Intelligent inspection system and method for VOC gas leakage based on gas cloud imaging

Abstract

The present invention discloses a VOC gas leakage intelligent inspection system and method based on gas-cloud imaging, belonging to the technical field of VOC gas leakage detection. The inspection system includes a mobile track and inspection points arranged along the monitored object, an intelligent slide-type robot moving along the mobile track, and a data processing system; the intelligent slide-type robot is provided with a gas-cloud imaging host, and the gas-cloud imaging host is provided with a medium-wave infrared movement, a visible light movement and a wireless communication module. The intelligent slide-type robot moves to each inspection point along the mobile track, and the gas-cloud imaging host takes infrared images and visible light images of each inspection point, and transmits the infrared images and visible light images to the data processing system through the wireless communication module. The present invention can timely and accurately lock the specific leakage point, determine the VOC leakage concentration, and realize the whole process of intelligent inspection from inspection, leakage detection and processing, leakage accident archiving, leakage warning, leakage accident processing to feedback of processing results.

Description

Technical Field

[0001] The present invention belongs to the technical field of VOC gas leakage detection, and relates to a VOC gas leakage intelligent inspection system and method based on gas cloud imaging. Background Art

[0002] There are many equipment (such as pumps, compressors, storage tanks) and pipelines and their components (such as valves and flanges) that are prone to leakage in petroleum and chemical enterprises. Once a leak occurs, a large amount of VOC gas or other toxic and hazardous substances are often released, which not only poses a safety hazard but also pollutes the environment. Timely discovery of leaking equipment and pipelines is of great significance for eliminating safety hazards and protecting the environment.

[0003] Traditional leak detection mainly relies on manual work, using portable hydrogen flame ionization detectors for on-site detection to promptly detect leaks, thereby repairing or replacing the detected leaks. However, this method consumes a lot of manpower and material resources, and cannot achieve real-time monitoring, and cannot timely monitor and capture potential safety hazards.

[0004] CN118959908A discloses a method and device for detecting oil pipeline leakage based on changes in TVOC concentration. A plurality of TVOC gas sampling detectors connected to a bus are (pre-)buried below the ground (G) at intervals along the oil (gas) pipeline section (P). Each TVOC gas sampling detector is connected via a power supply bus and a communication bus; the power supply bus and the communication bus are protected by an airtight casing, and the head end of the bus is connected to a monitoring device with bus communication, data display and storage functions, which is used for regular inspections to obtain the TVOC concentration of the sampling detector. By detecting the increment of the TVOC concentration change of the gas in a plurality of porous sampling tubes inserted into the soil, it is determined whether there is an oil pipeline leakage nearby, and the location of the leakage point is estimated based on the pipeline direction and the concentration increment. This technical solution requires the pre-burying of a large number of VOC sampling detectors. On the one hand, the monitoring cost is high, and on the other hand, it is not suitable for monitoring movable equipment.

[0005] CN114689804A discloses a method for intelligently monitoring dangerous gas leaks based on VOC gas cloud imaging, which uses electronic equipment to identify and process visible light image data and infrared light image data to intelligently monitor dangerous leaked gases. By using meteorological data to determine the azimuth of the leak source, a small amount of monitoring equipment can be used to monitor a large monitoring area, and the leak location can be quickly determined; the total amount of VOC is detected by a PID sensor, and a timely response to VOC leaks is achieved. This technical solution can monitor VOC leaks in a certain monitoring area and detect the total amount of VOC at the same time, but it cannot locate the specific leaking equipment or components, nor can it determine the VOC leakage concentration.

[0006] CN117516811A discloses a gas leakage monitoring system, method, device and inspection device, the system includes a first pan-tilt, a second pan-tilt, a gas cloud imaging camera, a laser detector and a computing platform; wherein the first pan-tilt, the second pan-tilt, the gas cloud imaging camera and the laser detector are respectively connected to the computing platform by signals. This technical solution requires the addition of a laser detector and a second pan-tilt in addition to the gas cloud imaging camera to determine the VOC leakage concentration. Summary of the invention

[0007] The purpose of the present invention is to provide a VOC gas leakage intelligent inspection system and method based on gas cloud imaging, which can timely and accurately lock the specific leakage point, determine the VOC leakage concentration, and realize the full process intelligent inspection of the monitoring object from inspection, leakage detection and processing, leakage accident archiving, leakage warning, leakage accident processing to processing result feedback. The purpose of the present invention is achieved through the following specific technical solutions.

