A method and system for detecting underground pipelines based on multiple information sources

Abstract

The application discloses a kind of underground pipeline detection method and system based on multiple information sources, which comprises the following steps: pipeline detection system obtains the historical information of regional pipeline, arranges pipeline detection sensor;Establish detection sub-region, obtain pipeline characteristic parameter;Design regional fault detection algorithm, obtain health center point, calculate regional fault coefficient, determine fault sub-region;Pipe robot is put into fault sub-region, design adaptive fault point detection algorithm, detect fault point, record fault point position;Output pipeline detection result, notify pipeline inspection personnel;The scheme realizes the fault detection of underground pipeline by using multiple information sources, improves the detection efficiency by using the historical information of pipeline and establishing the arrangement scheme of pipeline detection sensor, divides the entire pipeline by establishing sub-region, detects each sub-region separately, improves the detection efficiency and accuracy of pipeline fault, and reduces the safety risk caused by pipeline fault.

Description

Technical Field

[0001] This invention belongs to the field of pipeline inspection technology, and in particular relates to a method and system for underground pipeline inspection based on multiple information sources. Background Technology

[0002] As the "lifeline" of urban infrastructure, underground pipelines bear core functions such as water supply, gas transmission, and sewage discharge. Their health directly affects the safety of urban operations and the quality of life of residents. With the acceleration of urbanization, the scale of underground pipeline networks continues to expand, and the service life of pipelines increases year by year. Factors such as corrosion, settlement, and third-party damage lead to frequent failures such as cracks, leaks, and blockages in underground pipelines, seriously threatening social and economic security and people's lives.

[0003] Current mainstream underground pipeline inspection technologies have significant limitations: on the one hand, most inspection schemes rely on a single information source, resulting in low accuracy and error tolerance, and are prone to false positives; on the other hand, existing technologies lack correlation analysis of historical pipeline fault data. This "isolated inspection" model cannot optimize sensor deployment by incorporating the regional distribution patterns of pipeline faults, leading to either excessively high or low sensor density, thus increasing inspection costs. Furthermore, existing inspection technologies often only determine the general fault section, failing to quickly pinpoint specific fault sub-regions, requiring substantial manpower and resources for thorough investigation, resulting in low inspection efficiency. Therefore, there is an urgent need to propose an underground pipeline inspection method utilizing multiple information sources. Summary of the Invention

[0004] This invention provides a method and system for detecting underground pipelines based on multiple information sources, in order to solve the technical problems of current underground pipeline detection methods that rely on a single information source, have low detection accuracy, and cannot promptly identify underground pipeline fault points.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] On one hand, the present invention provides a method for detecting underground pipelines based on multiple information sources, which includes:

[0007] S1: The pipeline inspection system acquires historical information about regional pipelines, creates a fault distribution map, calculates the distribution coefficient, and deploys pipeline inspection sensors based on the distribution coefficient.

[0008] S2: Based on the number and distribution of pipeline detection sensors, establish detection sub-regions, obtain the initial detection parameters of pipeline detection sensors in each detection sub-region, perform feature processing on the initial detection parameters, and obtain pipeline feature parameters.

[0009] S3: Establish feature coordinates from pipeline feature parameters, design regional fault detection algorithm, obtain health center point, calculate regional fault coefficient from health center point and feature coordinates, and determine fault sub-region;

[0010] S4: Deploy pipeline robots to the faulty sub-region. The pipeline robots are equipped with industrial cameras to collect pipeline images. Design an adaptive fault point detection algorithm to detect fault points and record their locations.

[0011] S5: Output pipeline inspection results and notify pipeline inspection personnel.

[0012] Further, step S1 includes:

[0013] The pipeline inspection system acquires historical information about regional pipelines. Regional pipelines refer to pipelines within a continuous area, and these pipelines have single-layer continuity.

[0014] Historical information on regional pipelines includes the year the pipeline was manufactured, the actual location of pipeline failures and the number of actual failure locations since the year the pipeline was manufactured.

[0015] A fault distribution map is created based on the fault locations. The fault distribution map includes the overall pipeline route within the area and the actual fault locations.

[0016] Set distribution length Set the start and end points of the pipeline, starting from the start point and at intervals of... For each meter, a virtual fault point is set. The virtual fault point does not participate in the calculation of the distribution coefficient of the real fault point.

[0017] Based on the pipeline's production year and various failure points Calculate the distribution coefficient of each fault point within the sub-interval. Wherein, the sub-interval is a left-closed, right-open interval, and the distribution coefficient of the i-th fault point is... The specific calculation formula is as follows:

[0018]

[0019] in, Indicates the current year. This represents the production year of the pipeline at the i-th fault point. For virtual fault points, this value is the average production year of all pipelines. This represents the number of actual fault points surrounding the i-th fault point;

[0020] Extract the maximum value of the distribution coefficient within each subinterval, where the maximum value of the distribution coefficient within the i-th subinterval is... Initialize the evaluation coefficients The distribution coefficient and the number of pipeline detection sensors within the i-th sub-interval The relationship is shown in the following formula:

[0021]

[0022]

[0023]

[0024] When the value is zero, the pipeline detection sensor is positioned at the center of the i-th sub-interval. When the value is not zero, the pipeline detection sensors are evenly distributed within the i-th sub-interval.

