cross-sectional drawings;

[0063] Based on any pipe segment, when the actual inner diameter of the pipe segment is less than 90% of the design inner diameter in three or more consecutive cross sections, it is determined to be a pipe segment with reduced diameter. The starting mileage, ending mileage, and minimum actual inner diameter value are recorded. The range of pipe diameter reduction mileage is obtained based on the starting mileage and ending mileage.

[0064] Understandably, the significance of step S10 is as follows: to acquire three-dimensional point cloud data of the inner wall through a pipeline-specific laser scanning device, to reconstruct the three-dimensional mesh model of the pipeline inner wall with the help of the U-Net3+ deep learning model and generate a continuous cross-sectional view, to accurately identify pipe sections with reduced diameter based on the designed inner diameter, to achieve accurate visualization of the pipeline inner diameter status and quantitative location of abnormal pipe sections, and to provide basic data support for subsequent detection and analysis of abnormal pipe diameter.

[0065] Step S20: Using the IMU slope detection data of the municipal pipeline, generate slope curves for different pipe sections, and identify pipe sections with slope changes through the slope curves.

[0066] In step S20, the IMU slope detection data includes the pitch angle of the pipe axis, and the IMU slope detection data is acquired in the following way:

[0067] A high-precision IMU sensor (accelerometer accuracy ≤0.1mg, gyroscope accuracy ≤0.01° / h) is mounted on the pipeline inspection robot, which moves at a constant speed along the pipeline axis to collect data. The sampling frequency is set to 100Hz. The collected raw data is zero-drift calibrated (using the static calibration method, the zero-drift value is obtained and discarded after being left to stand on a water platform for 30 minutes) and integrated to obtain the pitch angle of the pipeline axis.

[0068] In step S20, the process of generating slope curves for different pipe sections and identifying pipe sections with varying slopes using these slope curves is as follows:

[0069] Based on any pipe segment, the pitch angle collected by the IMU is converted into a slope value, where the slope = tan(pitch angle) × 100%, and a slope curve is generated with the pipeline axis mileage as the horizontal axis and the slope value as the vertical axis.

[0070] Based on the slope curve, if the slope change (absolute value of the slope difference between two adjacent mileage points) of a pipe segment in the slope curve exceeds the slope change threshold (e.g., 2‰) for 5 or more consecutive data points, or if the difference between the maximum and minimum slope in the pipe segment exceeds the slope difference threshold (e.g., 5‰), then the pipe segment is determined to be a slope change pipe segment, and the starting mileage, ending mileage, and slope change range are recorded. Based on the starting mileage and ending mileage, the slope change mileage range is obtained.

[0071] If the slope change (absolute value of the slope difference between two adjacent mileage points) of a pipe segment in the slope curve does not exceed the slope change threshold (e.g., 2‰) for 5 or more consecutive data points, and the difference between the maximum and minimum slope within the pipe segment does not exceed the slope difference threshold (e.g., 5‰), then the pipe segment is determined to be a non-slope change pipe segment.

[0072] It should be noted that the slope change threshold (e.g., 2‰) and slope difference threshold (e.g., 5‰) are determined according to the municipal pipeline design specifications.

[0073] Understandably, the significance of step S20 lies in: using a high-precision IMU sensor mounted on a pipeline inspection robot to collect pipeline axis pitch angle data, calibrating and calculating it to convert it into slope values ​​and generate slope curves, identifying pipe sections with slope changes and those without based on slope change thresholds and slope difference thresholds derived from municipal design specifications, accurately capturing the dynamic change characteristics of pipeline slope, and providing reliable slope anomaly data for subsequent analysis of the correlation between pipe diameter and slope.

[0074] Step S30: Identify pipe sections at risk of misjudgment of pipe diameter by checking whether the pipe section with reduced diameter overlaps with the pipe section with changed slope and the linear gradient of the slope of the overlapping pipe section.

[0075] In step S30, the process of identifying pipe sections at risk of misjudgment of pipe diameter is as follows:

[0076] Calculate the overlapping mileage between the mileage range of pipe diameter reduction and the mileage range of slope change, and calculate the ratio of the overlapping mileage to the total mileage of pipe diameter reduction to obtain the overlapping mileage ratio.

