A method, system and platform for detecting equipment status in a gas pipeline network

Abstract

The present invention relates to a method, system, and platform for detecting the status of equipment within a gas pipeline network, comprising the following steps: obtaining instrument images of various devices within the gas pipeline network, wherein the instrument images are images including the dials of the instruments on the devices; for each instrument image, determining the dial reading in the instrument image based on the instrument image; for each instrument image, determining the operating status of the device corresponding to the instrument image based on the dial reading, wherein the operating status is either healthy or faulty; and for each device, detecting safety risks within the gas pipeline network based on the operating status. This method solves the problem that existing manual methods cannot detect safety hazards in gas pipeline networks in real time.

Description

Technical Field

[0001] The present invention relates to the field of gas pipeline network detection, and in particular to a method, system and platform for detecting the status of equipment in a gas pipeline network. Background Art

[0002] The gas pipeline network is a critical energy infrastructure. Gas pipeline network inspections are routine inspections of pipeline network facilities to promptly identify whether gas pipeline equipment is in good condition.

[0003] The inspection of gas pipeline network usually involves the inspection of equipment in the gas pipeline network, and the inspection of equipment in the gas pipeline network usually involves the inspection of the dial readings of the instruments on the equipment. If the dial readings are abnormal, it indicates that the equipment has failed, and then it is determined that the gas pipeline network has an abnormality. Since the inspection of instruments on the equipment is very boring and repetitive, especially field work is relatively difficult, and the gas pipeline network is in good condition in most cases, it is easy for on-site inspection personnel to be complacent and often fail to complete the inspection according to the prescribed procedures and standard content, resulting in inadequate inspections, which cannot fundamentally avoid potential accidents. Summary of the Invention

[0004] In order to overcome the problem that existing manual detection of potential safety hazards in gas pipeline networks cannot be performed in real time, the present invention provides a method, system and platform for detecting the status of equipment in a gas pipeline network.

[0005] In a first aspect, in order to solve the above technical problems, the present invention provides a method for detecting the status of equipment in a gas pipeline network, comprising the following steps:

[0006] Obtaining instrument images of various devices in the gas pipeline network, where the instrument images include dials of the instruments on the devices;

[0007] For each instrument image, determine the reading of the dial in the instrument image based on the instrument image;

[0008] For each instrument image, the operating status of the device corresponding to the instrument image is determined based on the dial reading. The operating status is either healthy or faulty.

[0009] For each device, the safety risks of the gas pipeline network are detected according to the operating status.

[0010] The beneficial effect of the device status detection method in the gas pipeline network provided by the present invention is: according to the instrument image of each device in the gas pipeline network, the dial reading in each instrument image is determined, and then the operating status of each device is determined according to the reading of each dial, so as to detect the safety risks of the gas pipeline network according to each operating status. This application avoids the need for manual on-site detection of each device, and the gas pipeline network can be detected in real time only by using the instrument image obtained in real time, which solves the problem that the existing manual detection of safety hazards of the gas pipeline network cannot be carried out in real time.

[0011] On the basis of the above technical solution, the method for detecting the status of equipment in a gas pipe network of the present invention can be further improved as follows.

[0012] Furthermore, in the above method, for each instrument image, determining the reading of the dial in the instrument image according to the instrument image specifically includes:

[0013] For each instrument image, determining connected body regions on the instrument image based on the instrument image, wherein each connected body region represents a region formed by each pixel point on the instrument image whose first pixel value is greater than a first threshold;

[0014] Determine the connected body area corresponding to each scale line on the instrument according to each connected body area, and use the connected body area corresponding to each scale line as the first target connected domain;

[0015] For each first target connected domain, determining a first center point of the first target connected domain;

[0016] For each watch dial, a first circle center is determined based on each first center point, and a first contour line corresponding to the watch dial is determined based on the first circle center and a preset radius, where the first circle center represents the center of the watch dial determined based on each first center point;

[0017] For each first target connected domain, determining a second pixel value corresponding to the first target connected domain, wherein the second pixel value represents the sum of the RGB values ​​of each pixel point on the first target connected domain;

[0018] The smallest pixel value among the second pixel values ​​is used as the target pixel value, and the position of the first target connected domain corresponding to the target pixel value is used as the pointer position; and according to the pointer position, a second contour line corresponding to the pointer position is determined;

[0019] Determining a first target center point from each first center point according to the first contour line and the second contour line;

[0020] For each dial, a reference point is determined according to a preset coordinate system. The preset coordinate system represents a coordinate system established with the center of the first circle as the origin, and the reference point represents the position corresponding to the zero scale on the dial;

[0021] For each dial, determine a first straight line between the reference point and the first center of the circle, and a second straight line between the first target center point and the first center of the circle. Based on the first straight line and the second straight line, determine the deflection angle of the second straight line relative to the first straight line. Based on the deflection angle and a preset relationship, determine the reading of the dial. The preset relationship represents the correspondence between the deflection angle and the reading of the dial.

[0022] The beneficial effect of adopting the above further scheme is: through the instrument image, each connected graph area on the instrument image is obtained, and the first target connected domain corresponding to the scale line is determined from each connected body area, and then the first center of the circle is determined according to the first center point of each first target connected domain, so as to determine the first contour line through the first center of the circle. At the same time, the position of the pointer is determined in each first target connected domain, and the second contour line corresponding to the pointer position is determined through the pointer position, so that the first target center point can be determined through the first contour line and the second contour line. Finally, the first straight line and the second straight line are determined according to the first target center point, so as to determine the deflection angle, and the dial reading can be determined through the deflection angle.