[0025] Further, step S2 includes: establishing detection sub-regions based on the number and distribution of pipeline detection sensors, with each detection sub-region and sub-interval corresponding to one another;

[0026] The initial detection parameters of the pipeline detection sensors in each detection sub-region are obtained, including pipeline pressure and flow rate. The initial detection parameters are then characterized to obtain the pipeline characteristic pressure and pipeline characteristic flow rate.

[0027] For pipeline pressure characterization, the pipeline pressure set of the i-th detection sub-region is: ,in , This represents the pipe pressure of the j-th pipe detection sensor in the i-th detection sub-region, where j represents the serial number and the characteristic pipe pressure of the i-th detection sub-region. The calculation formula is as follows:

[0028]

[0029] in, This represents the pipe pressure of the (j-1)th pipe detection sensor in the ith detection sub-region;

[0030] For pipeline flow characteristic processing, the pipeline flow set of the i-th detection sub-region is: ,in , This represents the pipe flow rate of the j-th pipe detection sensor in the i-th detection sub-region, and the characteristic pipe flow rate of the i-th detection sub-region. The calculation formula is as follows:

[0031]

[0032] in, This represents the pipe flow rate of the (j+1)th pipe detection sensor in the i-th detection sub-region;

[0033] The characteristic pressure and characteristic flow rate of the pipeline after characterization processing constitute the pipeline characteristic parameters. Further, step S3 includes:

[0034] The feature coordinates are established based on the pipe feature parameters. The feature coordinates of the i-th detection sub-region are ( , Design a regional fault detection algorithm and use the sparrow algorithm to determine the penalty coefficient in the regional fault detection algorithm. The detection sub-regions for anomalies are detected using regression analysis.

[0035] Let the plane equation of the classification line be:

[0036]

[0037] in, These are the weighting coefficients, and z is the bias value. ;

[0038] Introducing slack variables The following equation is satisfied:

[0039]

[0040] in, It is a sample classification function, and the detection result includes either a faulty or healthy result;

[0041] The objective function f of the regional fault detection algorithm is shown in the following equation:

[0042]

[0043] in, It is the number of sub-regions to be detected. Describes the Euclidean norm. Indicates the first There are slack variables, where g represents the ordinal number;

[0044] The underground pipeline inspection problem is transformed into a convex quadratic programming optimization problem. A regional fault detection algorithm is used to determine whether each detection sub-region is faulty or healthy. Training and test sets are constructed from empirical data of underground pipelines, with a data ratio of 7:3. Feature coordinates of healthy and faulty pipelines are randomly selected as training samples, and the remaining sample data is used as test samples. A penalty coefficient is determined. ;

[0045] After training, a health center point is obtained. The regional fault coefficient is calculated from the health center point and the feature coordinates. The regional fault coefficient is the Euclidean distance W from the feature coordinates of each detection sub-region to the health center point.

[0046] Initialize diagnostic thresholds If the regional failure coefficient W is greater than If the detected sub-region is faulty, then the fault coefficient W of the region is less than or equal to 1. If the result is positive, then the detected sub-region is considered healthy.

[0047] Further, step S4 includes:

[0048] A coordinate system is established with the starting position of the fault sub-region as the origin. A pipeline robot is deployed to the starting position of the fault sub-region. The pipeline robot is equipped with an industrial camera and an inertial measurement unit. The industrial camera acquires pipeline images, where the resolution of the pipeline images is [resolution missing]. ,in All values ​​are integers, the sampling frequency is V, and each pixel includes a temperature value and a grayscale value.

[0049] The inertial measurement unit records the real-time position of the pipeline robot, where the real-time position of the pipeline robot is the position of the center point of each pipeline image;

[0050] The design of an adaptive fault point detection algorithm to detect the location of fault points involves the following steps:

[0051] (1) Calculate the single-frame deviation coefficient:

[0052] The pipeline image is segmented to form There are several sub-images, where m represents an integer. Calculate the maximum grayscale value of each sub-image. The single-frame deviation coefficient is calculated from the maximum grayscale value of the sub-image. This is the single-frame deviation coefficient of the i-th pipeline image. The calculation formula is as follows:

[0053]

[0054] Where k represents the sequence number, which is an integer. This represents the maximum grayscale value of the k-th sub-image of the i-th pipeline image. This represents the average of the maximum grayscale values ​​in the i-th pipeline image;

[0055] (2) Locate pipeline images with fault points and set a single-frame deviation threshold. If the single-frame deviation coefficient of the i-th pipeline image is less than the single-frame deviation threshold, then there are no suspicious fault points in the image and it is marked as safe; if the single-frame deviation coefficient of the i-th pipeline image is greater than or equal to the single-frame deviation threshold, then there are suspicious fault points in the image and it is marked as fault.

