Water supply pipeline leakage intelligent diagnosis method based on YOLO deep learning

Abstract

The invention relates to the technical field of water supply pipeline leakage diagnosis, in particular to a water supply pipeline leakage intelligent diagnosis method based on YOLO deep learning. Comprising the following steps: acquiring a pipeline radar echo image; leakage image features are marked, and pipeline leakage image training is carried out by adopting a YOLO deep learning large model; judging and marking a leakage area; a radar image leakage area reflection signal is extracted, and the propagation speed of electromagnetic waves in the soil is calculated; calculating a soil dielectric constant of the leakage area; calculating an actual soil dielectric constant interval through the soil moisture content; judging whether the soil dielectric constant of the leakage area accords with an actual soil dielectric constant interval or not, verifying a YOLO diagnosis result, and outputting judgment data; through the above mode, the detection efficiency is improved, the misjudgment rate is reduced, the method is suitable for rapid inspection of large-scale water supply pipelines, and an efficient and reliable solution is provided for urban infrastructure management.

Description

Technical Field

[0001] The present invention relates to the technical field of leakage diagnosis of water supply pipelines, and particularly to an intelligent leakage diagnosis method for water supply pipelines based on YOLO deep learning. Background Art

[0002] As an important part of urban infrastructure, the safe operation of water supply pipelines is directly related to the stable development of residents' lives, industrial production, and social economy. Traditional water supply pipeline leakage detection methods mainly rely on manual inspections or simple sensor networks, which are inefficient and prone to missed detections. The existing determination of pipeline radar image leakage depends to a large extent on the subjective interpretation of radar images by technicians, lacking a unified objective standard, resulting in poor consistency of detection results and low efficiency. In addition, the interpretation of ground penetrating radar images requires rich experience, and the ability level of technicians directly affects the accuracy of diagnosis, which not only makes the detection process time-consuming and laborious but also easily introduces errors due to human factors. Although some studies have attempted to introduce image processing algorithms to assist in the analysis in recent years, these methods are often limited to simple image feature extraction and are difficult to cope with complex and changeable buried environments, unable to fundamentally solve the problems of low detection efficiency and high misjudgment rate.

[0003] On the other hand, there are various interference factors in the buried environment, such as cavities, stones, impurities, etc. These substances will significantly affect the radar echo signal, resulting in abnormal features unrelated to leakage in the image, thus increasing the difficulty of leakage diagnosis. Existing technologies usually cannot effectively distinguish these interference signals from real leakage signals, easily causing misjudgments. In addition, traditional methods lack dynamic analysis of soil dielectric constant changes and cannot make full use of the characteristics of electromagnetic wave propagation to verify leakage diagnosis results, further restricting the accuracy and reliability of detection.

[0004] Therefore, it is very necessary to propose an intelligent leakage diagnosis method for water supply pipelines that can improve the detection efficiency and reduce the misjudgment rate. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent leakage diagnosis method for water supply pipelines based on YOLO deep learning, aiming to improve the detection efficiency and reduce the misjudgment rate.

[0006] To achieve the above purpose, an intelligent leakage diagnosis method for water supply pipelines based on YOLO deep learning adopted by the present invention includes the following steps:

[0007] Use a ground penetrating radar to detect the pipeline and obtain the pipeline radar echo image;

[0008] Establish an image set of pipeline leakage states, mark the leakage image features, and use the YOLO deep learning large model to train the pipeline leakage images;

[0009] Based on the training results of YOLO deep learning, perform target image detection, identify the radar echo image of the leaking pipeline, determine and mark the leakage area;

[0010] Extract the reflection signal of the leakage area in the radar image, and use the two-way travel time of electromagnetic waves to calculate the propagation speed of electromagnetic waves in the soil;

[0011] Calculate the soil dielectric constant of the leakage area through the propagation speed of electromagnetic waves;

[0012] Using the Topp formula, calculate the actual soil dielectric constant range through the soil moisture content;

[0013] Judge whether the soil dielectric constant of the leakage area conforms to the actual soil dielectric constant range, verify the YOLO diagnosis result, and output the judgment data.

[0014] Among them, in the step of using a ground penetrating radar to detect the pipeline and obtain the pipeline radar echo image:

[0015] In the case of pipeline leakage, lay parallel to the pipeline laying direction, conduct long-distance detection directly above the pipeline, divide and calibrate at intervals of 5m during detection, and perform distance normalization processing to obtain multiple pipeline radar echo images, and the images are in JPG format.

