Autonomous cruise unmanned aerial vehicle inspection method and system for oil and gas pipe network

Abstract

The invention discloses an autonomous cruise unmanned aerial vehicle inspection method and system for an oil and gas pipe network, and belongs to the technical field of oil and gas pipeline safety monitoring, and the method comprises the steps: S10, obtaining related data of a target oil and gas pipeline, the related data comprising normal thermal imaging data, obtained by a target unmanned aerial vehicle, of the target oil and gas pipeline, obtaining a corresponding obtaining height when the target unmanned aerial vehicle obtains the normal thermal imaging data; and S20, processing the normal thermal imaging data, including distinguishing the parts of the oil and gas pipeline corresponding to the thermal imaging data in the normal thermal imaging data, converting the parts into thermal imaging images according to the thermal imaging data, and obtaining the parts corresponding to the fast temperature change in the converted thermal imaging images. Through data acquisition, accurate identification of a risk part, analysis of a risk trend, pipeline leakage verification and oil and gas transmission state evaluation, the oil and gas pipe network inspection efficiency and accuracy are improved, and a powerful guarantee is provided for safe operation of the oil and gas pipe network.

Description

Technical Field

[0001] This invention relates to the field of oil and gas pipeline safety monitoring technology, and in particular to an autonomous patrol drone inspection method and system for oil and gas pipeline networks. Background Technology

[0002] In real life, pipeline transportation has become one of the main methods of transporting oil and gas due to its advantages such as large transport capacity, good continuity and low cost. In order to ensure the long-term safe use of oil and gas pipelines, it is necessary to inspect them.

[0003] Regarding this research, application CN202211715495.7 provides a method and system for flight control of a UAV used for inspecting oil and gas pipelines in mountainous areas. Compared with the traditional PID controller (Proportional-Integration-Derivative controller), this technical solution uses a dynamic inverse control law to linearize and decouple the system, achieving good control performance for high angle-of-attack maneuvers of fixed-wing UAVs. In addition, the introduction of a disturbance observer can effectively estimate the total disturbance caused by wind disturbance and model uncertainty, and compensate it into the dynamic inverse controller, improving the robustness of the system. Considering the nonlinear characteristics of the system, it can solve the problem that commonly used linear control algorithms cannot meet the requirements of flight robustness.

[0004] Another application, CN202410138878.5, discloses a drone-based system for inspecting oil and gas pipeline leaks. This solution includes a drone equipped with a precision cruise system, a detection system, and a data processing system. The precision cruise system supports the drone in cruising along the oil and gas pipeline's path. The detection system detects leaks in the pipeline and obtains detection data, including both natural gas and oil leaks. The data processing system generates inspection results based on the detection data. This solution utilizes fiber optic sensing to detect oil leaks, offering high accuracy and a simple structure.

[0005] To ensure the long-term safe operation of oil and gas pipelines, leak detection is a crucial inspection item. When a pipeline leaks, the temperature around the leak point rises abnormally due to gas evaporation or frictional heat generation. Inspection drones, equipped with infrared thermal imagers, identify areas of abnormal temperature by capturing differences in thermal radiation. However, the detection algorithms of these technologies have weak generalization capabilities for unknown scenarios and are easily affected by factors such as lighting and seasonal changes. For example, infrared thermal imaging may falsely report abnormal pipeline heating in high-temperature summer environments, and it is prone to cross-sensitivity in complex gas environments, affecting the reliability of determining whether an oil and gas pipeline leak has occurred. Summary of the Invention

[0006] In view of the problems existing in the field of existing oil and gas pipeline safety monitoring technology, the present invention is proposed.

[0007] Therefore, one of the objectives of this invention is to provide an autonomous patrol drone inspection method and system for oil and gas pipeline networks. By acquiring data, accurately identifying risky locations, analyzing risk trends, verifying pipeline leaks, assessing oil and gas transportation status, and extracting features and establishing a database, this method improves the efficiency and accuracy of oil and gas pipeline network inspections, providing strong support for the safe operation of oil and gas pipeline networks.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0009] On the one hand, the present invention provides an autonomous patrol drone inspection method for oil and gas pipeline networks, comprising the following steps:

[0010] S10: Acquire relevant data about the target oil and gas pipeline, including normal thermal imaging data about the target oil and gas pipeline acquired by the target UAV, and the acquisition altitude corresponding to the acquisition altitude when the target UAV acquires the normal thermal imaging data.

[0011] S20: Process the normal thermal imaging data. The processing includes distinguishing the parts of the oil and gas pipelines corresponding to each thermal imaging data in the normal thermal imaging data, converting each thermal imaging data into a thermal imaging image, and obtaining the parts corresponding to rapid temperature changes in the converted thermal imaging image.