[0056] (3) Obtain a set of stable fault point images, count the number J of pipe images marked as faulty, reorder the pipe images marked as faulty, calculate the screening coefficient of each sub-image in the faulty pipe image, and the screening coefficient of the j-th sub-image of the i-th pipe image. The calculation formula is as follows:

[0057]

[0058] in, This represents the number of pixels surrounding the pixel with the maximum grayscale value in the j-th sub-image of the i-th pipeline image. This represents the temperature value of the pixel with the highest grayscale value in the j-th sub-image of the i-th pipeline image. Let represent the temperature value of the k-th pixel surrounding the pixel with the maximum grayscale value in the j-th sub-image of the i-th pipeline image, and let e represent the exponential function;

[0059] The selection coefficient for each pipeline image is the average of the selection coefficients for all sub-images;

[0060] Initialize the filtering threshold A. Calculate the adaptive filtering threshold based on the number of pipeline images marked as faulty, as shown in the following expression:

[0061]

[0062]

[0063] in, and Indicates the filter value;

[0064] If the screening coefficient of the i-th pipeline image is greater than the adaptive screening threshold, the image is marked as a safe and stable fault point image. If the screening coefficient of the i-th pipeline image is less than or equal to the adaptive screening threshold, the pipeline image is removed to obtain a set of stable fault point images.

[0065] (4) Calculate the location of the fault point. For each stable fault point image, calculate the location of each stable fault point image. The location of the fault point is the average of the locations of each stable fault point image.

[0066] On the other hand, the present invention also provides an underground pipeline detection system based on multiple information sources, which includes:

[0067] The sensor distribution planning module allows the pipeline detection system to acquire historical information about pipelines in the area, create a fault distribution map, calculate the distribution coefficient, and deploy pipeline detection sensors based on the distribution coefficient.

[0068] The region determination module establishes detection sub-regions based on the number and distribution of pipeline detection sensors, obtains the initial detection parameters of the detection sensors in each detection sub-region, performs feature processing on the initial detection parameters to obtain pipeline feature parameters, establishes feature coordinates from the pipeline feature parameters, designs a region fault detection algorithm to obtain the health center point, calculates the region fault coefficient from the health center point and feature coordinates, and determines the fault sub-region.

[0069] The module calculates the location of the fault point, deploys a pipeline robot to the fault sub-region, the pipeline robot is equipped with an industrial camera to collect pipeline images, an adaptive fault point detection algorithm is designed to detect the fault point, and the location of the fault point is recorded.

[0070] The results output module outputs the pipeline inspection results and notifies the pipeline inspection personnel.

[0071] The beneficial effects of the technical solution provided by this invention include at least the following:

[0072] 1. This invention completely solves the problem of blind sensor deployment in traditional detection by linking "historical information - fault distribution - sensor deployment," achieving precise allocation of detection resources. In the early stages of detection, this invention first acquires historical information about the regional pipeline, then creates a fault distribution map based on this data, and finally calculates the distribution coefficient to ensure no blind spots in high-fault areas. For low-fault areas, the number of sensors is appropriately reduced to avoid resource waste. This differentiated deployment method based on the distribution coefficient can reduce the number of sensors by 30%-40% compared to the traditional "uniform deployment" scheme, significantly reducing the procurement and installation costs of detection equipment. Simultaneously, because the sensors accurately cover high-fault areas, they can quickly detect parameter anomalies in those areas, shortening the initial fault detection response time by more than 50% compared to traditional solutions. This effectively improves the coverage efficiency and response speed of pipeline detection, making it particularly suitable for the detection of large-scale urban pipeline networks. Furthermore, by establishing detection sub-regions, this invention breaks down a large-scale pipeline network into several independent small regions. Sensor data within each sub-region is collected independently and analyzed centrally, avoiding data redundancy and interference problems in the traditional "overall collection-overall analysis" mode. It can also quickly locate sub-regions with abnormal data, narrowing the scope for subsequent fault diagnosis and further improving detection efficiency.

[0073] 2. This invention achieves precise location of faulty sub-regions by establishing feature coordinates, solving the pain point of traditional detection methods that "only know the faulty segment, but not the specific area". During the detection process, this invention first characterizes the initial detection parameters of each detection sub-region; then, based on these parameters, it establishes feature coordinates and combines pipeline information from network big data to design a regional fault detection algorithm to determine the health center point; finally, it calculates the Euclidean distance between the feature coordinates of each detection sub-region and the health center point to obtain the regional fault coefficient. Compared with traditional "human experience judgment", this quantitative calculation method improves the location accuracy of faulty sub-regions from "1-2km faulty segment" to "200-500m sub-region", reducing the location error by more than 70%.