[0023] Furthermore, in the above method, determining the connected body areas corresponding to the respective scale lines on the instrument according to the respective connected body areas includes:

[0024] For each connected body region, the area of ​​the connected body region is obtained. If the area is greater than the second threshold and less than the third threshold, the connected body region is the connected body region corresponding to the scale line corresponding to the connected body region.

[0025] The beneficial effect of adopting the above further scheme is: through the area of ​​each connected body area, some interference areas with an area greater than the second threshold and interference areas with an area less than the third threshold are excluded, thereby determining the connected body area corresponding to the scale line corresponding to the connected body area.

[0026] Furthermore, the method further comprises:

[0027] For each first target connected domain, convert the first target connected domain into a second target connected domain in a rectangular shape;

[0028] For each second target connected domain, a third target connected domain is determined, wherein for each third target connected domain, the third target connected domain represents an area formed by each pixel point in the second target connected domain whose third pixel value is greater than a fourth threshold;

[0029] For each first target connected component, determining a first center point of the first target connected component includes:

[0030] For each third target connected component, a first center point of the third target connected component is determined.

[0031] The beneficial effects of adopting the above further scheme are: converting the first target connected domain into a second target connected domain in a rectangular shape, further eliminating the interference of other connected body areas of similar shapes, and converting the second target connected domain into a third target connected domain, eliminating the influence of shadows on the scale lines caused by uneven lighting.

[0032] Furthermore, in the above method, for each dial, determining the first circle center according to each first center point includes:

[0033] For each dial, determine the first circle center based on each first center point using the RANSAC model;

[0034] The training process of the RANSAC model specifically includes:

[0035] S1, obtaining a training set, where the training set includes multiple second center points and the coordinates corresponding to each second center point in a preset coordinate system;

[0036] S2: Input the training set into the initial model, train the initial model, and output multiple target points. For each target point, the target point is either an interior point or an exterior point. The interior point represents each second circle center and the corresponding coordinates of each second circle center in the preset coordinate system. The exterior point represents each non-second circle center and the corresponding coordinates of each non-second circle center in the preset coordinate system.

[0037] S3, determining a loss value according to a first number of inliers and a second number of outliers in the target point;

[0038] S4, if the loss value meets the preset end condition, the initial model when the preset end condition is met is used as the RANSAC model; if the loss value does not meet the preset end condition, the network parameters of the initial network are adjusted, and the initial model is retrained according to the adjusted network parameters until the loss value of the initial model meets the preset end condition;

[0039] Determining a loss value according to a first number of inliers and a second number of outliers in the target point includes:

[0040] According to the first number of inliers and the second number of outliers in the target point, the loss value is determined by a first formula, wherein the first formula is:

[0041]

[0042]

[0043] Among them, k represents the number of current iterations, n inliers Indicates the first number, n outliersRepresents the second number, t represents the proportion of inliers in all target points, p represents the loss value, and n represents the number of second center points in the training set.

[0044] The beneficial effect of adopting the above further solution is that the first circle center can be directly determined from each first center point through the RANSAC model, which is more efficient and convenient.

[0045] Furthermore, in the above method, determining the first target center point according to the first contour line and the second contour line includes:

[0046] Determining a first intersection point and a second intersection point according to the first contour line and the second contour line;

[0047] Determine a second target center point based on the first intersection point, where the second target center point is the first center point closest to the first intersection point in a straight line among the first center points;

[0048] Determine a third target center point based on the second intersection point; the third target center point is the first center point of each first center point that is closest to the second intersection point in a straight line;

[0049] The first center point between the second target center point and the third target center point is taken as the first target center point.

[0050] The beneficial effect of adopting the above further scheme is: the first intersection point and the second intersection point are determined according to the first contour line and the second contour line, and then the second target center point and the third target center point are determined according to the first intersection point and the second intersection point, and the first center point between the second target center point and the third target center point is used as the first target center point, that is, the position of the first target center point is the reading of the dial.

[0051] In a second aspect, the present invention provides a system for detecting the status of equipment in a gas pipeline network, comprising:

[0052] The first acquisition module is used to acquire an instrument image of each device in the gas pipeline network, where the instrument image is an image including a dial of an instrument on the device;

[0053] A second acquisition module is configured to determine, for each instrument image, a reading of a dial in the instrument image based on the instrument image;

[0054] The third acquisition module is used to determine the operating status of the device corresponding to each instrument image based on the reading of the dial, where the operating status is a healthy state or a fault state;

[0055] The detection module is used to detect the safety risks of the gas pipeline network for each device according to its operating status.

[0056] In a third aspect, the present invention further provides a gas pipe network equipment status detection platform, comprising: an intelligent patrol monitoring terminal, an intelligent patrol cloud, and a patrol front end, wherein the intelligent patrol monitoring terminal is connected to the intelligent patrol cloud, and the intelligent patrol cloud is connected to the patrol front end;

[0057] Inspection front-end, used to obtain instrument images of various devices in the gas pipeline network. The instrument images include images of the dials of the instruments on the devices;

[0058] The intelligent inspection cloud is used to determine the dial reading of each instrument image based on the instrument image. For each instrument image, the operating status of the device corresponding to the instrument image is determined based on the dial reading, and the operating status is determined as healthy or faulty.