[0012] S30: Mark the area as a risk area and extract the feature changes of the thermal image of the risk area;

[0013] S40: Analyze the risk trend based on the aforementioned feature changes. The analysis method includes constructing a local thermal imaging trend and a global thermal imaging trend. The local thermal imaging trend involves selecting 3 to 5 observation points at the edge and center of the thermal imaging map corresponding to the aforementioned location, obtaining the color change between the observation points corresponding to the edge and the observation points corresponding to the center, and calculating the observation point with the deepest color change from the center to the edge based on the color change.

[0014] S50: Locate the observation point corresponding to the deepest color change, mark the observation point as a risk observation point, and verify whether there is a pipeline leak at the location. The verification method includes performing relevant calculations after the location is located. The relevant calculations include determining the color area corresponding to the risk observation point in the thermal image and calculating the diffusion rate of the color area towards the observation point at the center.

[0015] S60: A safety threshold for diffusion rate is preset. If the diffusion rate exceeds the safety threshold, it is determined that a leak has occurred at the risk location; otherwise, if the diffusion rate does not exceed the safety threshold, it is determined that the oil and gas transport at the risk location is abnormal.

[0016] S70: If it is determined that the oil and gas transport in the risk area is abnormal, then a global thermal imaging trend analysis is performed.

[0017] In a preferred embodiment of the present invention, in step S70, the analysis of global thermal imaging trends includes the following steps:

[0018] After identifying the location of the oil and gas pipeline corresponding to each thermal imaging data in the normal thermal imaging data, the number of the locations is counted.

[0019] Obtain the distance between the risk area and other areas, and based on the distance, obtain the 5 to 10 areas closest to the risk area, and mark the 5 to 10 areas as reference group areas;

[0020] When an abnormality in oil and gas transport is determined at the observation point, the correlation effect of the color area of ​​the risk observation point on each part of the reference group is obtained when the color area of ​​the risk observation point diffuses. The correlation effect includes whether the color of the thermal image corresponding to each part of the reference group changes to a darker trend when the color area of ​​the risk observation point diffuses.

[0021] If the color changes towards a darker color, the corresponding part in the reference group is considered to have abnormal oil and gas transport; if the color changes towards a lighter color, or if there is no color change, the corresponding part in the reference group is considered to have normal oil and gas transport.

[0022] In a preferred embodiment of the present invention, in step S30, feature changes in the thermal imaging image of the risk area are extracted, and the extraction steps are as follows:

[0023] The thermal image is preprocessed, including denoising and image enhancement.

[0024] Region segmentation is performed, including segmentation using a threshold method, which sets a threshold based on the temperature difference between the risk area and the surrounding normal area, and segments the risk area from the entire thermal image.

[0025] Feature extraction is performed, including temperature feature extraction, which includes calculating the average temperature value of all pixels within the risk area and recording the changes in the average temperature value at different times.

[0026] Feature extraction also includes texture feature extraction, which involves calculating the gray-level co-occurrence matrix (GLCM) of pixels within the risk area in different directions and extracting relevant feature parameters from the GLCM. These relevant feature parameters include energy, contrast, correlation, and entropy.

[0027] The extracted temperature and texture features are organized according to the time series, and feature change curves are plotted.

[0028] A correlation analysis is performed on the extracted temperature features and texture features;

[0029] The results of the correlation analysis are recorded and stored to establish a feature change database.

[0030] In a preferred embodiment of the present invention, in step S40, the observation point at the center with the deepest color change towards the observation point at the edge is calculated based on the color change, using the following formula:

[0031] ;

[0032] In the formula, This indicates the color difference between the observation point at the center and the observation point at the edge; the larger the value, the deeper the color change.

[0033] , , These represent the red, green, and blue components of the color at the observation point at the center in the RGB color space, respectively.

[0034] , , These represent the red, green, and blue components of the color at the observation point at the edge in the RGB color space, respectively.

[0035] In a preferred embodiment of the present invention, the following formula is also included:

[0036] ;

[0037] In the formula, This represents the hue difference between the observation point at the center and the observation point at the edge, with a value range of 0 to 180. The larger the value, the more obvious the hue change.

[0038] This represents the hue value of the observation point at the center, with a range of 0 to 360.

[0039] This represents the hue value of the observation point at the edge, with a value range of 0 to 360.

[0040] In a preferred embodiment of the present invention, in step S50, the diffusion rate of the color region toward the observation point at the center is calculated using the following formula:

[0041] ;

[0042] In the formula, This represents the average diffusion rate of the colored area toward the observation point at the center.

[0043] Indicates the number of points selected at the edge of the colored area;

[0044] The first one represents the edge of the color area The difference between the color value of each point and the color value of the observation point at the center;

[0045] The first one represents the edge of the color area The distance from each point to the observation point at the center.