[0074] 3. After identifying the faulty sub-region, this invention only requires deploying a pipeline robot to that sub-region, eliminating the need for a comprehensive inspection of the entire faulty section, significantly reducing the robot's operating range and time. For example, if a 10km-long water supply pipeline experiences pressure anomalies, the traditional approach requires deploying a robot to inspect the entire 10km pipeline, taking approximately 5 hours. However, this invention, through regional fault coefficient calculation, only requires deploying a robot to a 2km-long faulty sub-region, reducing inspection time to less than 1 hour and shortening the fault investigation cycle by 80%. Simultaneously, the industrial camera mounted on the pipeline robot can focus on the inner wall of the pipeline within the faulty sub-region, avoiding the cumbersome process of traditional "full-process shooting - post-process frame-by-frame analysis," further improving fault investigation efficiency, saving time for rapid fault repair, and effectively reducing the risk of safety accidents caused by pipeline faults. Furthermore, this invention, through the design of an adaptive fault point detection algorithm, addresses the problem of "fixed thresholds easily leading to misjudgments" in traditional image detection algorithms, achieving intelligent and accurate identification of pipeline fault points, significantly improving the reliability of detection results, and further reducing the waste of computing power during fault point detection. This represents a 21 percentage point improvement compared to the traditional fixed threshold algorithm (accuracy approximately 75%). This high-accuracy and high-precision fault point identification capability effectively avoids the repeated construction caused by "missed repairs" and "incorrect repairs" in traditional detection, reduces pipeline operation and maintenance costs, and ensures the long-term reliable operation of underground pipelines, which is of great significance for improving the operation and maintenance level of urban infrastructure. Attached Figure Description

[0075] Figure 1 This is a schematic diagram of the overall execution flow of an underground pipeline detection method based on multiple information sources, provided in an embodiment of the present invention. Detailed Implementation

[0076] The present invention will be further described below with reference to the accompanying drawings, but this is not intended to limit the present invention in any way. Any modifications or substitutions made based on the teachings of the present invention shall fall within the protection scope of the present invention.

[0077] Example 1

[0078] This embodiment provides a method for detecting underground pipelines based on multiple information sources, which can be implemented by electronic devices, such as... Figure 1 As shown. Specifically, the method in this embodiment includes the following steps:

[0079] The pipeline inspection system acquires historical information about pipelines in the area, creates a fault distribution map, calculates the distribution coefficient, and deploys pipeline inspection sensors based on the distribution coefficient. Specifically:

[0080] The pipeline inspection system acquires historical information about regional pipelines. It should be further clarified that the pipeline inspection system is the underground pipeline inspection system mentioned in this solution. Regional pipelines refer to pipelines within a certain continuous area. The pipelines have single-layer continuity. Due to the size of the pipelines, the distinction between intervals and regions is not made here. However, this solution is only for pipelines with single-layer continuity. Although it may also be applicable to complex pipelines, the effect will be limited.

[0081] Historical information on regional pipelines includes the year the pipeline was manufactured, the actual location of pipeline failures and the number of actual failure locations since the year the pipeline was manufactured.

[0082] It should be further explained that the production year of a pipeline is positively correlated with its failure rate. In particular, once a pipeline reaches its design lifespan, its replacement capacity and conditions become limited, and the failure rate will increase significantly. By utilizing the existing information of the pipeline itself, more detection sensors can be installed in high-failure areas to detect faults in a timely manner and reduce the potential risks caused by pipeline failures.

[0083] In addition, this solution is mainly for water supply pipelines, but it can also be used as a reference for other pipelines.

[0084] A fault distribution map is created based on the fault locations. The fault distribution map includes the overall pipeline route and the actual fault locations within the area. In this scheme, the purpose of the fault distribution map is to unify the information on the overall pipeline route and the actual fault locations within the area, so as to lay the foundation for subsequent determination of the distribution of detection sensors.

[0085] Set distribution length Set the start and end points of the pipeline, starting from the start point and at intervals of... For each meter, a virtual fault point is set. The virtual fault point does not participate in the calculation of the distribution coefficient of the real fault point.

[0086] It should be further explained that, in this scheme, the data is divisible by integers to facilitate calculation. For example, if the total length of the pipeline is 20,000 meters, the distribution length can be 200 meters or 400 meters. If the total length of the pipeline is not an integer, the last region needs to be extended or shortened when cutting. For example, if the total length of the pipeline is 20,001 meters, the distribution length can be 200 meters or 400 meters, then the length of the last region is 201 meters or 401 meters.

[0087] Based on the pipeline's production year and various failure points Calculate the distribution coefficient of each fault point within the sub-interval. The above sub-intervals are left-closed and right-open intervals, such as the first sub-interval being [0, ... The distribution coefficient of the i-th fault point The specific calculation formula is as follows:

[0088]

[0089] in, Indicates the current year. This represents the production year of the pipeline at the i-th fault point. For virtual fault points, this value is the average production year of all pipelines. This represents the number of actual fault points surrounding the i-th fault point;

[0090] To determine the required number of sensors within a given area, the maximum value of the distribution coefficient within each sub-interval is extracted. The maximum value of the distribution coefficient within the i-th sub-interval is... Initialize the evaluation coefficients The relationship between the distribution coefficient and the number of pipeline detection sensors within this sub-interval is shown in the following formula:

[0091]

[0092]

[0093]

[0094] in, This represents the number of pipe detection sensors in the i-th sub-interval;

[0095] when When the value is zero, the pipeline detection sensor is positioned at the center of the i-th sub-interval. When the value is not zero, the pipeline detection sensors are evenly distributed within the i-th sub-interval;

[0096] It should be further explained that the above methods can achieve efficient utilization of detection sensors, timely detection of pipeline faults, improve pipeline inspection efficiency, reduce socio-economic risks, and have certain economic and social value.