[0077] Among them, the total mileage of pipe diameter reduction is the difference between the maximum mileage and the minimum mileage within the pipe diameter reduction mileage range;

[0078] If the pipe section with reduced diameter and the pipe section with changed slope are the same pipe section and the overlapping mileage ratio is greater than or equal to the preset threshold, then the pipe section with reduced diameter is marked as an overlapping pipe section.

[0079] If the pipe section with reduced diameter and the pipe section with changed slope are not the same pipe section or the overlapping mileage ratio is less than a preset threshold (such as 70%), then the pipe section with reduced diameter will be marked as a non-overlapping pipe section.

[0080] Based on overlapping pipe sections, the slope variation curves within these sections are obtained and fitted using the least squares method. The coefficient of determination R0 of the fitted curve is then calculated. 2 ;

[0081] If the coefficient of determination R 2 If the value is greater than or equal to the determination coefficient threshold (e.g., 0.9), then the overlapping pipe section is marked as a pipe section at risk of misjudgment of pipe diameter;

[0082] If the coefficient of determination R 2 If the value is less than the determination coefficient threshold, the overlapping pipe section will be marked as a non-risk pipe section due to misjudgment of pipe diameter;

[0083] Understandably, if the pipe diameter reduction is a genuine design, the slope must be abrupt or a non-gradual change (for example, when the pipe diameter is reduced from DN500 to DN300, the slope may suddenly increase from 0.5% to 1.0% to match the flow rate, rather than changing slowly and linearly). If the "pipe diameter reduction" is due to measurement error (such as equipment calibration deviation or data entry error), then the actual slope of the pipe section should not have a design change. In this case, the extracted slope data is more likely to show a linear trend in random fluctuations (i.e., the slope itself is not actually adjusted, but only undergoes a small continuous change due to sampling error). Based on this, the linear fitting determination coefficient R² is introduced. The physical meaning of R² is: it represents the degree to which the fitted linear curve model interprets the slope data, and its value range is [0,1].

[0084] R²≥0.9: This means that more than 90% of the slope data can be explained by the linear curve model, indicating that the slope has a significant linear gradual trend. This is completely contradictory to the abrupt / discontinuous slope change that should occur when the actual pipe diameter changes. Therefore, it can be determined that the pipe diameter reduction is not the actual design, but a data misjudgment (such as measurement deviation or algorithm error), and should be marked as a pipe section with pipe diameter misjudgment risk.

[0085] Understandably, the significance of step S30 is as follows: by calculating the proportion of overlapping mileage between pipe sections with reduced diameter and pipe sections with changing slope, overlapping pipe sections are screened out. Then, the slope curves of the overlapping pipe sections are linearly fitted and the coefficient of determination R² is calculated. Based on whether R² reaches the threshold, it is determined whether the pipe section is a pipe section at risk of misjudgment of pipe diameter. By utilizing the contradiction between the linear gradual change characteristics of slope and the abrupt or discontinuous change characteristics of slope that should occur when the actual pipe diameter changes, the misjudgment of pipe diameter reduction caused by measurement or algorithm errors is initially investigated.

[0086] Step S40: By using historical slope change data and historical pipe diameter reduction data of the same type of pipe section, determine the coupling relationship between pipe section slope change and pipe diameter reduction, and combine the slope change and pipe diameter reduction of the pipe section with misjudged risk of pipe diameter to determine whether the pipe section slope change and pipe diameter reduction of the pipe section with misjudged risk of pipe diameter conform to the coupling relationship.

[0087] In step S40, the historical slope change data of the same type of pipe segment includes the slope change of the pipe segment at different time points within the historical service period (the period between the time of commissioning and the current time), and the historical pipe diameter reduction data includes the pipe diameter reduction of the pipe segment at different time points within the historical service period.

[0088] In step S40, the process of determining the coupling relationship between the change in pipe slope and the reduction in pipe diameter is as follows:

[0089] The slope changes at different time points are integrated into a slope change sequence, and the pipe diameter reduction at different time points is integrated into a pipe diameter reduction sequence.