[0046] In a preferred embodiment of the present invention, according to the calculation results, when the oil and gas transport at the risk location is determined to be abnormal, if the diffusion rate of the color area of ​​any observation point other than the risk observation point toward the observation point at the center does not exceed the safety threshold, then the oil and gas transport at the risk location is determined to be of a low degree of abnormality; otherwise, no determination is made.

[0047] In a preferred embodiment of the present invention, in the temperature feature extraction, the average temperature value of all pixels within the risk area is calculated according to the following formula:

[0048] ;

[0049] In the formula, This represents the average temperature value of all pixels within the risk area;

[0050] This represents the total number of pixels within the risk area.

[0051] Indicates the risk area and the first The temperature value corresponding to each pixel.

[0052] On the other hand, the present invention provides a system for an autonomous patrol drone inspection method for oil and gas pipeline networks as described above, comprising:

[0053] The data acquisition module is used to acquire relevant data about the target oil and gas pipeline. The relevant data includes normal thermal imaging data about the target oil and gas pipeline acquired by the target UAV, and the acquisition altitude corresponding to the acquisition altitude when the target UAV acquires the normal thermal imaging data.

[0054] The data processing module is used to process the normal thermal imaging data. The processing includes distinguishing the parts of the oil and gas pipelines corresponding to each thermal imaging data in the normal thermal imaging data, converting each thermal imaging data into a thermal imaging image, and obtaining the parts corresponding to rapid temperature changes in the converted thermal imaging image.

[0055] The feature extraction module is used to mark the location as a risk location and extract feature changes from the thermal image of the risk location.

[0056] A data fusion processing module, comprising an analysis unit, a calculation unit, and a judgment unit;

[0057] The analysis unit is used to analyze risk trends based on the feature changes. The analysis method includes constructing local thermal imaging trends and global thermal imaging trends. The local thermal imaging trend is to select 3 to 5 observation points at the edge and center of the thermal imaging map corresponding to the location, obtain the color changes between the observation points corresponding to the edge and the observation points corresponding to the center, and calculate the observation point with the darkest color change from the center to the observation point at the edge based on the color changes.

[0058] The calculation unit is used to locate the observation point corresponding to the deepest color change, mark the observation point as a risk observation point, and verify whether there is a pipeline leak at the location. The verification method includes performing relevant calculations after the location is located. The relevant calculations include determining the color area corresponding to the risk observation point in the thermal image and calculating the diffusion rate of the color area towards the observation point at the center.

[0059] The judgment unit is used to preset a safety threshold for diffusion rate. If the diffusion rate exceeds the safety threshold, it is determined that a leak has occurred at the risk location. Conversely, if the diffusion rate does not exceed the safety threshold, it is determined that the oil and gas transport at the risk location is abnormal. When the oil and gas transport at the risk location is determined to be abnormal, a global thermal imaging trend analysis is performed.

[0060] Beneficial effects:

[0061] 1. By processing normal thermal imaging data, different parts of oil and gas pipelines are distinguished and converted into thermal images, which can accurately identify parts with rapid temperature changes. These parts are often potential risk points. At the same time, the identified risk points are marked to facilitate subsequent key monitoring and analysis, thereby improving the accuracy of risk identification.

[0062] 2. By constructing local thermal imaging trends and global thermal imaging trends, risk trends are analyzed from both microscopic and macroscopic levels. Local analysis focuses on temperature changes within the risk area, while global analysis considers the impact of the risk area on the surrounding area, thus achieving a comprehensive risk assessment.

[0063] 3. By selecting observation points in the thermal imaging image, analyzing the color changes at the center and edges, and calculating the observation point with the deepest color change, the risk observation point can be located, providing a basis for risk verification. At the same time, by calculating the diffusion rate of the color area towards the observation point at the center, the temperature change rate of the risk area can be quantified, providing a scientific basis for judging whether the pipeline is leaking. Attached Figure Description

[0064] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0065] Figure 1 This is a schematic diagram of the modular structure of an autonomous patrol drone inspection system for oil and gas pipeline networks according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the method flow according to an embodiment of the present invention; The diagram is labeled as follows: 110 - Data acquisition module; 120 - Data processing module; 130 - Feature extraction module; 140 - Data fusion processing module; 1401 - Analysis unit; 1402 - Calculation unit; 1403 - Judgment unit. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0067] Because existing detection algorithms have weak generalization ability in unknown scenarios, the reliability of determining whether oil and gas pipelines are leaking is affected.

[0068] Based on this, the present invention proposes an autonomous patrol drone inspection method and system for oil and gas pipeline networks. Through data acquisition, accurate identification of risky locations, analysis of risk trends, verification of pipeline leaks, assessment of oil and gas transportation status, and feature extraction and database establishment, it improves the efficiency and accuracy of oil and gas pipeline network inspection and provides strong protection for the safe operation of oil and gas pipeline networks.