[0097] Based on the number and distribution of pipeline inspection sensors, detection sub-regions are established. Preliminary inspection parameters of the pipeline inspection sensors within each sub-region are obtained. These preliminary parameters are then characterized to obtain pipeline characteristic parameters. Specifically:

[0098] After the sensors are deployed, detection sub-regions are established based on the number and distribution of the pipeline detection sensors. Each detection sub-region and sub-interval corresponds to one another. It should be further noted that the detection sub-regions are given different names to distinguish between the pre-detection stage and the detection stage.

[0099] The initial detection parameters of the pipeline detection sensors in each detection sub-region are obtained, including pipeline pressure and flow rate. The initial detection parameters are then characterized to obtain the pipeline characteristic pressure and pipeline characteristic flow rate.

[0100] It should be further explained that pressure and flow values ​​are only instantaneous values. However, instantaneous values ​​are sporadic. Therefore, by performing characteristic processing on the initial detection parameters, the utilization effect and characteristic of the parameters can be improved, thereby further improving the accuracy of the detection.

[0101] For pipeline pressure characterization, the pipeline pressure set of the i-th detection sub-region is: ,in , This represents the pipe pressure of the j-th pipe detection sensor in the i-th detection sub-region, where j represents the serial number and the characteristic pipe pressure of the i-th detection sub-region. The calculation formula is as follows:

[0102]

[0103] in, This represents the pipe pressure of the (j-1)th pipe detection sensor in the ith detection sub-region;

[0104] For pipeline flow characteristic processing, the pipeline flow set of the i-th detection sub-region is: ,in , This represents the pipe flow rate of the j-th pipe detection sensor in the i-th detection sub-region, and the characteristic pipe flow rate of the i-th detection sub-region. The calculation formula is as follows:

[0105]

[0106] in, This represents the pipe flow rate of the (j+1)th pipe detection sensor in the i-th detection sub-region;

[0107] Feature coordinates are established based on pipeline characteristic parameters. A regional fault detection algorithm is designed to obtain a health center point. The regional fault coefficient is then calculated using the health center point and feature coordinates to determine the fault sub-regions. Specifically:

[0108] The characteristic pressure and characteristic flow rate of the pipeline after characterization constitute the characteristic parameters of the pipeline.

[0109] The feature coordinates are established based on the pipe feature parameters. The feature coordinates of the i-th detection sub-region are ( , Design a regional fault detection algorithm and use the sparrow algorithm to determine the penalty coefficient in the regional fault detection algorithm. The detection sub-regions for anomalies are detected using regression analysis.

[0110] Let the plane equation of the classification line be:

[0111]

[0112] in, These are the weighting coefficients, and z is the bias value. ;

[0113] Introducing slack variables The following equation is satisfied:

[0114]

[0115] in, It is a sample classification function. For this scheme, the detection result has only two outcomes: fault or health, thus transforming the complex decision-making problem into a simple binary problem.

[0116] The objective function f of the regional fault detection algorithm is shown in the following equation:

[0117]

[0118] in, It is the number of sub-regions to be detected. Denotes the Euclidean norm. Indicates the first There are slack variables, where g represents the index and N represents the number of feature coordinates;

[0119] The underground pipeline inspection problem is further transformed into a convex quadratic programming optimization problem. A regional fault detection algorithm is used to determine whether each detection sub-region is faulty or healthy. Training and test sets are constructed from empirical data of underground pipelines, with a data ratio of 7:3. Feature coordinates of healthy and faulty pipelines are randomly selected as training samples, and the remaining sample data is used as test samples. A penalty coefficient is determined. ;

[0120] It should be further explained that existing network big data information is used in the process of constructing the training set and the test set. Both the training set and the test set contain the feature coordinates of health and the feature coordinates of faults. By utilizing existing information and experience values, it is more conducive to detecting whether the pipeline has failed.

[0121] After training, a health center point is obtained. The regional fault coefficient is calculated from the health center point and the feature coordinates. The regional fault coefficient is the Euclidean distance W from the feature coordinates of each detection sub-region to the health center point.

[0122] Initialize diagnostic thresholds If the failure factor W is greater than If the detected sub-region is faulty, then the fault coefficient W is less than or equal to 1. If the result is positive, then the detected sub-region is considered healthy.

[0123] Deploy pipeline robots equipped with industrial cameras to the faulty sub-region. Use these robots to capture pipeline images, design an adaptive fault detection algorithm to detect fault points, and record their locations. Specifically:

[0124] After identifying the faulty sub-region, a coordinate system is established with the starting position of the faulty sub-region as the origin. A pipeline robot is deployed to the starting position of the faulty sub-region. The pipeline robot is equipped with an industrial camera and an inertial measurement unit. The industrial camera acquires images of the pipeline, where the resolution of the pipeline images is [resolution missing]. In this scheme, the resolution of the entire pipeline image is [missing information]. ,in All values ​​are integers, and the acquisition frequency is V. In this scheme, the acquisition frequency is 5 frames per second, and each pixel includes a temperature value and a grayscale value.

[0125] The inertial measurement unit records the real-time position of the pipeline robot, which is the position of the center point of each pipeline image.