[0059] The intelligent inspection and monitoring terminal is used to detect the safety risks of the gas pipeline network based on the operating status of each device.

[0060] In a fourth aspect, the present invention further provides an electronic device comprising a memory, a processor, and a program stored in the memory and running on the processor, wherein when the processor executes the program, the steps of the above-mentioned method for detecting the status of equipment in a gas pipeline network are implemented.

[0061] In a fifth aspect, the present invention also provides a computer-readable storage medium, which stores instructions. When the instructions are executed on a terminal device, the terminal device executes the steps of a method for detecting the status of equipment in a gas pipeline network as described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention is further described below with reference to the accompanying drawings and embodiments.

[0063] Figure 1 A flow chart of a method for detecting the status of equipment in a gas pipe network according to an embodiment of the present invention;

[0064] Figure 2 This is a structural diagram of a gas pipe network equipment status detection system according to an embodiment of the present invention;

[0065] Figure 3 This is a schematic structural diagram of a gas network equipment status detection platform according to an embodiment of the present invention;

[0066] Figure 4 This is a structural schematic diagram of a gas pipe network equipment status detection platform in another working state according to an embodiment of the present invention. DETAILED DESCRIPTION

[0067] The following examples are provided to further explain and supplement the present invention and do not constitute any limitation to the present invention.

[0068] The following describes a method, system, and platform for detecting the status of equipment in a gas pipeline network according to an embodiment of the present invention with reference to the accompanying drawings.

[0069] like Figure 1 As shown, a method for detecting the status of equipment in a gas pipe network according to an embodiment of the present invention includes the following steps:

[0070] Step 1: Obtain instrument images of various devices in the gas pipeline network. The instrument images are images including dials of instruments on the devices.

[0071] Optionally, before obtaining instrument images of each device in the gas pipeline network, it is necessary to develop an inspection management system, specifically including the following four aspects:

[0072] First, the configuration management of basic inspection data mainly includes:

[0073] Inspection system: inspection personnel, positions, and roles;

[0074] Inspection items: equipment name, gas pipeline location, and specific inspection content.

[0075] Second, develop periodic and customized inspection plans and goals for different equipment.

[0076] Third, according to the inspection plan, inspection task work orders are automatically generated on time and assigned to relevant personnel positions, so that inspection tasks are automatically sent to the terminals of corresponding inspection personnel and reminders are given. In special circumstances, special tasks can be quickly added. Special circumstances refer to situations that are not covered by the established inspection management system, such as temporary addition of inspection content to the equipment inspection items.

[0077] Fourth, detection status query: Based on the visual GIS system, the execution status of the detection task can be queried in real time, including the task completion progress, task safety risk prompts, whether there is any abnormal inspection information, etc., and whether the time and location of obtaining the instrument image are consistent with the inspection time and location.

[0078] In this embodiment, after the inspection management system is formulated, the safety risk detection of the gas pipeline network can be started and step one can be executed.

[0079] Step 2: For each instrument image, determine the dial reading in the instrument image according to the instrument image.

[0080] Optionally, for each instrument image, determining a dial reading in the instrument image according to the instrument image specifically includes:

[0081] S21, for each instrument image, determine each connected body region on the instrument image according to the instrument image, for each connected body region, the connected body region represents the region formed by each pixel point on the instrument image whose first pixel value is greater than a first threshold.

[0082] Optionally, the first threshold is selected according to actual conditions.

[0083] Optionally, before determining the connected body areas on the instrument image, the instrument image can be Gaussian smoothed to eliminate the interference of noise on the instrument image, and then features on the instrument image can be extracted using methods such as Hough transform and region growing segmentation algorithm to obtain a feature image corresponding to the instrument image. Finally, the connected body areas on the feature image are determined based on the first threshold.

[0084] S22 , determining the connected body areas corresponding to the respective scale lines on the meter according to the respective connected body areas, and taking the connected body areas corresponding to the respective scale lines as the first target connected domains.

[0085] Optionally, determining the connected body areas corresponding to the respective scale lines on the meter according to the respective connected body areas includes:

[0086] For each connected body region, the area of ​​the connected body region is obtained. If the area is greater than the second threshold and less than the third threshold, the connected body region is the connected body region corresponding to the scale line corresponding to the connected body region.

[0087] In this embodiment, since the scale lines are relatively thin short lines, the connected body area corresponding to the scale lines is determined by the area of ​​the connected body area. Based on this, the interference areas with an area of ​​the connected body area greater than the second threshold are excluded (such as the white background part of the dial), and the interference areas with an area of ​​the connected body area less than the third threshold are excluded (such as numbers).

[0088] S23 : For each first target connected component, determine a first center point of the first target connected component.

[0089] Optionally, the method further includes:

[0090] For each first target connected domain, convert the first target connected domain into a second target connected domain in a rectangular shape;

[0091] For each second target connected domain, a third target connected domain is determined, wherein for each third target connected domain, the third target connected domain represents an area formed by each pixel point in the second target connected domain whose third pixel value is greater than a fourth threshold;

[0092] For each first target connected component, determining a first center point of the first target connected component includes:

[0093] For each third target connected component, a first center point of the third target connected component is determined.

[0094] In this embodiment, the second threshold and the third threshold are selected according to actual conditions.

[0095] In this embodiment, a bilinear interpolation method may be used to transform the first target connected domain into a second target connected domain in a rectangular shape, thereby further eliminating interference from other connected body regions with shapes similar to the first target connected domain.