[0008] The primary aspect of the present invention is to provide a VOC gas leakage intelligent inspection system based on gas cloud imaging, including a mobile track and inspection points arranged along the monitored object, an intelligent slide-type robot moving along the mobile track, and a data processing system; the intelligent slide-type robot is provided with a gas cloud imaging host, and the gas cloud imaging host is provided with a medium-wave infrared movement, a visible light movement and a wireless communication module. The intelligent slide-type robot moves to each inspection point along the mobile track, and the gas cloud imaging host takes infrared images and visible light images of each inspection point, and transmits the infrared images and visible light images to the data processing system through the wireless communication module.

[0009] Furthermore, the data processing system includes a leakage analysis module, a leakage point locking module, a display module, an information push module and an information storage module. The leakage analysis module determines whether there is a leakage by analyzing the infrared images and visible light images of each inspection point taken by the gas cloud imaging host; if it is determined that there is a leakage, the leakage point locking module locks the specific leakage point through analysis and determines the VOC concentration of the leakage point. The display module displays the leakage point and its VOC concentration. The information push module pushes the leakage point and its VOC concentration information to the staff for processing. The information storage module stores the leakage point and its VOC concentration information and the processing process and results.

[0010] Furthermore, it also includes a wireless charging module. When the power level of the intelligent slide-type robot is lower than a set value, the intelligent slide-type robot moves to the vicinity of the wireless charging module, and the wireless charging module charges the intelligent slide-type robot.

[0011] Another aspect of the present invention provides a VOC gas leakage intelligent inspection based on gas cloud imaging, which is implemented by the above-mentioned VOC gas leakage intelligent inspection system, including the following steps:

[0012] The S1 intelligent slide-rail robot inspects along the moving track, and the gas cloud imaging host takes infrared images and visible light images of each inspection point, and transmits the infrared images and visible light images to the data processing system through the wireless communication module;

[0013] After the leakage analysis module of the S2 data processing system finds the VOC leakage gas cloud, it creates a VOC leakage event record M;

[0014] S3 The leakage analysis module of the data processing system further processes and analyzes the VOC leakage event record M of step S2, realizes the conversion of the total amount of VOC leakage originally monitored by gas cloud imaging to the VOC leakage concentration, that is, the conversion of ppm-m to ppm, and draws a VOC gas cloud imaging concentration distribution map at the same time;

[0015] S4: a leakage point locking module of the data processing system locks the leakage point based on the VOC leakage concentration calculated in step S3;

[0016] The display module of the S5 data processing system displays the leak point and its VOC concentration.

[0017] The information push module of the S6 data processing system pushes the leakage point and its VOC concentration information to the staff for processing;

[0018] The information storage module of the S7 data processing system stores the leakage point and its VOC concentration information as well as the processing process and results.

[0019] Furthermore, the specific method of step S3 is:

[0020] S3-1 Read the preliminary leakage event M recorded in step 2S, which includes a real-time infrared image and currently captured VOC leakage information, the leakage information includes a leakage gas cloud image G and a total leakage amount Vm (in ppm-m, indicating the total leakage amount when the VOC leakage gas cloud thickness is 1 meter).

[0021] S3-2 Based on the corresponding relationship between the size, distance and image pixels of the infrared core imaging object, the size relationship between the gas cloud pixels in the infrared image and the actual space is calculated by formula 1 and formula 2:

[0022] F(x)=Tan(hAng / 2 / 180) × dis / (W / 2) Formula (1)

[0023] F(y)= Tan(vAng / 2 / 180) × dis / (H / 2) Formula (2)

[0024] Among them, F(x) represents the calculation of the horizontal size of a pixel, in m;

[0025] F(y) represents the vertical dimension of a pixel, in m.

[0026] Tan represents the tangent operation;

[0027] hAng represents the horizontal field angle of the infrared movement, a known quantity, in degrees;

[0028] vAng represents the vertical field angle of the infrared movement, a known quantity, in degrees;

[0029] dis represents the distance between the current monitoring point and the movement, which is a known quantity in meters;

[0030] W represents the width of the infrared core imaging image, in meters;

[0031] H represents the height of the infrared core imaging image, in meters;

[0032] The height and width of each pixel are obtained by the above formula, and the area represented by each pixel in the gas cloud image G is calculated.