[0126] The design of an adaptive fault point detection algorithm to detect the location of fault points involves the following steps:

[0127] (1) Calculate the single-frame deviation coefficient. Segment the pipeline image to form... In this scheme, there are 8 sub-images, where m represents an integer. The maximum grayscale value of each sub-image is calculated. The single-frame deviation coefficient is calculated from the maximum grayscale value of the sub-image. This is the single-frame deviation coefficient of the i-th pipeline image. The calculation formula is as follows:

[0128]

[0129] Where k represents the sequence number, which is an integer. This represents the maximum grayscale value of the k-th sub-image of the i-th pipeline image. This represents the average of the maximum grayscale values ​​in the i-th pipeline image;

[0130] It should be further explained that the purpose of the single-frame deviation coefficient is to find the specific location of the fault point. When finding the specific location of the fault point, in order to facilitate the calculation, the average value of all gray values ​​in the neighboring pipe images can also be compared. If they are similar, the pipe image with the larger value is the location of the fault point. This approach is more economical for hardware devices with insufficient computing power.

[0131] (2) Locate pipeline images with fault points and set a single-frame deviation threshold. If the single-frame deviation coefficient of the i-th pipeline image is less than the single-frame deviation threshold, then there are no suspicious fault points in the image and it is marked as safe. If the single-frame deviation coefficient of the i-th pipeline image is greater than or equal to the single-frame deviation threshold, then there are suspicious fault points in the image and it is marked as fault.

[0132] (3) Obtain a set of stable fault point images, count the number J of pipe images marked as faults, reorder the pipe images marked as faults, and calculate the screening coefficient of each sub-image in the above pipe images, and the screening coefficient of the j-th sub-image of the i-th pipe image. The calculation formula is as follows:

[0133]

[0134] in, This represents the number of pixels surrounding the pixel with the maximum grayscale value in the j-th sub-image of the i-th pipeline image. This represents the temperature value of the pixel with the highest grayscale value in the j-th sub-image of the i-th pipeline image. Let represent the temperature value of the k-th pixel surrounding the pixel with the maximum grayscale value in the j-th sub-image of the i-th pipeline image, and let e represent the exponential function;

[0135] The selection coefficient for each pipeline image is the average of the selection coefficients for all sub-images;

[0136] Initialize the filtering threshold A. Calculate the adaptive filtering threshold based on the number of pipeline images marked as faulty, as shown in the following expression:

[0137]

[0138]

[0139] in, and Indicates the filter value;

[0140] If the screening coefficient of the i-th pipeline image is greater than the adaptive screening threshold, the image is marked as a safe and stable fault point image. If the screening coefficient of the i-th pipeline image is less than or equal to the adaptive screening threshold, the pipeline image is removed to obtain a set of stable fault point images.

[0141] (4) Calculate the location of the fault point. For each stable fault point image, calculate the location of each stable fault point image. The location of the fault point is the average of the locations of each stable fault point image.

[0142] Output the pipeline inspection results and notify the pipeline inspection personnel.

[0143] Example 2

[0144] This embodiment provides an underground pipeline detection system based on multiple information sources, which includes the following modules:

[0145] The sensor distribution planning module allows the pipeline detection system to acquire historical information about pipelines in the area, create a fault distribution map, calculate the distribution coefficient, and deploy pipeline detection sensors based on the distribution coefficient.

[0146] The region determination module establishes detection sub-regions based on the number and distribution of pipeline detection sensors, obtains the initial detection parameters of the detection sensors in each detection sub-region, performs feature processing on the initial detection parameters to obtain pipeline feature parameters, establishes feature coordinates from the pipeline feature parameters, designs a region fault detection algorithm to obtain the health center point, calculates the region fault coefficient from the health center point and feature coordinates, and determines the fault sub-region.

[0147] The module calculates the location of the fault point, deploys a pipeline robot to the fault sub-region, the pipeline robot is equipped with an industrial camera to collect pipeline images, an adaptive fault point detection algorithm is designed to detect the fault point, and the location of the fault point is recorded.

[0148] The results output module outputs the pipeline inspection results and notifies the pipeline inspection personnel.

[0149] As used herein, the term "preferred" is meant as an example, illustration, or illustration. Any aspect or design described herein as "preferred" need not be construed as being more advantageous than other aspects or designs. Rather, the use of the term "preferred" is intended to present the concept in a specific manner. As used in this application, the term "or" is intended to mean an inclusive "or" rather than an exclusionary "or." That is, unless otherwise specified or clear from the context, "X uses A or B" naturally includes either of the permutations. That is, if X uses A; X uses B; or X uses both A and B, then "X uses A or B" is satisfied in any of the foregoing examples.

[0150] Furthermore, although this disclosure has been shown and described with respect to one or more implementations, equivalent variations and modifications will occur to those skilled in the art based on a reading and understanding of this specification and the accompanying drawings. This disclosure includes all such modifications and variations and is limited only by the scope of the appended claims. In particular, with respect to the various functions performed by the aforementioned components (e.g., elements, etc.), the terminology used to describe such components is intended to correspond to any component (unless otherwise indicated) that performs the specified function of said component (e.g., is functionally equivalent to it), even if structurally not equivalent to the disclosed structure performing the functions in the exemplary implementations of this disclosure shown herein. Moreover, although specific features of this disclosure have been disclosed with respect to only one of several implementations, such features may be combined with one or more features of other implementations that may be desirable and advantageous for a given or particular application. Furthermore, with regard to the use of the terms “comprising,” “having,” “containing,” or variations thereof in the Detailed Description or claims, such terms are intended to be included in a manner similar to the term “including.”