[0090] It should be noted that the slope change sequence and the pipe diameter reduction sequence are the same in the time dimension and contain the same corresponding time points;

[0091] The Pearson correlation coefficient between the slope change sequence and the pipe diameter reduction sequence was calculated using the Pearson correlation coefficient method, and the coupled correlation value was obtained after absolute value conversion.

[0092] In some embodiments, the coupling association value is compared with the coupling association threshold;

[0093] If the coupling correlation value is greater than or equal to the coupling correlation threshold, it indicates that there is a linear coupling relationship between the change in pipe slope and the reduction in pipe diameter;

[0094] If the coupling correlation value is less than the coupling correlation threshold, it indicates that there is a non-linear coupling relationship between the change in pipe slope and the reduction in pipe diameter;

[0095] In step S40, the process of determining whether the slope change and diameter reduction of the pipe section at risk of misjudged pipe diameter conform to a coupling relationship, by combining the slope change of the pipe section at risk of misjudged pipe diameter and the pipe diameter reduction, is as follows:

[0096] If there is a linear coupling relationship between the change in pipe slope and the reduction in pipe diameter, then the least squares method is used to linearly fit the slope change sequence and the pipe diameter reduction sequence to obtain the coupling model between the pipe slope and the pipe diameter reduction. For example, the coupling model is: ΔD=k×ΔS+b (where ΔD is the pipe diameter reduction, ΔS is the slope change, k is the coupling coefficient, and b is a constant term). The values ​​of k and b are calculated by fitting using the least squares method.

[0097] If there is a nonlinear coupling relationship between the change in pipe slope and the reduction in pipe diameter, then nonlinear fitting is performed on the slope change sequence and the pipe diameter reduction sequence. Nonlinear fitting includes quadratic function fitting, exponential function fitting and logarithmic function fitting, etc., as shown in Table 1.

[0098] Table 1: Nonlinear Fitting Methods;

[0099]

[0100] After performing nonlinear fitting, the fitting model with the highest goodness of fit is selected as the coupling model between pipe slope and pipe diameter reduction.

[0101] Obtain the minimum actual inner diameter value of the pipe section at risk of misdiagnosis, and perform absolute difference processing with the design inner diameter of the pipe section to obtain the current pipe diameter reduction. Based on the range of slope variation in the risky pipe section determined by pipe diameter misjudgment, the absolute difference between the maximum and minimum slope values ​​within the range of slope variation is calculated to obtain the current slope variation. ;

[0102] The current slope change of the pipe section at risk of misjudging pipe diameter. Substituting into the coupling model, we obtain the theoretical pipe diameter reduction for pipe sections at risk of misjudgment. ;

[0103] The current pipe diameter reduction of the pipe section at risk of misjudgment. Compared with the theoretical pipe diameter reduction The diameter reduction deviation between them, i.e.:

[0104] If the pipe diameter reduction deviation is less than or equal to the allowable error (5% of the pipe diameter design inner diameter), then the slope change of the pipe section at risk of pipe diameter misjudgment is considered to be coupled with the pipe diameter reduction.

[0105] If the deviation of pipe diameter reduction is greater than the allowable error, the slope change of the pipe section at risk of pipe diameter misjudgment does not have a coupling relationship with the pipe diameter reduction.

[0106] Understandably, the significance of step S40 is as follows: based on historical data of similar pipe sections, the coupling relationship (linear or nonlinear) between slope change and pipe diameter reduction is determined by the Pearson correlation coefficient, and a coupling model is established. The current slope change of the pipe section at risk of misjudgment of pipe diameter is substituted into the model to obtain the theoretical pipe diameter reduction. This is compared with the actual pipe diameter reduction. Whether the deviation is within the allowable range is used to determine whether the two conform to the coupling relationship, further verifying the rationality of the pipe diameter reduction judgment and providing a basis for the final determination of whether there is a misjudgment.

[0107] If step S50 does not meet the requirements, then combine the RANSAC algorithm to convert the diameter of the inclined section into the actual inner diameter according to the slope inclination angle, and finally determine whether the pipe section with the risk of misjudgment of pipe diameter has a reduced pipe diameter.