[0096] In this embodiment, since the scale lines on the dial are all vertically upward, various shadows will be generated when the lighting is uneven. In order to avoid the influence of the shadows, the third target connected domain is determined according to the fourth threshold.

[0097] S24, for each dial, determine the first center of the circle according to each first center point, and determine the first contour line corresponding to the dial according to the first center of the circle and the preset radius, where the first center of the circle represents the center of the dial determined according to each first center point.

[0098] Optionally, for each dial, determining the first circle center according to each first center point includes:

[0099] For each dial, determine the first circle center based on each first center point using the RANSAC model;

[0100] The training process of the RANSAC model specifically includes:

[0101] S1, obtaining a training set, where the training set includes multiple second center points and the coordinates corresponding to each second center point in a preset coordinate system;

[0102] S2: Input the training set into the initial model, train the initial model, and output multiple target points. For each target point, the target point is either an interior point or an exterior point. The interior point represents each second circle center and the corresponding coordinates of each second circle center in the preset coordinate system. The exterior point represents each non-second circle center and the corresponding coordinates of each non-second circle center in the preset coordinate system.

[0103] S3, determining a loss value according to a first number of inliers and a second number of outliers in the target point;

[0104] S4, if the loss value meets the preset end condition, the initial model when the preset end condition is met is used as the RANSAC model; if the loss value does not meet the preset end condition, the network parameters of the initial network are adjusted, and the initial model is retrained according to the adjusted network parameters until the loss value of the initial model meets the preset end condition;

[0105] Determining a loss value according to a first number of inliers and a second number of outliers in the target point includes:

[0106] According to the first number of inliers and the second number of outliers in the target point, the loss value is determined by a first formula, wherein the first formula is:

[0107]

[0108]

[0109] Among them, k represents the number of current iterations, n inliers Indicates the first number, n outliers Represents the second number, t represents the proportion of inliers in all target points, p represents the loss value, and n represents the number of second center points in the training set.

[0110] S25 , for each third target connected component, determine a second pixel value corresponding to the third target connected component, where each second pixel value represents the sum of RGB values ​​of each pixel point in the third target connected component.

[0111] S26 , taking the smallest pixel value among the second pixel values ​​as the target pixel value, and taking the position of the first target connected domain corresponding to the target pixel value as the pointer position; and determining the second contour line corresponding to the pointer position according to the pointer position.

[0112] In this embodiment, since the pointer is dark gray (black RGB = 0, 0, 0) when it is on the scale line, the second pixel value on the third target connected domain containing the pointer is smaller than the second pixel value on the third target connected domain not containing the pointer. Based on this, by comparing each second pixel value, the smallest second pixel value is selected as the target pixel value, so that the position of the first target connected domain corresponding to the target pixel value is used as the pointer position.

[0113] S27 , determining a first target center point from each first center point according to the first contour line and the second contour line.

[0114] Optionally, determining the first target center point according to the first contour line and the second contour line includes:

[0115] Determining a first intersection point and a second intersection point according to the first contour line and the second contour line;

[0116] Determine a second target center point based on the first intersection point, where the second target center point is the first center point closest to the first intersection point in a straight line among the first center points;

[0117] Determine a third target center point based on the second intersection point; the third target center point is the first center point of each first center point that is closest to the second intersection point in a straight line;

[0118] The first center point between the second target center point and the third target center point is taken as the first target center point.

[0119] S28, for each dial, determine a reference point according to a preset coordinate system, where the preset coordinate system represents a coordinate system established with the first center of the circle as the origin, and the reference point represents a position corresponding to the 0 scale on the dial.

[0120] S29. For each dial, determine the first straight line between the reference point and the first center of the circle, and the second straight line between the first target center point and the first center of the circle. Based on the first straight line and the second straight line, determine the deflection angle of the second straight line relative to the first straight line. Based on the deflection angle and the preset relationship, determine the reading of the dial. The preset relationship represents the correspondence between the deflection angle and the reading of the dial.

[0121] In this embodiment, the preset relationship includes the correspondence between each deflection angle and the dial reading. For example, a deflection angle of 30° indicates that the second straight line is deflected by 30° relative to the first straight line, that is, the pointer position is at a position deflected 30° clockwise from the 0 scale. At this time, the dial reading is 3 (the dial reading is +1 for every 10° interval).

[0122] Step 3: For each instrument image, determine the operating status of the device corresponding to the instrument image based on the dial reading. The operating status is either a healthy state or a fault state.

[0123] In this embodiment, after obtaining the dial reading, the operating status of the equipment can be determined through, for example, an intelligent monitoring model. The intelligent monitoring model integrates classic machine learning methods, such as linear regression algorithm, BP neural network algorithm, decision tree, support vector machine, KNN, and integrated learning methods such as XGboost, LGBM, Random Forest, and a deep convolutional neural network formed by an integrated learning method based on Stacking. It can establish a potential connection between the historical equipment dial reading and the operating status. Therefore, after the dial reading obtained in step three is input into the intelligent monitoring model, the operating status of the equipment can be predicted through the intelligent monitoring model.

[0124] Step 4: For each device, detect the safety risks of the gas pipeline network based on its operating status.