[0151] The functional units in this invention embodiment can be integrated into a processing module, or each unit can exist physically separately, or multiple units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. The aforementioned devices or systems can execute the storage methods in the corresponding method embodiments.

[0152] In summary, the above embodiments are one implementation of the present invention, but the implementation of the present invention is not limited to the embodiments described above. Any changes, modifications, substitutions, combinations, or simplifications made that deviate from the spirit and principle of the present invention should be considered equivalent substitutions and are included within the protection scope of the present invention.

Claims

1. A method for detecting underground pipelines based on multiple information sources, characterized in that, Includes the following steps: S1: The pipeline inspection system acquires historical information about regional pipelines, creates a fault distribution map, calculates the distribution coefficient, and deploys pipeline inspection sensors based on the distribution coefficient. S2: Based on the number and distribution of pipeline detection sensors, establish detection sub-regions, obtain the initial detection parameters of pipeline detection sensors in each detection sub-region, perform feature processing on the initial detection parameters, and obtain pipeline feature parameters. S3: Establish feature coordinates from pipeline feature parameters, design regional fault detection algorithm, obtain health center point, calculate regional fault coefficient from health center point and feature coordinates, and determine fault sub-region; Step S3 includes: The feature coordinates are established based on the pipe feature parameters. The feature coordinates of the i-th detection sub-region are ( , Design a regional fault detection algorithm and use the sparrow algorithm to determine the penalty coefficient in the regional fault detection algorithm. The detection sub-regions for anomalies are detected using regression analysis. Let the plane equation of the classification line be: ； in, These are the weighting coefficients, and z is the bias value. ; Introducing slack variables The following equation is satisfied: ； in, It is a sample classification function, and the detection result includes either a faulty or healthy result; The objective function f of the regional fault detection algorithm is shown in the following equation: ； in, It is the number of sub-regions to be detected. Denotes the Euclidean norm. Indicates the first There are slack variables, where g represents the ordinal number; The underground pipeline inspection problem is transformed into a convex quadratic programming optimization problem. A regional fault detection algorithm is used to determine whether each detection sub-region is faulty or healthy. Training and test sets are constructed from empirical data of underground pipelines, with a data ratio of 7:

3. Feature coordinates of healthy and faulty pipelines are randomly selected as training samples, and the remaining sample data is used as test samples. A penalty coefficient is determined. ; After training, a health center point is obtained. The regional fault coefficient is calculated from the health center point and the feature coordinates. The regional fault coefficient is the Euclidean distance W from the feature coordinates of each detection sub-region to the health center point. Initialize diagnostic thresholds If the regional failure coefficient W is greater than If the detected sub-region is faulty, then the fault coefficient W of the region is less than or equal to 1. If so, then the detected sub-region is healthy; S4: Deploy pipeline robots to the faulty sub-region. The pipeline robots are equipped with industrial cameras to collect pipeline images. Design an adaptive fault point detection algorithm to detect fault points and record their locations. Step S4 includes: A coordinate system is established with the starting position of the fault sub-region as the origin. A pipeline robot is deployed to the starting position of the fault sub-region. The pipeline robot is equipped with an industrial camera and an inertial measurement unit. The industrial camera acquires pipeline images, where the resolution of the pipeline images is [resolution missing]. ,in All values ​​are integers, the sampling frequency is V, and each pixel includes a temperature value and a grayscale value. The inertial measurement unit records the real-time position of the pipeline robot, where the real-time position of the pipeline robot is the position of the center point of each pipeline image; The design of an adaptive fault point detection algorithm to detect the location of fault points involves the following steps: (1) Calculate the single-frame deviation coefficient: The pipeline image is segmented to form There are several sub-images, where m represents an integer. Calculate the maximum grayscale value of each sub-image. The single-frame deviation coefficient is calculated from the maximum grayscale value of the sub-image. This is the single-frame deviation coefficient of the i-th pipeline image. The calculation formula is as follows: ； Where k represents the sequence number, which is an integer. This represents the maximum grayscale value of the k-th sub-image of the i-th pipeline image. This represents the average of the maximum grayscale values ​​in the i-th pipeline image; (2) Locate pipeline images with fault points and set a single-frame deviation threshold. If the single-frame deviation coefficient of the i-th pipeline image is less than the single-frame deviation threshold, then there are no suspicious fault points in the image and it is marked as safe; if the single-frame deviation coefficient of the i-th pipeline image is greater than or equal to the single-frame deviation threshold, then there are suspicious fault points in the image and it is marked as fault. (3) Obtain a set of stable fault point images, count the number J of pipe images marked as faulty, reorder the pipe images marked as faulty, calculate the screening coefficient of each sub-image in the faulty pipe image, and the screening coefficient of the j-th sub-image of the i-th pipe image. The calculation formula is as follows: ； in, This represents the number of pixels surrounding the pixel with the maximum grayscale value in the j-th sub-image of the i-th pipeline image. This represents the temperature value of the pixel with the highest grayscale value in the j-th sub-image of the i-th pipeline image. Let represent the temperature value of the k-th pixel surrounding the pixel with the maximum grayscale value in the j-th sub-image of the i-th pipeline image, and let e represent the exponential function; The selection coefficient for each pipeline image is the average of the selection coefficients for all sub-images; Initialize the filtering threshold A. Calculate the adaptive filtering threshold based on the number of pipeline images marked as faulty, as shown in the following expression: ； ； in, and Indicates the filter value; If the screening coefficient of the i-th pipeline image is greater than the adaptive screening threshold, the image is marked as a safe and stable fault point image. If the screening coefficient of the i-th pipeline image is less than or equal to the adaptive screening threshold, the pipeline image is removed to obtain a set of stable fault point images. (4) Calculate the location of the fault point. For each stable fault point image, calculate the location of each stable fault point image. The location of the fault point is the average of the locations of each stable fault point image. S5: Output pipeline inspection results and notify pipeline inspection personnel.