[0033] S3-3 grayscale conversion is performed on the leaking gas cloud image G, and the grayscale value of each pixel in the gas cloud is read. According to the gas cloud imaging principle, the grayscale value is a direct reflection of the gas cloud leakage concentration, that is, the higher the gas cloud leakage concentration, the smaller the grayscale value, and the darker the color value when reflected in the grayscale image. In order to facilitate subsequent calculations, the image is inverted, that is, the grayscale value of each pixel in the leaking gas cloud image after grayscale conversion is subtracted from 255 in sequence.

[0034] S3-4 Matrix Z is established by using the grayscale data of the image after inverting in step S3-3, and the grayscale values ​​in the matrix Z are integrated to obtain the total grayscale value Gm of the leaking gas cloud. There is a conversion factor t between the total grayscale value Gm and the leakage volume Vm, t=Vm / Gm.

[0035] S3-5 Considering the influence of the background, an area of ​​8 to 12 pixels outside the gas cloud is selected and averaged as the background grayscale Gb, and then the grayscale value of each pixel in the matrix Z is subtracted from Gb.

[0036] S3-6 Use the conversion factor t to perform multiplication operations with the factors in the matrix Z in sequence to obtain the VOC leakage concentration distribution matrix data GRz in units of ppm; use the color caliper to color and draw the image to obtain the VOC leakage concentration distribution graph GRb in units of ppm.

[0037] At a certain moment, if there are multiple leakage points leaking simultaneously within the imaging field of the gas cloud imaging, the gas cloud imaging will form an overall leakage cloud map and cannot give specific information on which leakage points are leaking. The leaked gas will diffuse under the influence of the wind. Generally, the concentration near the leakage point is high and relatively uniform, while the concentration in the diffusion area is low, with a large change gradient and a non-uniform state. According to this principle, step S4 realizes locking the leakage point, and the specific method is as follows.

[0038] S4-1 Preset a set Loc of all possible positions where VOC leakage may occur within the inspection point field of view of the monitored object. Loc contains the positions and their unique numbers of the positions where VOC leakage may occur.

[0039] S4-2 Based on the VOC leakage concentration distribution matrix GRz calculated in step S3-6, design an inspection matrix Ji. This inspection matrix is a matrix with dimensions of i×k.

[0040] i = GRz_h / 15 Formula (3)

[0041] k = GRz_c / 15 Formula (4)

[0042] Where, GRz_h represents the number of rows of the matrix GRz;

[0043] GRz_c represents the number of columns of the matrix GRz;

[0044] In actual calculation, if i<10, then i takes the empirical value 10; if k<10, then k takes the empirical value 10.

[0045] S4-3 Use Ji to start from the position (0, 0) in the matrix GRz, first slide along the X direction, and then along the Y direction in a loop with step sizes of i / 2 and k / 2 respectively, and calculate the average gray value of the area covered by the points at the sliding positions of Ji to obtain the average concentration distribution matrix GRzavg in ppm units.

[0046] S4-4 Continue to use Ji to repeat the above operation in GRzavg. While calculating the mean value of the area at the sliding position of Ji, calculate the variance of the values in this area, so as to obtain the secondary mean matrix GRzavg_dub and the variance matrix GRzavg_var.

[0047] S4-5 Analyze the characteristic selection functions of GRzavg_dub and GRzavg_var in turn:

[0048] F(x,y) = GRzavg_dub(x,y)>avg&GRzavg_var(x,y)<avg / vr Formula (5)

[0049] That is, make a logical judgment on whether GRzavg_dub (x, y) is greater than avg, and GRzavg_var (x, y) is less than avg / vr. If the judgment value is true, F(x, y) = 1, otherwise F(x, y) = 0;

[0050] Among them, GRzavg_dub (x, y) is the GRzavg_dub value at the coordinate (x, y);

[0051] GRzavg_var(x,y) is the value of GRzavg_var at the coordinate (x,y);

[0052] avg is the average value of GRZ;

[0053] vr is an empirical value, usually 10;

[0054] When F(x,y)=1, it indicates that there is a leakage point in the area. At this time, a circular match is performed with the data set Loc. If a leakage point Loc(i) happens to be within the range of the matrix GRzavg determined in step S4-3, the position of the leakage point is locked, the leakage point number is recorded and marked as the leakage point R(i). At the same time, the average concentration and maximum concentration of the matrix GRzavg are calculated, which are the data information of the current leakage point. If multiple leakage points are matched, the leakage point with the closest square distance is taken as the leakage point R(i). Finally, the leakage point, concentration and other information are marked on the leakage gas cloud concentration distribution map.