[0016] Among them, in the step of extracting the reflection signal of the leakage area in the radar image and using the two-way travel time of electromagnetic waves to calculate the propagation speed of electromagnetic waves in the soil:

[0017] The calculation method of the propagation speed of electromagnetic waves in the soil is:

[0018]

[0019] Among them, v is the propagation speed of electromagnetic waves in the medium, with the unit of m / s; d is the buried depth of the pipeline, with the unit of m; TWTT is the two-way travel time of electromagnetic waves, that is, the total time it takes for electromagnetic waves to be emitted from the radar to the pipeline and reflected back to the radar receiver, with the unit of s.

[0020] Among them, in the step of calculating the soil dielectric constant of the leakage area through the propagation speed of electromagnetic waves:

[0021] The calculation method of the soil dielectric constant of the leakage area is:

[0022]

[0023] Among them, ε r is the soil dielectric constant; c is the propagation speed of light in a vacuum, taking 3×10 8 m / s; v is the propagation speed of electromagnetic waves in the medium, with the unit of m / s.

[0024] Among them, in the step of calculating the actual soil dielectric constant range through soil moisture content using the Topp formula:

[0025] The calculation method of the actual soil dielectric constant is as follows:

[0026]

[0027] Among them, ε r is the soil dielectric constant; θ v is the percentage of soil moisture content;

[0028] It is defined that the water content ≥ 30% is wet soil, and the water content of 100% is completely waterlogged soil;

[0029] When the water content is 30%: ε r = 3.03 + 9.3×0.3 + 146×0.3 2 - 76.7×0.3 3 = 16.89;

[0030] When the water content is 100%: ε r = 3.03 + 9.3×1 + 146×1 2 - 76.7×1 3 = 81.63;

[0031] Then the soil dielectric constant range is [16, 82].

[0032] Among them, in the step of judging whether the soil dielectric constant of the leakage area conforms to the actual soil dielectric constant range, verifying the YOLO diagnosis result, and outputting the judgment data:

[0033] If the soil dielectric constant of the leakage area conforms to the actual soil dielectric constant range, it is the leakage area.

[0034] Among them, in the step of judging whether the soil dielectric constant of the leakage area conforms to the actual soil dielectric constant range, verifying the YOLO diagnosis result, and outputting the judgment data:

[0035] If the soil dielectric constant of the leakage area does not conform to the actual soil dielectric constant range, it is not the leakage area.

[0036] An intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning of the present invention uses a ground penetrating radar to detect the pipeline, obtains the pipeline radar echo image; establishes an image set of pipeline leakage states, marks the leakage image features, and uses the YOLO deep learning large model to train the pipeline leakage images; performs target image detection based on the YOLO deep learning training results, identifies the radar echo image of the leakage pipeline, determines and marks the leakage area; extracts the reflection signal in the leakage area of the radar image, and calculates the propagation speed of electromagnetic waves in the soil by using the two-way travel time of electromagnetic waves; calculates the soil dielectric constant of the leakage area through the propagation speed of electromagnetic waves; uses the Topp formula to calculate the actual soil dielectric constant interval through the soil moisture content; determines whether the soil dielectric constant in the leakage area conforms to the actual soil dielectric constant interval, verifies the YOLO diagnosis result, and outputs the judgment data; realizes improving the detection efficiency, reducing the misjudgment rate, is applicable to the rapid inspection of large-scale water supply pipelines, and provides an efficient and reliable solution for urban infrastructure management. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0038] Figure 1 is a flowchart of the steps of the intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning of the present invention.

[0039] Figure 2 is a schematic diagram of the process of the intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning of the present invention.

[0040] Figure 3 is a schematic diagram of pipeline laying for pipeline detection of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application.

[0042] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms "a", "the", and "said" used in this application and the appended claims are also intended to include the plural forms unless the context clearly dictates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0043] It should be understood that although the terms first, second, third, etc. may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of this application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to a determination".

[0044] Please refer to Figures 1 to 3 , the present invention provides an intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning, including the following steps:

[0045] S100: Use ground penetrating radar to detect the pipeline and obtain the pipeline radar echo image.

[0046] In this embodiment, ground penetrating radar is used to detect the pipeline and obtain the pipeline radar echo image. As Figure 3 shown, in the case of pipeline leakage, parallel to the pipeline laying direction, long-distance detection is carried out directly above the pipeline. During detection, calibration is carried out by dividing at intervals of 5m, and distance normalization processing is performed to obtain multiple pipeline radar echo images, and the images are in JPG format.

[0047] S200: Establish an image set of pipeline leakage states, mark the characteristics of leakage images, and use the YOLO deep learning large model to train pipeline leakage images.