[0125] like Figure 2 As shown, a gas network equipment status detection system according to an embodiment of the present invention includes:

[0126] The first acquisition module 202 is used to acquire an instrument image of each device in the gas pipeline network, where the instrument image is an image including a dial of an instrument on the device;

[0127] The second acquisition module 203 is configured to determine, for each instrument image, a reading of a dial in the instrument image according to the instrument image;

[0128] The third acquisition module 204 is used to determine the operating status of the device corresponding to each instrument image based on the reading of the dial, where the operating status is a healthy state or a fault state;

[0129] The detection module 205 is used to detect the safety risk of the gas network for each device according to the operating status.

[0130] Optionally, the second obtaining module 203 includes:

[0131] A first determining module is configured to determine, for each instrument image, connected body regions on the instrument image based on the instrument image, wherein each connected body region represents a region formed by pixels on the instrument image having a first pixel value greater than a first threshold;

[0132] A second determining module is configured to determine, based on each connected body area, the connected body area corresponding to each scale line on the meter, and use the connected body area corresponding to each scale line as a first target connected domain;

[0133] A third determining module is configured to determine, for each first target connected domain, a first center point of the first target connected domain;

[0134] a fourth determining module, configured to determine, for each watch dial, a first circle center based on each first center point, and determine a first contour line corresponding to the watch dial based on the first circle center and a preset radius, wherein the first circle center represents the center of the watch dial determined based on each first center point;

[0135] a fifth determining module, for each first target connected domain, determining a second pixel value corresponding to the first target connected domain, wherein each second pixel value represents the sum of RGB values ​​of each pixel point on the first target connected domain;

[0136] a sixth determining module, configured to use the smallest pixel value among the second pixel values ​​as the target pixel value, and the position of the first target connected domain corresponding to the target pixel value as the pointer position; and determine the second contour line corresponding to the pointer position according to the pointer position;

[0137] A seventh determining module, configured to determine a first target center point from each first center point according to the first contour line and the second contour line;

[0138] an eighth determining module, configured to determine, for each dial, a reference point according to a preset coordinate system, where the preset coordinate system represents a coordinate system established with the first center of the circle as the origin, and the reference point represents a position corresponding to a zero mark on the dial;

[0139] The ninth determination module is used to determine, for each dial, a first straight line between the reference point and the first center of the circle, and a second straight line between the first target center point and the first center of the circle; based on the first straight line and the second straight line, determine a deflection angle of the second straight line relative to the first straight line; and based on the deflection angle and a preset relationship, determine a reading of the dial, wherein the preset relationship represents a correspondence between the deflection angle and the reading of the dial.

[0140] Optionally, the second determining module is specifically configured to:

[0141] For each connected body region, the area of ​​the connected body region is obtained. If the area is greater than the second threshold and less than the third threshold, the connected body region is the connected body region corresponding to the scale line corresponding to the connected body region.

[0142] Optionally, the system further includes:

[0143] A tenth determination module is configured to convert each first target connected domain into a second target connected domain in a rectangular shape, and determine a third target connected domain for each second target connected domain, wherein the third target connected domain represents an area formed by each pixel point in the second target connected domain whose third pixel value is greater than a fourth threshold.

[0144] The third determining module is specifically configured to:

[0145] For each third target connected component, a first center point of the third target connected component is determined.

[0146] Optionally, when determining the first center of the circle through the first unit, the fourth determining module is specifically configured to:

[0147] For each dial, determine the first circle center based on each first center point using the RANSAC model;

[0148] The training process of the above RANSAC model specifically includes:

[0149] S1, obtaining a training set, where the training set includes multiple second center points and the coordinates corresponding to each second center point in a preset coordinate system;

[0150] S2: Input the training set into the initial model, train the initial model, and output multiple target points. For each target point, the target point is either an interior point or an exterior point. The interior point represents each second circle center and the corresponding coordinates of each second circle center in the preset coordinate system. The exterior point represents each non-second circle center and the corresponding coordinates of each non-second circle center in the preset coordinate system.

[0151] S3, determining a loss value according to a first number of inliers and a second number of outliers in the target point;

[0152] S4, if the loss value meets the preset end condition, the initial model when the preset end condition is met is used as the RANSAC model; if the loss value does not meet the preset end condition, the network parameters of the initial network are adjusted, and the initial model is retrained according to the adjusted network parameters until the loss value of the initial model meets the preset end condition;

[0153] Determining a loss value according to a first number of inliers and a second number of outliers in the target point includes:

[0154] According to the first number of inliers and the second number of outliers in the target point, the loss value is determined by a first formula, wherein the first formula is:

[0155]

[0156]

[0157] Among them, k represents the number of current iterations, n inliers Indicates the first number, n outliers Represents the second number, t represents the proportion of inliers in all target points, p represents the loss value, and n represents the number of second center points in the training set.

[0158] Optionally, the seventh determination module determines the first target center point through the second unit, wherein the second unit is specifically configured to:

[0159] Determining a first intersection point and a second intersection point according to the first contour line and the second contour line;

[0160] Determine a second target center point based on the first intersection point, where the second target center point is the first center point closest to the first intersection point in a straight line among the first center points;

[0161] Determine a third target center point based on the second intersection point; the third target center point is the first center point of each first center point that is closest to the second intersection point in a straight line;

[0162] The first center point between the second target center point and the third target center point is taken as the first target center point.