[0055] Furthermore, the specific method of step S5 is:

[0056] S5-1 3D modeling of the monitored object to be inspected is performed to form a three-dimensional model diagram of the monitored object;

[0057] S5-2: numbering each possible leakage point in the three-dimensional model, and the numbering data is consistent with the numbering of Loc in step S4-1;

[0058] S5-3 The leakage point information R(i) calculated in step S4 is sequentially displayed on the corresponding leakage points on the three-dimensional model diagram in a highlighted manner and its leakage concentration is marked to form a leakage point monitoring information summary diagram covering the entire monitoring object.

[0059] Furthermore, the specific method of step S6 is: when the leakage concentration of the leakage point exceeds the set threshold value, it is marked as an excessive leakage event and pushed to the administrator as an alarm message. After receiving the alarm message, the administrator promptly delegates relevant personnel to the site to handle it and feedback the processing results.

[0060] The present invention provides a VOC gas leakage intelligent inspection system and method based on gas cloud imaging, which can effectively, in real time, and unattendedly realize intelligent monitoring of VOC leakage points such as pipelines and devices in chemical parks. The system runs 24 hours a day, 7 days a week, and combined with 3D modeling, it intuitively displays the real-time status of possible leakage points in various pipelines and production devices in the chemical park, greatly reducing the traditional manpower investigation costs. After detecting VOC leakage at the leakage point, it can promptly record and push alarm information to the administrator, so that the administrator can delegate maintenance personnel to deal with it in time. On the one hand, timely discovery of leakage events and alarm of leakage events can reduce environmental pollution and possible production safety accidents. On the other hand, it reduces a large amount of manual investigation costs and avoids the safety hazards caused by manual investigation. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a structural schematic diagram of a VOC gas leakage intelligent inspection system based on gas cloud imaging provided by the present invention.

[0062] Figure numerals: 1-intelligent slide-rail robot, 2-gas-cloud imaging host, 3-mobile track, 4-wireless charging module, 5-inspection point.

[0063] Figure 2 The present invention provides a flow chart of a VOC gas leakage intelligent inspection method based on gas cloud imaging.

[0064] Figure 3 It is a schematic diagram of the calculation principle of leakage point locking data.

[0065] Figure 4 This is a leak point locking result diagram from a certain intelligent inspection.

[0066] Figure 5 It is a 3D model diagram of the intelligent inspection results of a certain device. DETAILED DESCRIPTION

[0067] The technical solution of the present invention is described clearly and completely below in conjunction with the accompanying drawings of the specification. Obviously, the described implementation is only a part of the implementation of the present invention, not all of the implementations. Based on the implementation of the present invention, all other implementations obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0068] An intelligent VOC gas leakage inspection system based on gas cloud imaging, the structure is as follows Figure 1As shown, it includes a mobile track 3, a wireless charging module 4 and a patrol point 5 arranged along the monitored object, an intelligent slide-type robot 1 moving along the mobile track 3, and a data processing system; the intelligent slide-type robot 1 is provided with a gas cloud imaging host 2, and the gas cloud imaging host 2 is provided with a medium-wave infrared movement, a visible light movement and a wireless communication module. The intelligent slide-type robot 1 moves along the mobile track 3 to each patrol point 5, and the gas cloud imaging host takes infrared images and visible light images of each patrol point 5, and transmits the infrared images and visible light images to the data processing system through the wireless communication module.