[0108] In step S50, using the RANSAC algorithm, the diameter of the inclined section is converted into the actual inner diameter based on the slope angle. The final process for determining whether the pipe section at risk of misjudgment of pipe diameter has experienced pipe diameter reduction is as follows:

[0109] Based on the risk of misjudged pipe diameter, the RANSAC algorithm is used to fit the angle θ (slope inclination angle) between the pipe axis and the cross section, and the true inner diameter D of each cross section is calculated according to the geometric relationship: D = D_slope × cosθ (D_slope is the diameter of the inclined section reconstructed from the point cloud).

[0110] The actual inner diameter of the pipe section at risk of misjudgment is re-compared with the design inner diameter;

[0111] If there are still three or more consecutive cross sections that are less than 90% of the design inner diameter, then the pipe diameter is determined to have been reduced. Conversely, if there are no three or more consecutive cross sections that are less than 90% of the design inner diameter, then the pipe diameter is determined not to have been reduced.

[0112] Understandably, the significance of step S50 is as follows: when the slope change of the pipe section with the risk of misjudged pipe diameter does not have a coupling relationship with the pipe diameter reduction, the RANSAC algorithm is used to fit the angle between the pipe axis and the cross section, the diameter of the inclined section is converted into the true inner diameter, and then compared with the design inner diameter to finally determine whether the pipe section really has a pipe diameter reduction, correct the measurement deviation caused by the inclination of the cross section, and ensure the accuracy of the pipe diameter detection results.

[0113] Example 2

[0114] Based on the same inventive concept as the municipal pipeline inspection method based on non-destructive testing and artificial intelligence in the foregoing embodiments, such as Figure 3 As shown, this application provides a municipal pipeline inspection system based on non-destructive testing and artificial intelligence, wherein the system specifically includes:

[0115] The initial identification module for pipe diameter reduction uses point cloud data of pipe diameters of different sections of municipal pipelines to perform AI reconstruction of municipal pipelines, generate continuous cross-sectional views of the inner diameter of different pipe sections, and identify pipe diameter reduction sections by comparing with the designed inner diameter of the pipe sections.

[0116] The slope change pipe segment identification module generates slope curves for different pipe segments using IMU slope detection data of municipal pipelines, and identifies pipe segments with slope changes through the slope curves.

[0117] The pipe diameter misjudgment analysis module identifies pipe sections at risk of misjudgment by checking whether pipe sections with reduced diameters overlap with pipe sections with changing slopes and by assessing the linear gradient of the slope in overlapping pipe sections.

[0118] The pipe diameter-slope coupling analysis module determines the coupling relationship between pipe slope changes and pipe diameter reduction by using historical slope change data and historical pipe diameter reduction data of the same type of pipe section. It also combines the slope changes and pipe diameter reduction of pipe sections with misjudged risk of pipe diameter to determine whether the slope changes and pipe diameter reduction of pipe sections with misjudged risk of pipe diameter conform to the coupling relationship.

[0119] If the pipe diameter reduction determination module does not meet the requirements, it combines the RANSAC algorithm to convert the diameter of the inclined section into the actual inner diameter based on the slope inclination angle, and finally determines whether the pipe section at risk of misjudgment of pipe diameter has experienced pipe diameter reduction.

[0120] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for municipal pipe inspection based on non-destructive probing and artificial intelligence, characterized in that: The method comprises the following steps: Step S10, AI reconstruction of the municipal pipeline is performed through pipe diameter point cloud data of different pipe sections of the municipal pipeline, a continuous profile graph of the inner diameters of different pipe sections is generated, and a pipe diameter reduction pipe section is identified by comparing the design inner diameters of the pipe sections; Step S20, a slope curve of different pipe sections is generated through IMU slope detection data of the municipal pipeline, and a slope change pipe section is identified through the slope curve; Step S30, whether the pipe diameter reduction pipe section and the slope change pipe section coincide and the linear gradient of the slope of the coinciding pipe section are determined to identify a pipe diameter misjudgment risk pipe section; Step S40, a coupling relationship between the slope change and the pipe diameter reduction of the pipe section is determined through historical slope change data and historical pipe diameter reduction data of the same type of pipe section, and whether the slope change and the pipe diameter reduction of the pipe diameter misjudgment risk pipe section meet the coupling relationship is determined in combination with the slope change and the pipe diameter reduction of the pipe diameter misjudgment risk pipe section; Step S50, if not, the RANSAC algorithm is combined, the inclined section diameter is converted into the real inner diameter according to the slope inclination angle, and whether the pipe diameter misjudgment risk pipe section has the pipe diameter reduction is finally determined.