[0069] The present solution will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0070] Reference Figures 1 to 2 As one embodiment of the present invention, this embodiment provides an autonomous patrol drone inspection method for oil and gas pipeline networks, comprising the following steps:

[0071] S10: Acquire relevant data about the target oil and gas pipeline. The relevant data includes normal thermal imaging data about the target oil and gas pipeline acquired by the target UAV, as well as the acquisition altitude corresponding to the acquisition altitude when the target UAV acquires the normal thermal imaging data.

[0072] The drone carries thermal imaging equipment to acquire normal thermal imaging data and corresponding altitude information of the target oil and gas pipeline.

[0073] Record the altitude to eliminate thermal imaging errors caused by differences in flight altitude and improve data consistency;

[0074] S20: Process the normal thermal imaging data. The processing includes distinguishing the parts of the oil and gas pipelines corresponding to each thermal imaging data in the normal thermal imaging data, converting each thermal imaging data into a thermal imaging image, and obtaining the parts corresponding to rapid temperature changes in the converted thermal imaging image.

[0075] Distinguish the pipe parts corresponding to the thermal imaging data, convert them into thermal images, and identify areas with rapid temperature changes as potential risk areas;

[0076] It should be noted that the process of converting thermal imaging data into thermal images includes data preprocessing, pseudo-color mapping, and image generation. This technique of converting thermal imaging data into thermal images is well known to those skilled in the art and will not be described in detail here.

[0077] The term "rapid temperature change" refers to a temperature change cycle of 10 seconds. If the temperature changes more than twice within this cycle, it indicates that the temperature is changing rapidly.

[0078] S30: Mark the area as a risk area, and extract the feature changes of the thermal image of the risk area. The extraction steps are as follows:

[0079] The thermal image is preprocessed, including noise reduction and image enhancement.

[0080] Region segmentation is performed, including segmentation using a threshold method, which sets a threshold based on the temperature difference between the risk area and the surrounding normal area, and segments the risk area from the entire thermal image.

[0081] Feature extraction is performed, including temperature feature extraction. Temperature feature extraction involves calculating the average temperature value of all pixels within the risk area and recording the changes in the average temperature value at different times (e.g., during multiple drone inspections). By analyzing the trend of the average temperature change, it is possible to preliminarily determine whether there is an abnormal increase or decrease in temperature at the risk area, and the rate of temperature change. For example, if the average temperature rises continuously in a short period of time, it may indicate a risk of leakage or overheating at that location.

[0082] Feature extraction also includes texture feature extraction, which involves calculating the gray-level co-occurrence matrix (GLCM) of pixels within the risk area in different directions (such as 0°, 45°, 90°, 135°) and extracting relevant feature parameters from the GLCM. These relevant feature parameters include energy, contrast, correlation, and entropy.

[0083] These feature parameters can reflect information such as texture roughness and regularity in thermal images. For example, the energy feature value reflects the uniformity of the image texture; the higher the energy value, the more uniform the image texture. The contrast feature value reflects the magnitude of grayscale contrast in the thermal image; the higher the contrast, the clearer the texture of the thermal image. By analyzing changes in these texture feature parameters, it is possible to detect whether abnormal changes have occurred in the surface texture of risky areas, thereby determining whether there are problems such as corrosion or wear in oil and gas pipelines.

[0084] The extracted temperature and texture features are organized according to the time series, and feature change curves are plotted.

[0085] By observing these curves and analyzing the changing trends of each feature over time, it is possible to determine whether there are any abnormal change patterns. For example, if the temperature feature curve shows a continuous upward trend and the rate of increase exceeds the normal range; or if the area parameter in the shape feature suddenly increases, these may indicate potential problems in the risk area.

[0086] Correlation analysis was performed on the extracted temperature and texture features;

[0087] For example, by analyzing whether the texture of the risk area remains uniform and clear or undergoes significant changes when the temperature rises abnormally, correlation analysis can provide a more comprehensive understanding of the characteristic change mechanism of the risk area and improve the accuracy of fault diagnosis.

[0088] The results of the correlation analysis are recorded and stored to establish a feature change database.

[0089] S40: Analyze risk trends based on feature changes. The analysis methods include constructing local thermal imaging trends and global thermal imaging trends. For local thermal imaging trends, select 3 to 5 observation points at the edge and center of the thermal imaging map corresponding to the location. Among the observation points corresponding to the center, obtain the color changes between the observation points corresponding to the edge. Calculate the observation point at the center with the deepest color change towards the observation point at the edge based on the color changes.

[0090] In this embodiment, the color with the deepest change is the reddest color, such as rainbow or iron red.