[0069] The data processing system includes a leakage analysis module, a leakage point locking module, a display module, an information push module and an information storage module. The leakage analysis module determines whether there is a leakage by analyzing the infrared images and visible light images of each inspection point taken by the gas cloud imaging host; if it is determined that there is a leakage, the leakage point locking module locks the specific leakage point through analysis and determines the VOC concentration of the leakage point. The display module displays the leakage point and its VOC concentration. The information push module pushes the leakage point and its VOC concentration information to the staff for processing. The information storage module stores the leakage point and its VOC concentration information and the processing process and results. Example

[0070] A VOC gas leakage intelligent inspection method based on gas cloud imaging, the process is as follows Figure 2 As shown, the VOC gas leakage intelligent inspection system of Example 1 is implemented, including the following steps.

[0071] The S1 intelligent slide-rail robot inspects along the moving track. The gas-cloud imaging host takes infrared images and visible light images of each inspection point and transmits the infrared images and visible light images to the data processing system through the wireless communication module.

[0072] After the leakage analysis module of the S2 data processing system finds the VOC leakage gas cloud, it creates a VOC leakage event record M.

[0073] The leakage analysis module of the S3 data processing system further processes and analyzes the VOC leakage event record M of step S2 to realize the conversion of the total VOC leakage originally monitored by the gas cloud imaging to the VOC leakage concentration, that is, the conversion from ppm-m to ppm, and draws a VOC gas cloud imaging concentration distribution map at the same time.

[0074] S4: a leakage point locking module of the data processing system locks the leakage point based on the VOC leakage concentration calculated in step S3;

[0075] The display module of the S5 data processing system displays the leak point and its VOC concentration;

[0076] The information push module of the S6 data processing system pushes the leakage point and its VOC concentration information to the staff for processing;

[0077] The information storage module of the S7 data processing system stores the leakage point and its VOC concentration information as well as the processing process and results.

[0078] The specific method of step S3 is:

[0079] S3-1 Read the preliminary leakage event M recorded in step 2S, which includes a real-time infrared image and currently captured VOC leakage information, the leakage information includes a leakage gas cloud image G and a total leakage amount Vm (in ppm-m, indicating the total leakage amount when the VOC leakage gas cloud thickness is 1 meter).

[0080] S3-2 Based on the corresponding relationship between the size, distance and image pixels of the infrared core imaging object, the size relationship between the gas cloud pixels in the infrared image and the actual space is calculated by formula 1 and formula 2:

[0081] F(x)=Tan(hAng / 2 / 180) × dis / (W / 2) Formula (1)

[0082] F(y)= Tan(vAng / 2 / 180) × dis / (H / 2) Formula (2)

[0083] Among them, F(x) represents the calculation of the horizontal size of a pixel, in m;

[0084] F(y) represents the vertical dimension of a pixel, in m.

[0085] Tan represents the tangent operation;

[0086] hAng represents the horizontal field angle of the infrared movement, a known quantity, in degrees;

[0087] vAng represents the vertical field angle of the infrared movement, a known quantity, in degrees;

[0088] dis represents the distance between the current monitoring point and the movement, which is a known quantity in meters;

[0089] W represents the width of the infrared core imaging image, in meters;

[0090] H represents the height of the infrared core imaging image, in meters;

[0091] The height and width of each pixel are obtained by the above formula, and the area represented by each pixel in the gas cloud image G is calculated.

[0092] S3-3 grayscale conversion is performed on the leaking gas cloud image G, and the grayscale value of each pixel in the gas cloud is read. According to the gas cloud imaging principle, the grayscale value is a direct reflection of the gas cloud leakage concentration, that is, the higher the gas cloud leakage concentration, the smaller the grayscale value, and the darker the color value when reflected in the grayscale image. In order to facilitate subsequent calculations, the image is inverted, that is, the grayscale value of each pixel in the leaking gas cloud image after grayscale conversion is subtracted from 255 in sequence.

[0093] S3-4 Matrix Z is established by using the grayscale data of the image after inverting in step S3-3, and the grayscale values ​​in the matrix Z are integrated to obtain the total grayscale value Gm of the leaking gas cloud. There is a conversion factor t between the total grayscale value Gm and the leakage volume Vm, t=Vm / Gm.

[0094] S3-5 Considering the influence of the background, an area of ​​8 to 12 pixels outside the gas cloud is selected and averaged as the background grayscale Gb, and then the grayscale value of each pixel in the matrix Z is subtracted from Gb.