2. The underground pipeline detection method based on multiple information sources according to claim 1, characterized in that, Step S1 includes: The pipeline inspection system acquires historical information about regional pipelines. Regional pipelines refer to pipelines within a continuous area, and these pipelines have single-layer continuity. Historical information on regional pipelines includes the year the pipeline was manufactured, the actual location of pipeline failures and the number of actual failure locations since the year the pipeline was manufactured. A fault distribution map is created based on the fault locations. The fault distribution map includes the overall pipeline route within the area and the actual fault locations. Set distribution length Set the start and end points of the pipeline, starting from the start point and at intervals of... For each meter, a virtual fault point is set. The virtual fault point does not participate in the calculation of the distribution coefficient of the real fault point. Based on the pipeline's production year and various failure points Calculate the distribution coefficient of each fault point within the sub-interval. Wherein, the sub-interval is a left-closed, right-open interval, and the distribution coefficient of the i-th fault point is... The specific calculation formula is as follows: ； in, Indicates the current year. This represents the production year of the pipeline at the i-th fault point. For virtual fault points, this value is the average production year of all pipelines. This represents the number of actual fault points surrounding the i-th fault point; Extract the maximum value of the distribution coefficient within each subinterval, where the maximum value of the distribution coefficient within the i-th subinterval is... Initialize the evaluation coefficients The distribution coefficient and the number of pipeline detection sensors within the i-th sub-interval The relationship is shown in the following formula: ； ； ； when When the value is zero, the pipeline detection sensor is positioned at the center of the i-th sub-interval. When the value is not zero, the pipeline detection sensors are evenly distributed within the i-th sub-interval.

3. The underground pipeline detection method based on multiple information sources according to claim 2, characterized in that, Step S2 includes: Based on the number and distribution of pipeline detection sensors, detection sub-regions are established, and each detection sub-region and sub-interval corresponds to one; The initial detection parameters of the pipeline detection sensors in each detection sub-region are obtained, including pipeline pressure and flow rate. The initial detection parameters are then characterized to obtain the pipeline characteristic pressure and pipeline characteristic flow rate. For pipeline pressure characterization, the pipeline pressure set of the i-th detection sub-region is: ,in , This represents the pipe pressure of the j-th pipe detection sensor in the i-th detection sub-region, where j represents the serial number and the characteristic pipe pressure of the i-th detection sub-region. The calculation formula is as follows: ； in, This represents the pipe pressure of the (j-1)th pipe detection sensor in the ith detection sub-region; For pipeline flow characteristic processing, the pipeline flow set of the i-th detection sub-region is: ,in , This represents the pipe flow rate of the j-th pipe detection sensor in the i-th detection sub-region, and the characteristic pipe flow rate of the i-th detection sub-region. The calculation formula is as follows: ； in, This represents the pipe flow rate of the (j+1)th pipe detection sensor in the i-th detection sub-region; The characteristic pressure and characteristic flow rate of the pipeline after characterization constitute the characteristic parameters of the pipeline.

4. A multi-information-source-based underground pipeline detection system, characterized in that, include: The sensor distribution planning module allows the pipeline detection system to acquire historical information about pipelines in the area, create a fault distribution map, calculate the distribution coefficient, and deploy pipeline detection sensors based on the distribution coefficient. The region determination module establishes detection sub-regions based on the number and distribution of pipeline detection sensors, obtains the initial detection parameters of the detection sensors in each detection sub-region, performs feature processing on the initial detection parameters to obtain pipeline feature parameters, establishes feature coordinates from the pipeline feature parameters, designs a region fault detection algorithm to obtain the health center point, calculates the region fault coefficient from the health center point and feature coordinates, and determines the fault sub-region. The module calculates the location of the fault point, deploys a pipeline robot to the fault sub-region, the pipeline robot is equipped with an industrial camera to collect pipeline images, an adaptive fault point detection algorithm is designed to detect the fault point, and the location of the fault point is recorded. The results output module outputs the pipeline inspection results and notifies the pipeline inspection personnel. To implement the underground pipeline detection method based on multiple information sources as described in any one of claims 1-3.

Images