[0048] In this embodiment, after obtaining the pipeline radar echo image, an image set of pipeline leakage states is established, the characteristics of leakage images are marked, and the YOLO deep learning large model is used to train pipeline leakage images.

[0049] S300: Perform target image detection based on the YOLO deep learning training results, identify the radar echo image of the leaking pipeline, and determine and mark the leakage area;

[0050] In this embodiment, target image detection is performed based on the YOLO deep learning training results, the radar echo image of the leaking pipeline is identified, and the leakage area is determined and marked, such as leakage area i and leakage area j.

[0051] S400: Extract the reflection signals in the leakage areas of the radar images, and calculate the propagation speed of electromagnetic waves in the soil by using the two-way travel time of electromagnetic waves.

[0052] In this embodiment, extract the reflection signals in leakage area i and leakage area j of the radar image, and calculate the propagation speed of electromagnetic waves in the soil by using the two-way travel time of electromagnetic waves. The calculation method of the propagation speed of electromagnetic waves in the soil is as follows:

[0053]

[0054] Among them, v is the propagation speed of electromagnetic waves in the medium, with the unit of m / s; d is the burial depth of the pipeline, with the unit of m; TWTT is the two-way travel time of electromagnetic waves, that is, the total time it takes for electromagnetic waves to be emitted from the radar to the pipeline and then reflected back to the radar receiver, with the unit of s.

[0055] For example, if the burial depth of the pipeline in image i is 1.5 m and the two-way travel time of electromagnetic waves is 20 ns, and the burial depth of the pipeline in image j is 1.5 m and the two-way travel time of electromagnetic waves is 77 ns, then:

[0056]

[0057] S500: Calculate the soil dielectric constant of the leakage area through the propagation speed of electromagnetic waves.

[0058] In this embodiment, calculate the soil dielectric constant of the leakage area through the propagation speed of electromagnetic waves. The calculation method of the soil dielectric constant of the leakage area is as follows:

[0059]

[0060] Among them, ε r is the soil dielectric constant; c is the propagation speed of light in a vacuum, taking 3×10 8 m / s; v is the propagation speed of electromagnetic waves in the medium, with the unit of m / s.

[0061] Then for leakage area i and leakage area j:

[0062]

[0063] S600: Use the Topp formula to calculate the actual soil dielectric constant range through the soil moisture content.

[0064] In this embodiment, use the Topp formula to calculate the actual soil dielectric constant range through the soil moisture content. The calculation method of the actual soil dielectric constant is as follows:

[0065]

[0066] Among them, ε r is the soil dielectric constant; θ vis the percentage of soil moisture content;

[0067] Under normal dry conditions, without rainfall and no pipeline leakage, the soil water content is usually relatively low, about 10%. When the water content rate ≥ 30%, it is wet soil, and there may be pipeline leakage. In the case of large leakage, the soil is completely soaked by water, and the water content is nearly 100%;

[0068] Define the soil with water content ≥ 30% as wet soil, and the soil with water content 100% as completely water-soaked soil;

[0069] When the water content is 30%: ε r = 3.03 + 9.3×0.3 + 146×0.3 2 - 76.7×0.3 3 = 16.89;

[0070] When the water content is 100%: ε r = 3.03 + 9.3×1 + 146×1 2 - 76.7×1 3 = 81.63;

[0071] Then the soil dielectric constant range is [16, 82].

[0072] S700: Determine whether the soil dielectric constant in the leakage area conforms to the actual soil dielectric constant range, verify the YOLO diagnosis result, and output the judgment data.

[0073] In this embodiment, determine whether the soil dielectric constant in the leakage area conforms to the actual soil dielectric constant range, verify the YOLO diagnosis result. If the soil dielectric constant in the leakage area conforms to the actual soil dielectric constant range, it is the leakage area; if the soil dielectric constant in the leakage area does not conform to the actual soil dielectric constant range, it is not the leakage area. Then, the pipeline area of the i image does not belong to the leakage area, and the pipeline area of the j image is confirmed as the leakage area.