[0095] S3-6 Use the conversion factor t to perform multiplication operations with the factors in the matrix Z in sequence to obtain the VOC leakage concentration distribution matrix data GRz in units of ppm; use the color caliper to color and draw the image to obtain the VOC leakage concentration distribution graph GRb in units of ppm.

[0096] The principle of locking the leakage point in step S4 is as follows Figure 3 As shown in the figure, the grid data of number a is the GRz matrix data calculated in step S3, in which the part with numerical values ​​is the grayscale data of the leakage gas cloud concentration, and the part without numerical values ​​indicates the area where the leakage gas cloud has not spread. Number b is a schematic diagram of the sliding calculation of Ji in the matrix GRz, which will first slide along the X direction and then along the Y direction, and calculate the average grayscale of the area covered by Ji to construct a new matrix GRzavg, as shown in the figure by number c. Number d is a schematic diagram of the sliding calculation of Ji in the matrix GRzavg, which will first slide along the X direction and then along the Y direction, and calculate the variance of the values ​​in the area while calculating the mean of the sliding position area of ​​Ji, thereby obtaining the GRzavg_dub quadratic mean matrix (as shown in number e) and the GRzavg_var variance matrix (as shown in number f).

[0097] The specific method is as follows.

[0098] S4-1 pre-sets a set of all points Loc where VOC leakage may occur within the field of view of the inspection point of the monitored object, Loc including the location of the point where VOC leakage may occur and its unique number.

[0099] S4-2 Based on the VOC leakage concentration distribution matrix GRz calculated in step S3-6, design a test matrix Ji. This test matrix is a matrix with dimensions i×k.

[0100] i = GRz_h / 15 Equation (3)

[0101] k = GRz_c / 15 Equation (4)

[0102] Among them, GRz_h represents the number of rows of the matrix GRz;

[0103] GRz_c represents the number of columns of the matrix GRz;

[0104] In actual calculation, if i < 10, then i takes the empirical value 10; if k < 10, then k takes the empirical value 10.

[0105] S4-3 Use Ji to start from the position (0, 0) in the matrix GRz, first slide along the X direction, and then along the Y direction in a loop with step sizes of i / 2 and k / 2 respectively, and calculate the average gray value of the area covered by the sliding position points of Ji to obtain the average concentration distribution matrix GRzavg in units of Ji.

[0106] S4-4 Continue to use Ji to repeat the above operation in GRzavg. While calculating the mean value of the area where Ji slides, calculate the variance of the values in this area, so as to obtain the secondary mean matrix GRzavg_dub and the variance matrix GRzavg_var.

[0107] S4-5 Analyze the characteristic selection functions of GRzavg_dub and GRzavg_var in turn:

[0108] F(x,y) = GRzavg_dub(x,y) > avg & GRzavg_var(x,y) < avg / vr Equation (5)

[0109] That is, make a logical judgment on whether GRzavg_dub(x,y) is greater than avg and whether GRzavg_var(x,y) is less than avg / vr. If the judgment value is true, then F(x,y) = 1; otherwise, F(x,y) = 0;

[0110] Among them, GRzavg_dub(x,y) is the value of GRzavg_dub at the coordinate (x,y);

[0111] GRzavg_var(x,y) is the value of GRzavg_var at the coordinate (x,y);

[0112] avg is the average value of GRZ;

[0113] vr is an empirical value, usually 10;

[0114] When F(x,y)=1, it indicates that there is a leakage point in the area. At this time, a circular match is performed with the data set Loc. If a leakage point Loc(i) happens to be within the range of the matrix GRzavg determined in step S4-3, the position of the leakage point is locked, the leakage point number is recorded and marked as the leakage point R(i). At the same time, the average concentration and maximum concentration of the matrix GRzavg are calculated, which are the data information of the current leakage point. If multiple leakage points are matched, the leakage point with the closest square distance is taken as the leakage point R(i). Finally, the leakage point, concentration and other information are marked on the leakage gas cloud concentration distribution map.

[0115] The specific method of step S5 is:

[0116] S5-1 3D modeling of the monitored object to be inspected is performed to form a three-dimensional model diagram of the monitored object;

[0117] S5-2: numbering each possible leakage point in the three-dimensional model, and the numbering data is consistent with the numbering of Loc in step S4-1;

[0118] S5-3 The leakage point information R(i) calculated in step S4 is sequentially displayed on the corresponding leakage points on the three-dimensional model diagram in a highlighted manner and its leakage concentration is marked to form a leakage point monitoring information summary diagram covering the entire monitoring object.