[0091] Construct local thermal imaging trends (color changes at observation points) and global thermal imaging trends (the impact of risky areas on the surrounding areas);

[0092] Local analysis focuses on changes within the risk area, while global analysis assesses the chain reaction, thus achieving a comprehensive risk assessment.

[0093] S50: Locate the observation point corresponding to the deepest color change, mark the observation point as a risk observation point, and verify whether there is a pipeline leak at the location. The verification method includes performing relevant calculations after the location is located. The relevant calculations include determining the color area corresponding to the risk observation point in the thermal image and calculating the diffusion rate of the color area towards the observation point at the center.

[0094] S60: A preset safety threshold for diffusion rate is used. If the diffusion rate exceeds the safety threshold, a leak is determined to have occurred at the risk location. Conversely, if the diffusion rate does not exceed the safety threshold, an abnormality in oil and gas transport at the risk location is determined.

[0095] The observation point with the most pronounced color change was identified as the risk observation point, and the diffusion rate of the color region was calculated to verify the leak.

[0096] By calculating the diffusion rate, we can distinguish between actual leaks and transport anomalies, thus reducing false alarms.

[0097] It should be noted that those skilled in the art are well aware that the setting of this safety threshold is flexible and should be set or adjusted according to factors such as pipe material and operating pressure.

[0098] For example, different materials (such as steel pipes and plastic pipes) have different thermal conductivity coefficients, which leads to temperature anomalies and differences in diffusion rates. Metal pipes conduct electricity quickly, so the diffusion rate threshold should be higher; plastic pipes conduct electricity slowly, so the threshold can be appropriately lowered.

[0099] Furthermore, when high-pressure pipelines leak, the medium ejects at a faster speed and diffuses at a higher rate, requiring a higher threshold to avoid false alarms; while the opposite is true for low-pressure pipelines.

[0100] S70: If abnormal oil and gas transport is determined in a risky area, a global thermal imaging trend analysis is performed. The analysis steps include:

[0101] After identifying the location of the oil and gas pipeline corresponding to each thermal imaging data in the normal thermal imaging data, the number of locations is counted.

[0102] Obtain the distance between the risk area and other areas, and based on the distance, identify the 5 to 10 areas closest to the risk area, and mark these 5 to 10 areas as reference group areas;

[0103] When an abnormality in oil and gas transport is determined at an observation point, the correlation between the color area of ​​the risk observation point and the corresponding parts in the reference group is obtained as the color area of ​​the risk observation point diffuses. The correlation includes whether the color of the thermal image corresponding to each part in the reference group changes to a darker color as the color area of ​​the risk observation point diffuses.

[0104] If the color changes towards a darker color, the corresponding part in the reference group is considered to have abnormal oil and gas transport; if the color changes towards a lighter color, or if there is no color change, the corresponding part in the reference group is considered to have normal oil and gas transport.

[0105] Based on the above, in S40, the observation point at the center with the deepest color change towards the edge is calculated according to the following formula:

[0106] ;

[0107] In the formula, This indicates the color difference between the observation point at the center and the observation point at the edge; the larger the value, the deeper the color change.

[0108] , , These represent the red, green, and blue components of the color at the observation point at the center in the RGB color space, respectively.

[0109] , , These represent the red, green, and blue components of the color at the observation point at the edge in the RGB color space, respectively.

[0110] Explanation of the calculation steps for the formula in this embodiment:

[0111] Obtain the RGB component values ​​of the center observation point and the edge observation point respectively;

[0112] Substitute the corresponding component values ​​into the formula to calculate the sum of squares of the differences between the three components;

[0113] The color difference is obtained by taking the square root and the square root of the square. ;

[0114] For each observation point at the center, calculate its relationship with each observation point at the edge. Find the value that is the largest. The observation point at the edge corresponding to the value has the deepest color change compared to the observation point at the center.

[0115] Furthermore, it also includes calculations based on the following formula:

[0116] ;

[0117] In the formula, This represents the hue difference between the observation point at the center and the observation point at the edge, with a value range of 0 to 180. The larger the value, the more obvious the hue change.

[0118] This represents the hue value of the observation point at the center, with a range of 0 to 360.

[0119] This represents the hue value of the observation point at the edge, with a range of 0 to 360.

[0120] Explanation of the calculation steps for the formula in this embodiment:

[0121] Obtain the hue values ​​of the center observation point and the edge observation points. and ;

[0122] calculate If the result is greater than 180, subtract the result from 360 to get the result. ;

[0123] For each observation point at the center, calculate its relationship with each observation point at the edge. Find the value that is the largest. The observation point at the edge corresponding to the value has the deepest hue difference from the observation point at the center;

[0124] It should be noted that the two formulas above are used to calculate the color difference and hue difference between the observation points at the center and the edge, respectively. The connection between the two is that they are used together to accurately locate the risk observation points and improve the accuracy of leakage judgment through multi-dimensional color feature analysis.