2. The method for municipal pipeline inspection based on non-destructive detection and artificial intelligence according to claim 1, characterized in that: The process of generating the continuous profile graph of the inner diameters of different pipe sections is: A U-Net3+ deep learning model is used to reconstruct the inner wall of the pipe diameter point cloud data, the model input is point cloud voxel data, and the output is a three-dimensional grid model of the inner wall of the pipe; along the axis direction of the pipe, cross sections are intercepted at fixed intervals, the inner diameter data of each cross section is calculated, and a continuous profile graph with the axis length as the horizontal axis and the inner diameter as the vertical axis is generated.

3. The method for municipal pipeline inspection based on non-destructive detection and artificial intelligence according to claim 2, characterized in that: The process of identifying the pipe diameter reduction pipe section is: The pipe section design inner diameter data in the pipe design drawing is called, and the actual inner diameter in the continuous profile graph is compared point by point; Based on any pipe section, when the actual inner diameter of the pipe section is less than 90% of the design inner diameter for 3 or more cross sections in succession, the pipe diameter reduction pipe section is determined, and the starting mileage, the ending mileage and the minimum actual inner diameter value are recorded, and the pipe diameter reduction mileage range is obtained according to the starting mileage and the ending mileage.

4. The method for municipal pipeline inspection based on non-destructive detection and artificial intelligence according to claim 1, characterized in that: The process of generating the slope curve of different pipe sections and identifying the slope change pipe section through the slope curve is: Based on any pipe section, the pitch angle collected by the IMU is converted into a slope value, and a slope curve is generated with the pipe axis mileage as the horizontal axis and the slope value as the vertical axis; Based on the slope curve, if the slope change amount of the pipe section in the slope curve exceeds the slope change threshold value for 5 or more in succession, or the difference between the maximum slope and the minimum slope in the pipe section exceeds the slope difference threshold value, the pipe section is determined as a slope change pipe section, and the starting mileage, the ending mileage and the slope change range are recorded, and the slope change mileage range is obtained according to the starting mileage and the ending mileage.

5. The method for municipal pipeline inspection based on non-destructive detection and artificial intelligence according to claim 4, characterized in that: The process of identifying the pipe diameter misjudgment risk pipe section is: The coincidence mileage proportion is obtained, if the pipe diameter reduction pipe section and the slope change pipe section are the same pipe section and the coincidence mileage proportion meets the requirements, the pipe diameter reduction pipe section is marked as a coinciding pipe section; The slope change curve in the coinciding pipe section is obtained, and the least square method is used for fitting, and the determination coefficient of the fitted curve is calculated after fitting; The coinciding pipe section meeting the determination coefficient requirement is marked as a pipe diameter misjudgment risk pipe section.

6. The method for municipal pipeline inspection based on non-destructive detection and artificial intelligence according to claim 5, characterized in that: The acquisition method of the coincidence mileage proportion is: Calculate the overlapping mileage between the pipe diameter reduction mileage range and the slope change mileage range, and calculate the proportion of the overlapping mileage with the total pipe diameter reduction mileage to obtain the proportion of the overlapping mileage. The total pipe diameter reduction mileage is the difference between the maximum mileage and the minimum mileage in the pipe diameter reduction mileage range.