[0119] The specific method of step S6 is: when the leakage concentration of the leakage point exceeds the set threshold, it is marked as an excessive leakage event and pushed to the administrator as an alarm message. After receiving the alarm message, the administrator promptly delegates relevant personnel to the site to handle it and feedback the processing results.

[0120] like Figure 4 As shown, the leakage point is locked during an inspection. Through the processing method provided by the present invention, the specific leakage point can be analyzed from the VOC leakage gas cloud captured by the gas cloud imaging, and the leakage point position can be located, and the leakage point number and the current leakage concentration can be marked, which intuitively reflects which leakage points in the current inspection field of view have leakage behavior and their corresponding numbers and leakage concentrations.

[0121] like Figure 5 As shown, it is an intelligent inspection result display for the entire monitored object based on the 3D model. It comprehensively utilizes the leakage information monitored at each inspection point, combines it with the 3D model of the device, highlights the leakage points with leakage behavior, and marks their numbers and current leakage concentrations, intuitively and globally displaying the health status of the monitored object.

[0122] Although the embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and cannot be understood as limiting 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 without departing from the principles and purpose of the present invention. The protection scope of the present invention is defined by the claims and their equivalent technical solutions.

Claims

1. The VOC gas leakage intelligent inspection method based on gas cloud imaging is characterized by: It is realized by a VOC gas leakage intelligent inspection system, which includes a mobile track and inspection points arranged along the monitored object, an intelligent slide-type robot moving along the mobile track, and a data processing system; the intelligent slide-type robot is provided with a gas cloud imaging host, and the gas cloud imaging host is provided with a medium-wave infrared movement, a visible light movement and a wireless communication module; the data processing system includes a leakage analysis module, a leakage point locking module, a display module, an information push module and an information storage module; The intelligent inspection method comprises the following steps: The S1 intelligent slide-rail robot inspects along the moving track, and the gas cloud imaging host takes infrared images and visible light images of each inspection point, and transmits the infrared images and visible light images to the data processing system through the wireless communication module; After the leakage analysis module of the S2 data processing system finds the VOC leakage gas cloud, it creates a VOC leakage event record M; S3 The leakage analysis module of the data processing system further processes and analyzes the VOC leakage event record M of step S2, realizes the conversion of the total amount of VOC leakage originally monitored by gas cloud imaging to the VOC leakage concentration, and draws the VOC gas cloud imaging concentration distribution map at the same time; the specific method is: S3-1 reading the preliminary leakage event recorded in step S2, which includes a real-time infrared image and currently captured VOC leakage information, wherein the leakage information includes a leakage gas cloud image G and a total leakage amount Vm; S3-2 Based on the corresponding relationship between the size, distance and image pixels of the infrared core imaging object, the size relationship between the gas cloud pixels in the infrared image and the actual space is calculated by formula 1 and formula 2: F(x)=Tan(hAng / 2 / 180) × dis / (W / 2) Formula (1) F(y) = Tan(vAng / 2 / 180) × dis / (H / 2) Formula (2) Among them, F(x) represents the calculation of the horizontal size of a pixel, in units of m; F(y) represents the vertical dimension of a pixel, in m. Tan represents the tangent operation; hAng represents the horizontal field angle of the infrared movement, a known quantity, in degrees; vAng represents the vertical field angle of the infrared movement, a known quantity, in degrees; dis represents the distance between the current monitoring point and the movement, which is a known quantity in meters; W represents the width of the infrared core imaging image, in meters; H represents the height of the infrared core imaging image, in meters; The height and width of each pixel obtained by the above formula are used to calculate the area represented by each pixel in the leaking gas cloud image G; S3-3 converts the leaking gas cloud image G into grayscale, reads the grayscale value of each pixel in the gas cloud, and then inverts the image; S3-4 Establish a matrix Z with the grayscale data after the image is inverted in step S3-3, integrate the grayscale values ​​in the matrix Z to obtain the total