[0125] The first formula, in the RGB color space, calculates the difference between the red, green, and blue components of the observation point at the center and the observation point at the edge, and quantifies the depth of color change using Euclidean distance; it is used to identify the area with the most significant temperature anomaly in the thermal image (the location with the deepest color change), providing a quantitative basis for the preliminary screening of risk observation points.

[0126] The second formula calculates the hue difference between the observation points at the center and the edge in the hue component of the HSV color space, and takes the minimum value to handle the cyclical characteristics of hue (such as 0° and 360° being the same).

[0127] The second formula supplements the shortcomings of RGB color differences, especially in low contrast or when colors are similar (such as light-colored areas), where hue differences can capture changes more sensitively; at the same time, it avoids the influence of RGB value fluctuations caused by lighting or material reflection on the results.

[0128] First, the candidate points with the most significant color changes are screened using the RGB difference formula, and then their significance is verified using the hue difference formula to avoid misjudgment caused by a single indicator.

[0129] Combining both methods allows for a more accurate determination of the boundaries of temperature diffusion. For example, RGB differences indicate the diffusion range, while hue differences confirm whether the diffusion direction is consistent.

[0130] In S50, the diffusion rate of the colored area toward the observation point at the center is calculated using the following formula:

[0131] ;

[0132] In the formula, This represents the average diffusion rate of a colored area toward the observation point at its center (the unit depends on the color value and the distance unit; for example, if the color value is a grayscale value (dimensionless) and the distance unit is pixels, then the rate unit is "grayscale value / pixel").

[0133] Indicates the number of points selected at the edge of the colored area;

[0134] The first one represents the edge of the color area The difference between the color value of each point and the color value of the observation point at the center;

[0135] The first one represents the edge of the color area The distance from each point to the observation point at the center (the unit is related to the unit of the average diffusion rate, such as a pixel).

[0136] Explanation of the calculation steps for the formula in this embodiment:

[0137] Determine the color value of the observation point at the center;

[0138] Select at the edge of the color area There are several points, and the distance from each point to the observation point at the center is measured. Simultaneously, the color value of each point is obtained;

[0139] Will , , Substitute into the formula to calculate the average diffusion rate .

[0140] Based on the above, according to the calculation results, when an abnormality in oil and gas transport is determined in a risky location, if the diffusion rate of the colored area of ​​any observation point other than the risk observation point towards the observation point at the center does not exceed the safety threshold, then the oil and gas transport in the risky location is determined to be of a low degree of abnormality; otherwise, it is not determined.

[0141] In temperature feature extraction, the average temperature value of all pixels within the risk area is calculated using the following formula:

[0142] ;

[0143] In the formula, This represents the average temperature value of all pixels within the risk area.

[0144] This represents the total number of pixels within the risk area.

[0145] Indicates the risk area and the first The temperature value corresponding to each pixel;

[0146] Explanation of the calculation steps for the formula in this embodiment:

[0147] Determine the total number of pixels within the risk area. ;

[0148] Obtain each pixel separately Corresponding temperature value ;

[0149] Will and Substitute into the formula to calculate the average temperature value .

[0150] Based on the above, this application improves the efficiency and accuracy of oil and gas pipeline network inspection by acquiring data, accurately identifying risk locations, analyzing risk trends, verifying pipeline leaks, assessing oil and gas transportation status, and extracting features and establishing a database, thus providing a strong guarantee for the safe operation of oil and gas pipeline networks.

[0151] This embodiment, in conjunction with the above-described autonomous patrol UAV inspection method for oil and gas pipeline networks, also proposes a working system applied to this method, as follows:

[0152] The data acquisition module 110 is used to acquire relevant data about the target oil and gas pipeline. The relevant data includes normal thermal imaging data about the target oil and gas pipeline acquired by the target UAV, and the acquisition altitude corresponding to the acquisition altitude when the target UAV acquires the normal thermal imaging data.

[0153] The data processing module 120 is used to process normal thermal imaging data. The processing includes distinguishing the parts of oil and gas pipelines corresponding to each thermal imaging data in the normal thermal imaging data, converting each thermal imaging data into a thermal imaging image, and obtaining the parts corresponding to rapid temperature changes in the converted thermal imaging image.

[0154] Feature extraction module 130 is used to mark the parts as risk parts and extract the feature changes of the thermal image of the risk parts;

[0155] The data fusion processing module 140 includes an analysis unit 1401, a calculation unit 1402, and a judgment unit 1403.