7. The method for municipal pipeline inspection based on non-destructive detection and artificial intelligence according to claim 6, characterized in that: The process of determining the coupling relationship between the slope change and the pipe diameter reduction of the pipe section is: According to historical slope change data and historical pipe diameter reduction data, obtain the slope change amount and the pipe diameter reduction amount at different time points and integrate them respectively to obtain a slope change sequence and a pipe diameter reduction sequence; Calculate the Pearson correlation coefficient between the slope change sequence and the pipe diameter reduction sequence, and obtain the coupling correlation value after absolute value processing; If the coupling correlation value is greater than or equal to the coupling correlation threshold, it indicates that the pipe section slope change and the pipe diameter reduction have a linear coupling relationship, otherwise, they have a nonlinear coupling relationship.

8. The method for municipal pipeline inspection based on non-destructive detection and artificial intelligence according to claim 7, characterized in that: The process of determining whether the pipe section slope change and the pipe diameter reduction of the pipe diameter misjudgment risk section conform to the coupling relationship is: If there is a linear coupling relationship, linear fitting is performed on the slope change sequence and the pipe diameter reduction sequence by the least squares method to obtain a coupling model between the pipe section slope and the pipe diameter reduction; If there is a nonlinear coupling relationship, nonlinear fitting is performed on the slope change sequence and the pipe diameter reduction sequence, and the fitting model with the highest goodness of fit is selected as the coupling model between the pipe section slope and the pipe diameter reduction; Obtain the minimum actual inner diameter value of the pipe diameter misjudgment risk section, and perform absolute difference processing with the design inner diameter of the pipe section to obtain the current pipe diameter reduction amount. According to the slope change range of the pipe diameter misjudgment risk section, perform absolute difference processing on the maximum slope value and the minimum slope value in the slope change range to obtain the current slope change amount; Substitute the current slope change amount of the pipe diameter misjudgment risk section into the coupling model to obtain the theoretical pipe diameter reduction amount of the pipe diameter misjudgment risk section; Calculate the pipe diameter reduction deviation between the current pipe diameter reduction amount and the theoretical pipe diameter reduction amount of the pipe diameter misjudgment risk section; If the pipe diameter reduction deviation is greater than the allowable error, it is determined that the slope change and the pipe diameter reduction of the pipe diameter misjudgment risk section do not conform to the coupling relationship, otherwise, they conform.

9. The method for municipal pipeline inspection based on non-destructive detection and artificial intelligence according to claim 1, characterized in that: The process of finally determining whether the pipe diameter misjudgment risk section has pipe diameter reduction is: Based on the pipe diameter misjudgment risk section, fit the angle between the pipeline axis and the cross section by the RANSAC algorithm, and calculate the true inner diameter of each cross section according to the geometric relationship; If there are still three or more consecutive cross sections that are less than 90% of the design inner diameter, it is determined that pipe diameter reduction occurs, otherwise, it does not occur.

10. A municipal pipeline inspection system based on non-destructive probing and artificial intelligence, characterized in that, The system is used to execute the method of any one of claims 1-9, and the system comprises: A pipe diameter reduction pipe section initial identification module, which performs AI reconstruction of the municipal pipeline by using pipe diameter point cloud data of different pipe sections of the municipal pipeline, generates continuous profile graphs of the inner diameters of different pipe sections, and identifies pipe diameter reduction pipe sections by comparing the design inner diameters of the pipe sections; A slope change pipe section identification module, which generates slope curves of different pipe sections by using IMU slope detection data of the municipal pipeline, and identifies slope change pipe sections by using the slope curves. The pipe diameter misjudgment analysis module identifies the pipe diameter misjudgment risk pipe section by whether the pipe diameter reduction pipe section and the slope change pipe section coincide and the linear gradual change of the slope of the coinciding pipe section; The pipe diameter-slope coupling analysis module determines the coupling relationship between the slope change of the pipe section and the pipe diameter reduction through the historical slope change data and the historical pipe diameter reduction data of the same type of pipe section, and judges whether the slope change of the pipe section and the pipe diameter reduction of the pipe diameter misjudgment risk pipe section conform to the coupling relationship in combination with the slope change and the pipe diameter reduction of the pipe diameter misjudgment risk pipe section; The pipe diameter reduction determination module, if not, converts the oblique section diameter into the real inner diameter according to the slope inclination angle in combination with the RANSAC algorithm, and finally determines whether the pipe diameter misjudgment risk pipe section appears the pipe diameter reduction.

Images