[0163] like Figure 3 As shown, a gas pipe network equipment status detection system of this embodiment includes: an intelligent patrol monitoring terminal, an intelligent patrol cloud and a patrol front end, the intelligent patrol monitoring terminal is connected to the intelligent patrol cloud, and the intelligent patrol cloud is connected to the patrol front end, wherein:

[0164] The inspection front end is used to obtain instrument images of various devices in the gas pipeline network. The instrument image is an image containing the dials of the instruments on the equipment.

[0165] The intelligent inspection cloud is used to determine the dial reading of each instrument image based on the instrument image. For each instrument image, the operating status of the device corresponding to the instrument image is determined based on the dial reading, and the operating status is determined as healthy or faulty.

[0166] The intelligent inspection and monitoring terminal is used to detect the safety risks of the gas pipeline network based on the operating status of each device.

[0167] like Figure 3 As shown in the figure, the process between the inspection front-end and the intelligent inspection cloud mainly includes pushing tasks, collecting data, obtaining results, and determining results.

[0168] Push tasks refer to the established inspection management system. For example, if the inspection management system stipulates periodic inspection of device A, the intelligent inspection cloud will push this task to the inspection front-end, which will periodically obtain the instrument image of device A.

[0169] The collected data refers to the instrument image of device A collected by the inspection front end;

[0170] The acquisition result refers to whether the intelligent inspection cloud detects whether the instrument image of device A has been obtained. If so, a confirmation message is sent to the inspection front end;

[0171] Result confirmation refers to the feedback information that the inspection front-end feeds back to the intelligent inspection cloud based on the confirmation information.

[0172] The functions between the smart inspection cloud and the smart inspection monitoring terminal mainly include task control, status acquisition, and exception control, including:

[0173] Task control refers to the established inspection management system. For example, the inspection management system specifies periodic inspection of device A. This task is pushed to the intelligent inspection cloud, which then pushes it to the inspection front-end for execution.

[0174] Obtaining status means that the intelligent inspection cloud sends the operating status of device A to the intelligent inspection monitoring terminal;

[0175] Abnormal control refers to the operating status of device A fed back by the intelligent inspection cloud. If the operating status is a fault state, corresponding countermeasures need to be executed and fed back to the intelligent inspection cloud.

[0176] Optional, such as Figure 4 As shown, the intelligent inspection cloud also includes the following functions:

[0177] 1. The functions of inspection management, data processing, real-time inspection, exception handling and monitoring platform include the following:

[0178] Inspection anomaly registration: This system not only records the operating status of the equipment but also registers anomalies discovered during on-site inspections, including the appearance, time, degree, and duration of the anomaly. Major hidden dangers and accidents in gas pipeline networks include leaks, rust spots, pressure occupation, mechanical damage, exposed pipes, and line construction. By registering and counting these hidden dangers and accidents, the intelligent inspection cloud automatically calculates the probability of each event, identifies high-incidence events, customizes inspection content specifications, and scientifically arranges line inspection tasks, thereby more effectively reducing the possibility of accidents.

[0179] Abnormal handling: Equipment operators can manage equipment health based on inspection abnormality registration, such as launching emergency plans or inspection work orders in real time. By recording and tracking data throughout the entire process until the complete handling is completed and the loop is closed, the entire abnormality handling process can be visualized and reproducible on-site.

[0180] Inspection ledger management: This system manages the inspection work order results and enables effective query, analysis, and report output of the inspection ledger. By setting different time periods and equipment objects, intuitive curves are drawn for equipment-related parameters. By analyzing the relationship between the curves and the equipment's standard operating parameters, the operating status of the equipment can be understood. This provides decision makers with relevant information for analyzing equipment operation and enabling them to make timely and accurate decisions.

[0181] Inspection data statistics: This function enables statistical analysis of inspection results. By statistically analyzing and managing these data, the inherent patterns of equipment degradation or failure can be discovered, and equipment life cycle data can be gradually established and improved to guide equipment maintenance. Regular inspection statistical analysis reports can be generated.

[0182] Inspection performance management: Based on the execution of inspection tasks (such as the presence and completion rates of inspection personnel and equipment), automatic inspection KPI (Key Performance Indication) indicators are generated to conduct performance appraisals, thereby improving the efficiency of on-site inspection personnel.

[0183] 6. Data upload and data labeling: realize data interaction with the inspection front end and the intelligent inspection cloud.

[0184] 7. Model training, model testing, and model downloading: used for training, testing, and downloading RANSAC models and intelligent monitoring models.

[0185] 8. Identification model library and device knowledge base: used to store RANSAC models and intelligent monitoring models, as well as various types of device information.

[0186] Optionally, the inspection front end can use portable devices, such as intelligently controlled high-definition cameras, drones, robots, etc., which can be controlled through a smartphone APP to collect, upload and receive data in real time. At the same time, the inspection front end also includes the following functions:

[0187] 1. Receive inspection tasks and start the inspection process: According to the established inspection management system, the corresponding inspection personnel receive push tasks in real time.

[0188] 2. Automatic positioning: Automatically locate the location of the equipment currently being inspected, interactively determine the equipment that needs to be inspected within the current location range, and the inspection front end collects and reports data on the current equipment according to specifications, completes the inspection work, and waits for the intelligent inspection cloud and real-time inspection results and confirmation to complete the inspection task process of the current equipment.

[0189] 3. Data collection: Inspection personnel can use the operation interface provided by the inspection instrument (such as APP) to automatically collect data information through equipment coding, quickly enter it, or record videos / images according to regulations; or high-definition equipment can automatically collect images at a fixed time; the collected equipment-related dynamic data, including multimedia information such as photos, entered information, three-dimensional coordinates, on-site status and other data are transmitted to the cloud.