[0156] Analysis unit 1401 is used to analyze risk trends based on feature changes. The analysis methods include constructing local thermal imaging trends and global thermal imaging trends. Local thermal imaging trends involve selecting 3 to 5 observation points at the edge and center of the thermal imaging map corresponding to the location, obtaining the color changes between the observation points corresponding to the edge and the center, and calculating the darkest color change from the observation point at the center to the observation point at the edge based on the color changes.

[0157] The calculation unit 1402 is used to locate the observation point corresponding to the deepest color change, mark the observation point as a risk observation point, and verify whether there is a pipeline leak at the location. The verification method includes performing relevant calculations after the location is located. The relevant calculations include determining the color area corresponding to the risk observation point in the thermal image and calculating the diffusion rate of the color area towards the observation point at the center.

[0158] The judgment unit 1403 is used to preset a safety threshold for diffusion rate. If the diffusion rate exceeds the safety threshold, it is determined that a leak has occurred at the risk location. Conversely, if the diffusion rate does not exceed the safety threshold, it is determined that the oil and gas transport at the risk location is abnormal. When the oil and gas transport at the risk location is determined to be abnormal, a global thermal imaging trend analysis is performed.

[0159] In summary, this invention improves the efficiency and accuracy of oil and gas pipeline network inspections by acquiring data, accurately identifying risk locations, analyzing risk trends, verifying pipeline leaks, assessing oil and gas transportation status, and extracting features and establishing a database, thus providing a strong guarantee for the safe operation of oil and gas pipeline networks.

[0160] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for autonomous patrol drone inspection of oil and gas pipeline networks, characterized in that, Includes the following steps: S10: Acquire relevant data about the target oil and gas pipeline, including normal thermal imaging data about the target oil and gas pipeline acquired by the target UAV, and the acquisition altitude corresponding to the acquisition altitude when the target UAV acquires the normal thermal imaging data. S20: Process the normal thermal imaging data. The processing includes distinguishing the parts of the oil and gas pipelines corresponding to each thermal imaging data in the normal thermal imaging data, converting each thermal imaging data into a thermal imaging image, and obtaining the parts corresponding to rapid temperature changes in the converted thermal imaging image. S30: Mark the area as a risk area and extract the feature changes of the thermal image of the risk area; S40: Analyze risk trends based on the aforementioned feature changes. The analysis methods include constructing local thermal imaging trends and global thermal imaging trends. The local thermal imaging trend is to select 3 to 5 observation points at the edge and center of the thermal imaging map corresponding to the location, obtain the color change between the observation points corresponding to the edge and the observation points corresponding to the center, and calculate the observation point with the darkest color change from the center to the observation point at the edge based on the color change. S50: Locate the observation point corresponding to the deepest color change, mark the observation point as a risk observation point, and verify whether there is a pipeline leak at the location. The verification method includes performing relevant calculations after the location is located. The relevant calculations include determining the color area corresponding to the risk observation point in the thermal image and calculating the diffusion rate of the color area towards the observation point at the center. S60: A preset safety threshold for diffusion rate is established. If the diffusion rate exceeds the safety threshold, it is determined that a leak has occurred at the risk location. Conversely, if the diffusion rate does not exceed the safety threshold, the oil and gas transport at the risk location is determined to be abnormal. S70: If it is determined that the oil and gas transport in the risk area is abnormal, then a global thermal imaging trend analysis is performed.

2. The autonomous patrol drone inspection method for oil and gas pipeline networks as described in claim 1, characterized in that, In step S70, a global thermal imaging trend analysis is performed, including the following steps: After identifying the location of the oil and gas pipeline corresponding to each thermal imaging data in the normal thermal imaging data, the number of the locations is counted. Obtain the distance between the risk area and other areas, and based on the distance, obtain the 5 to 10 areas closest to the risk area, and mark the 5 to 10 areas as reference group areas; When an abnormality in oil and gas transport is determined at the observation point, the correlation effect of the color area of ​​the risk observation point on each part of the reference group is obtained when the color area of ​​the risk observation point diffuses. The correlation effect includes whether the color of the thermal image corresponding to each part of the reference group changes to a darker trend when the color area of ​​the risk observation point diffuses. If the color changes towards a darker color, the corresponding part in the reference group is considered to have abnormal oil and gas transport; if the color changes towards a lighter color, or if there is no color change, the corresponding part in the reference group is considered to have normal oil and gas transport.

3. The autonomous patrol drone inspection method for oil and gas pipeline networks as described in claim 1, characterized in that, In step S30, feature changes in the thermal image of the risk area are extracted. The extraction steps are as follows: The thermal image is preprocessed, including denoising and image enhancement. Region segmentation is performed, including segmentation using a threshold method, which sets a threshold based on the temperature difference between the risk area and the surrounding normal area, and segments the risk area from the entire thermal image. Feature extraction is performed, including temperature feature extraction, which includes calculating the average temperature value of all pixels within the risk area and recording the changes in the average temperature value at different times. Feature extraction also includes texture feature extraction, which involves calculating the gray-level co-occurrence matrix (GLCM) of pixels within the risk area in different directions and extracting relevant feature parameters from the GLCM. These relevant feature parameters include energy, contrast, correlation, and entropy. The extracted temperature and texture features are organized according to the time series, and feature change curves are plotted. A correlation analysis is performed on the extracted temperature features and texture features; The results of the correlation analysis are recorded and stored to establish a feature change database.