[0074] In the present invention, first, ground penetrating radar is used for pipeline detection to obtain pipeline radar echo images; then, an image set of pipeline leakage states is established, the characteristics of leakage images are marked, and a large YOLO deep learning model is used for training pipeline leakage images; then, based on the training results of YOLO deep learning, target image detection is carried out to identify the radar echo images of leaking pipelines, and the leakage areas are determined and marked; then, the reflection signals in the leakage areas of the radar images are extracted, and the electromagnetic wave two-way travel time is used to calculate the propagation speed of electromagnetic waves in the soil; the soil dielectric constant in the leakage area is calculated through the electromagnetic wave propagation speed; then, using the Topp formula, the actual soil dielectric constant range is calculated through the soil moisture content; finally, it is judged whether the soil dielectric constant in the leakage area conforms to the actual soil dielectric constant range, the YOLO diagnosis result is verified, and the judgment data is output; thus, the detection efficiency is improved, the misjudgment rate is reduced, it is applicable to the rapid inspection of large-scale water supply pipelines, and an efficient and reliable solution is provided for urban infrastructure management.

[0075] After considering the specification and the practice disclosed herein, those skilled in the art will readily conceive of other embodiments of the present application. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include known common knowledge or conventional technical means in the technical field not disclosed in the present application.

[0076] It should be understood that the present application is not limited to the precise structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. An intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning, characterized in that: The steps include: Use ground penetrating radar to detect pipelines and obtain pipeline radar echo images; Establish a pipeline leakage status image set, mark the leakage image features, and use the YOLO deep learning model to train pipeline leakage images; Based on the YOLO deep learning training results, target image detection is performed to identify the radar echo image of the leaking pipeline, and the leakage area is determined and marked; Extract the reflected signal from the leakage area of ​​the radar image, and use the two-way travel time of the electromagnetic wave to calculate the propagation speed of the electromagnetic wave in the soil; Calculate the soil dielectric constant of the leakage area by the propagation velocity of electromagnetic waves; Using the Topp formula, the actual soil dielectric constant range is calculated through soil moisture content; Determine whether the soil dielectric constant in the leakage area conforms to the actual soil dielectric constant range, verify the YOLO diagnosis results, and output the judgment data.

2. The intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning according to claim 1, characterized in that: In the steps of using ground penetrating radar to detect pipelines and obtain pipeline radar echo images: In case of pipeline leakage, long-distance detection is carried out directly above the pipeline in parallel with the pipeline laying direction. The detection is divided and calibrated at intervals of 5m, and the distance is homogenized to obtain multiple pipeline radar echo images in JPG format.

3. The intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning as claimed in claim 2, characterized in that: In the steps of extracting the reflected signal of the leakage area of ​​the radar image and using the two-way travel time of the electromagnetic wave to calculate the propagation speed of the electromagnetic wave in the soil: The propagation speed of electromagnetic waves in soil is calculated as: Among them, v is the propagation speed of electromagnetic waves in the medium, in meters per second; d is the buried depth of the pipeline, in meters; TWTT is the two-way travel time of the electromagnetic wave, that is, the total time it takes for the electromagnetic wave to be emitted from the radar to the pipeline and reflected back to the radar receiver, in seconds.

4. The intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning as claimed in claim 3 is characterized in that: In the steps of calculating the soil dielectric constant of the leakage area by electromagnetic wave propagation velocity: The soil dielectric constant in the leakage area is calculated as: Among them, ε r is the dielectric constant of the soil; c is the speed of light in a vacuum, which is 3×10 8 m / s; v is the propagation speed of electromagnetic waves in the medium, measured in meters per second.

5. The intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning as claimed in claim 4, characterized in that: In the step of calculating the actual soil dielectric constant range by soil moisture content using the Topp formula: The actual soil dielectric constant is calculated as: Among them, ε r is the soil dielectric constant; θ v is the soil moisture content percentage; A soil with a water content of ≥30% is defined as moist soil, and a soil with a water content of 100% is defined as soil completely soaked in water; When the water content is 30%: ε r =3.03+9.3×0.3+146×0.3 2 -76.7×0.3 3 =16.89; When the water content is 100%: ε r =3.03+9.3×1+146×1 2 -76.7×1 3 =81.63; Then the range of soil dielectric constant is [16,82].

6. The intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning according to claim 5, characterized in that: In the step of judging whether the soil dielectric constant of the leakage area conforms to the actual soil dielectric constant range, verifying the YOLO diagnosis result, and outputting the judgment data: If the soil dielectric constant of the leakage area conforms to the actual soil dielectric constant range, it is a leakage area.

7. The intelligent diagnosis method for water supply pipeline leakage based on YOLO deep learning according to claim 6, characterized in that: In the step of judging whether the soil dielectric constant of the leakage area conforms to the actual soil dielectric constant range, verifying the YOLO diagnosis result, and outputting the judgment data: If the soil dielectric constant in the leakage area does not conform to the actual soil dielectric constant range, it is not a leakage area.

Images