[0190] 4. Data pre-identification and processing: Depending on the functions and integration complexity of the inspection front-end, some basic data pre-identification and processing capabilities can be provided, such as automatic recognition of simple instrument data based on AI algorithms. The inspection front-end directly obtains the scale data shown in the instrument image, and the inspection personnel confirm and directly upload the result data.

[0191] 5. Data transmission: Data, videos and images collected on-site, real-time inspection information, pre-processing result data, etc. can be transmitted to the intelligent inspection cloud in real time through 4G / 5G / WiFi of handheld devices, or networks such as the Internet of Things; data will be automatically cached in areas without network signals and transmitted immediately after the network is connected.

[0192] 6. Real-time acquisition of inspection analysis results and on-site confirmation by inspection personnel: The intelligent inspection cloud pushes the real-time inspection results of the equipment to the terminal (user), and the user interactively confirms to complete the inspection work of the equipment in this task.

[0193] 7. Inspection task completion statistics: statistical analysis of inspection history records of this terminal (user).

[0194] Optionally, the intelligent inspection and monitoring terminal is a computer, which can display various types of information through the screen for the terminal (user) to use and analyze, and then generate corresponding push tasks and control commands through the computer.

[0195] An electronic device according to an embodiment of the present invention includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements some or all steps of the above-mentioned method for detecting the status of equipment in a gas pipeline network.

[0196] Among them, the electronic device can be a computer, and correspondingly, its program is computer software. The above-mentioned parameters and steps in an electronic device of the present invention can refer to the parameters and steps in the embodiment of a method for detecting the status of equipment in a gas pipeline network above, and will not be repeated here.

[0197] Those skilled in the art will appreciate that the present invention may be implemented as a system, method, or computer program product. Therefore, the present disclosure may be specifically implemented in the following forms, namely: in the form of complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, the present invention may also be implemented in the form of a computer program product in one or more computer-readable media, the computer-readable media containing computer-readable program code. Computer-readable storage media may be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination thereof.

[0198] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0199] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for detecting the status of equipment in a gas pipeline network, characterized in that: The following steps are involved: Acquire an instrument image of each device in the gas pipeline network, wherein the instrument image is an image including a dial of an instrument on the device; For each of the instrument images, determining a reading of the dial in the instrument image according to the instrument image; For each of the instrument images, determining the operating status of the device corresponding to the instrument image according to the reading of the dial, wherein the operating status is a healthy state or a fault state; For each of the devices, detecting the safety risk of the gas network according to the operating status; For each instrument image, determining the reading of the dial in the instrument image according to the instrument image specifically includes: For each instrument image, determining connected body regions on the instrument image based on the instrument image, wherein each connected body region represents a region formed by pixel points on the instrument image whose first pixel value is greater than a first threshold; Determine, based on each of the connected body areas, a connected body area corresponding to each of the scale lines on the meter, and use the connected body area corresponding to each of the scale lines as a first target connected domain; For each of the first target connected domains, determining a first center point of the first target connected domain; For each of the watch dials, determining a first circle center according to each of the first center points, and determining a first contour line corresponding to the watch dial according to the first circle center and a preset radius, wherein the first circle center represents the center of the watch dial determined according to each of the first center points; For each of the first target connected domains, determining a second pixel value corresponding to the first target connected domain, where for each second pixel value, the second pixel value represents the sum of RGB values ​​of each pixel point on the first target connected domain; taking the smallest pixel value among the second pixel values ​​as the target pixel value, taking the position of the first target connected domain corresponding to the target pixel value as the pointer position; and determining a second contour line corresponding to the pointer position based on the pointer position; determining a first target center point from each of the first center points according to the first contour line and the second contour line; For each of the dials, a reference point is determined according to a preset coordinate system, where the preset coordinate system represents a coordinate system established with the first center of the circle as the origin, and the reference point represents a position corresponding to the zero mark on the dial; For each of the dials, determine a first straight line between the reference point and the first center of the circle, and a second straight line between the first target center point and the first center of the circle; based on the first straight line and the second straight line, determine a deflection angle of the second straight line relative to the first straight line; based on the deflection angle and a preset relationship, determine the reading of the dial; the preset relationship represents the correspondence between the deflection angle and the reading of the dial.

2. The method according to claim 1, characterized in that Determining the connected body area corresponding to each scale line on the meter according to each connected body area includes: For each of the connected body regions, the area of ​​the connected body region is obtained. If the area is greater than the second threshold and less than the third threshold, the connected body region is the connected body region corresponding to the scale line corresponding to the connected body region.

3. The method according to claim 1, characterized in that Also includes: For each of the first target connected domains, convert the first target connected domain into a second target connected domain in a rectangular shape; For each of the second target connected domains, determining a third target connected domain, wherein for each of the third target connected domains, the third target connected domain represents an area formed by each pixel point in the second target connected domain whose third pixel value is greater than a fourth threshold; The step of determining, for each of the first target connected domains, a first center point of the first target connected domain includes: For each of the third target connected components, a first center point of the third target connected component is determined.