4. The autonomous patrol drone inspection method for oil and gas pipeline networks as described in claim 1, characterized in that, In step S40, the observation point at the center with the deepest color change towards the edge is calculated based on the color change, using the following formula: ； In the formula, This indicates the color difference between the observation point at the center and the observation point at the edge; the larger the value, the deeper the color change. , , These represent the red, green, and blue components of the color at the observation point at the center in the RGB color space, respectively. , , These represent the red, green, and blue components of the color at the observation point at the edge in the RGB color space, respectively.

5. The autonomous patrol drone inspection method for oil and gas pipeline networks as described in claim 4, characterized in that, It also includes calculations based on the following formula: ； In the formula, This represents the hue difference between the observation point at the center and the observation point at the edge, with a value range of 0 to 180. The larger the value, the more obvious the hue change. This represents the hue value of the observation point at the center, with a range of 0 to 360. This represents the hue value of the observation point at the edge, with a value range of 0 to 360.

6. The autonomous patrol drone inspection method for oil and gas pipeline networks as described in claim 1, characterized in that, In step S50, the diffusion rate of the color region toward the observation point at the center is calculated using the following formula: ； In the formula, This represents the average diffusion rate of the colored area toward the observation point at the center. Indicates the number of points selected at the edge of the colored area; The first one represents the edge of the color area The difference between the color value of each point and the color value of the observation point at the center; The first one represents the edge of the color area The distance from each point to the observation point at the center.

7. A method for autonomous patrol drone inspection of oil and gas pipeline networks as described in any one of claims 5 to 6, characterized in that, Based on the calculation results, when the oil and gas transport in the risky location is determined to be abnormal, if the diffusion rate of the color area of ​​any observation point other than the risk observation point towards the observation point at the center does not exceed the safety threshold, then the oil and gas transport in the risky location is determined to be of a low degree of abnormality; otherwise, no determination is made.

8. The autonomous patrol drone inspection method for oil and gas pipeline networks as described in claim 3, characterized in that, In temperature feature extraction, the average temperature value of all pixels within the risk area is calculated using the following formula: ； In the formula, This represents the average temperature value of all pixels within the risk area; This represents the total number of pixels within the risk area. Indicates the risk area and the first The temperature value corresponding to each pixel.

9. A system applied to the autonomous patrol drone inspection method for oil and gas pipeline networks as described in claim 1, characterized in that, include: The data acquisition module is used to acquire relevant data about the target oil and gas pipeline. The relevant data includes normal thermal imaging data about the target oil and gas pipeline acquired by the target UAV, and the acquisition altitude corresponding to the acquisition altitude when the target UAV acquires the normal thermal imaging data. The data processing module is used to process the normal thermal imaging data. The processing includes distinguishing the parts of the oil and gas pipelines corresponding to each thermal imaging data in the normal thermal imaging data, converting each thermal imaging data into a thermal imaging image, and obtaining the parts corresponding to rapid temperature changes in the converted thermal imaging image. The feature extraction module is used to mark the location as a risk location and extract feature changes from the thermal image of the risk location. A data fusion processing module, comprising an analysis unit, a calculation unit, and a judgment unit; The analysis unit is used to analyze risk trends based on the changes in the features. The analysis methods include constructing local thermal imaging trends and global thermal imaging trends. The local thermal imaging trend is to select 3 to 5 observation points at the edge and center of the thermal imaging map corresponding to the location, obtain the color change between the observation points corresponding to the edge and the observation points corresponding to the center, and calculate the observation point with the darkest color change from the center to the observation point at the edge based on the color change. The calculation unit is used to locate the observation point corresponding to the deepest color change, mark the observation point as a risk observation point, and verify whether there is a pipeline leak at the location. The verification method includes performing relevant calculations after the location is located. The relevant calculations include determining the color area corresponding to the risk observation point in the thermal image and calculating the diffusion rate of the color area towards the observation point at the center. The judgment unit is used to preset a safety threshold for the diffusion rate. If the diffusion rate exceeds the safety threshold, it is determined that a leak has occurred at the risk location. Conversely, if the diffusion rate does not exceed the safety threshold, the oil and gas transport at the risk location is determined to be abnormal. And when anomalies in oil and gas transport are detected in the risk areas, global thermal imaging trend analysis is performed.

Images