grayscale value Gm of the leaking gas cloud, and calculate the conversion factor t between the total grayscale value Gm and the leakage volume Vm, t=Vm / Gm; S3-5 selects an area of ​​8 to 12 pixels outside the leaking gas cloud and averages it as the background grayscale Gb, then subtracts Gb from the grayscale value of each pixel in the matrix Z; S3-6 Use the conversion factor t to perform multiplication operations with the factors in the matrix Z in sequence to obtain the VOC leakage concentration distribution matrix data GRz; use the color scale caliper to color and draw the image to obtain the VOC leakage concentration distribution map GRb; S4 The leakage point locking module of the data processing system locks the leakage point based on the VOC leakage concentration calculated in step S3; the specific method is: S4-1 pre-sets a set of all points Loc where VOC leakage may occur within the field of view of the inspection point of the monitored object, Loc including the location of the point where VOC leakage may occur and its unique number; S4-2 Based on the VOC leakage concentration distribution matrix GRz calculated in step S3-6, a test matrix Ji is designed. The test matrix is ​​a matrix with a dimension of i×k. i = GRz_h / 15 Formula (3) k = GRz_c / 15 Formula (4) Among them, GRz_h represents the number of rows of matrix GRz; GRz_c represents the number of columns of matrix GRz; In actual calculations, if i<10, i takes the empirical value 10, and if k<10, k takes the empirical value 10; S4-3 Use Ji to start from the position (0, 0) in the matrix GRz, first along the X direction, then along the Y direction according to the cycle sliding with the step length of i / 2 and k / 2 respectively, and calculate the average gray value of the area covered by the sliding position point of Ji, and obtain the average concentration distribution matrix GRzavg; S4-4 Continue to use Ji to repeat the operation of the previous step in GRzavg, and calculate the variance of the mean of the area where Ji slides while calculating the mean of the area, so as to obtain the quadratic mean matrix GRzavg_dub and the variance matrix GRzavg_var; S4-5 Analyze the feature selection functions of GRzavg_dub and GRzavg_var in turn: F(x,y) = GRzavg_dub(x,y)> avg & GRzavg_var(x,y) < avg / vr Formula (5) Among them, GRzavg_dub (x, y) is the GRzavg_dub value at the coordinate (x, y); GRzavg_var(x,y) is the value of GRzavg_var at the coordinate (x,y); avg is the average value of GRZ; vr is the experience value; When F(x,y)=1, it indicates that there is a leakage point in the area. At this time, a circular match is performed with Loc. If a leakage point Loc(i) happens to be within the range of the matrix GRzavg determined in step S4-3, the position of the leakage point is locked, the leakage point number is recorded and marked as the leakage point R(i). At the same time, the average concentration and maximum concentration of the matrix GRzavg are calculated, which are the data information of the current leakage point. If multiple leakage points are matched, the leakage point with the closest square distance is taken as the leakage point R(i). Finally, the leakage point and concentration information are marked on the leakage gas cloud concentration distribution map. The display module of the S5 data processing system displays the leak point and its VOC concentration; The information push module of the S6 data processing system pushes the leakage point and its VOC concentration information to the staff for processing; The information storage module of the S7 data processing system stores the leakage point and its VOC concentration information as well as the processing process and results.

2. The VOC gas leakage intelligent inspection method according to claim 1 is characterized in that: The VOC gas leakage intelligent inspection system also includes a wireless charging module.

3. The VOC gas leakage intelligent inspection method according to claim 1 is characterized in that: The specific method of step S5 is: S5-1 3D modeling of the monitored object to be inspected is performed to form a three-dimensional model diagram of the monitored object; S5-2: numbering each possible leakage point in the three-dimensional model, and the numbering data is consistent with the numbering of Loc in step S4-1; S5-3 The leakage point information R(i) calculated in step S4 is sequentially displayed on the corresponding leakage points on the three-dimensional model diagram in a highlighted manner and its leakage concentration is marked to form a leakage point monitoring information summary diagram covering the entire monitoring object.

4. The VOC gas leakage intelligent inspection method according to claim 3 is characterized in that: The specific method of step S6 is: when the leakage concentration of the leakage point exceeds the set threshold, it is marked as an excessive leakage event and pushed to the administrator as an alarm message. After receiving the alarm message, the administrator appoints relevant personnel to handle it on site and feedback the processing results.

Images