4. The method according to claim 1, wherein For each of the dials, determining the first circle center according to each of the first center points includes: For each of the dials, determining a first circle center using a RANSAC model according to each of the first center points; The training process of the RANSAC model specifically includes: S1, obtaining a training set, where the training set includes a plurality of second center points and the coordinates corresponding to each of the second center points in a preset coordinate system; S2, inputting the training set into an initial model, training the initial model, and outputting a plurality of target points, wherein each target point is an interior point or an exterior point, wherein the interior point represents each second circle center and the corresponding coordinates of each second circle center in a preset coordinate system, and the exterior point represents each non-second circle center and the corresponding coordinates of each non-second circle center in the preset coordinate system; S3, determining a loss value according to a first number of the inliers and a second number of the outliers in the target point; S4, if the loss value satisfies a preset end condition, using the initial model when the preset end condition is satisfied as the RANSAC model; if the loss value does not satisfy the preset end condition, adjusting the network parameters of the initial network, and retraining the initial model according to the adjusted network parameters until the loss value of the initial model satisfies the preset end condition; The determining the loss value according to the first number of the inliers and the second number of the outliers in the target point includes: A loss value is determined according to a first number of the inliers and a second number of the outliers in the target point using a first formula, wherein the first formula is: Among them, k represents the number of current iterations, n inliers Indicates the first number, n outliers Represents the second number, t represents the proportion of inliers in all target points, p represents the loss value, and n represents the number of second center points in the training set.

5. The method according to claim 1, wherein The determining of the first target center point according to the first contour line and the second contour line includes: determining a first intersection point and a second intersection point according to the first contour line and the second contour line; Determine a second target center point based on the first intersection point, where the second target center point is the first center point closest to the first intersection point in a straight line among the first center points; Determine a third target center point based on the second intersection point; the third target center point is the first center point of each of the first center points that is closest to the second intersection point in a straight line; The first center point between the second target center point and the third target center point is used as the first target center point.

6. A gas network equipment status detection system, characterized in that: include: A first acquisition module is used to acquire an instrument image of each device in the gas pipeline network, wherein the instrument image is an image including a dial of an instrument on the device; A second acquisition module is configured to determine, for each of the instrument images, a reading of the dial in the instrument image based on the instrument image; A third acquisition module is configured to determine, for each of the instrument images, an operating status of the device corresponding to the instrument image based on the reading of the dial, where the operating status is a healthy state or a faulty state; a detection module, configured to detect the safety risk of the gas network for each of the devices according to the operating status; The second acquisition module 203 includes: A first determining module is configured to determine, for each instrument image, connected body regions on the instrument image based on the instrument image, wherein each connected body region represents a region formed by pixels on the instrument image having a first pixel value greater than a first threshold; A second determining module is configured to determine, based on each connected body area, the connected body area corresponding to each scale line on the meter, and use the connected body area corresponding to each scale line as a first target connected domain; A third determining module is configured to determine, for each first target connected domain, a first center point of the first target connected domain; a fourth determining module, configured to determine, for each watch dial, a first circle center based on each first center point, and determine a first contour line corresponding to the watch dial based on the first circle center and a preset radius, wherein the first circle center represents the center of the watch dial determined based on each first center point; a fifth determining module, for each first target connected domain, determining a second pixel value corresponding to the first target connected domain, wherein each second pixel value represents the sum of RGB values ​​of each pixel point on the first target connected domain; a sixth determining module, configured to use the smallest pixel value among the second pixel values ​​as the target pixel value, and the position of the first target connected domain corresponding to the target pixel value as the pointer position; and determine the second contour line corresponding to the pointer position according to the pointer position; A seventh determining module, configured to determine a first target center point from each first center point according to the first contour line and the second contour line; an eighth determining module, configured to determine, for each dial, a reference point according to a preset coordinate system, where the preset coordinate system represents a coordinate system established with the first center of the circle as the origin, and the reference point represents a position corresponding to a zero mark on the dial; The ninth determination module is used to determine, for each dial, a first straight line between the reference point and the first center of the circle, and a second straight line between the first target center point and the first center of the circle; based on the first straight line and the second straight line, determine a deflection angle of the second straight line relative to the first straight line; and based on the deflection angle and a preset relationship, determine a reading of the dial, wherein the preset relationship represents a correspondence between the deflection angle and the reading of the dial.

7. A gas network equipment status detection platform, characterized in that: A method for detecting the status of equipment in a gas pipe network as claimed in claim 1, comprising: an intelligent patrol monitoring terminal, an intelligent patrol cloud, and a patrol front end, wherein the intelligent patrol monitoring terminal is connected to the intelligent patrol cloud, and the intelligent patrol cloud is connected to the patrol front end; The inspection front end is used to obtain instrument images of various devices in the gas pipeline network, and the instrument images are images including the dials of the instruments on the devices; The intelligent inspection cloud is configured to determine, for each instrument image, a reading of the dial in the instrument image based on the instrument image, and to determine, for each instrument image, an operating status of the device corresponding to the instrument image based on the reading of the dial, where the operating status is a healthy state or a faulty state; The intelligent inspection and monitoring terminal is used to detect the safety risks of the gas pipeline network for each of the devices according to the operating status.

8. An electronic device comprising a memory, a processor, and a program stored in the memory and running on the processor, characterized in that: When the processor executes the program, the steps of the method for detecting the status of equipment in a gas pipeline network as described in any one of claims 1 to 5 are implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed on the terminal device, the terminal device executes the steps of the method for detecting the status of equipment in a gas pipeline network as described in any one of claims 1 to 5.

Images