Weak current monitoring system and method based on intelligent internet of things

Abstract

The invention discloses a weak current monitoring system and method based on intelligent Internet of Things, and relates to the technical field of weak current monitoring, and the system comprises a real-time data monitoring unit, an equipment recognition unit, a fault detection unit, a fault positioning unit and an early warning prompt unit. According to the invention, the real-time data monitoring unit collects data in the weak current equipment and the area where the weak current equipment is located, so that an operator can timely obtain a series of data of a specified area to monitor changes in the area in real time, and the equipment identification unit analyzes all collected infrared image data to obtain the type of the equipment, so that the accuracy of the equipment identification is improved. In combination with a specific area corresponding to image data, related information of the equipment can be further analyzed, and meanwhile, a fault detection unit is utilized to judge a fault state corresponding to each type of equipment when abnormity is found, so that maintenance personnel can take measures in time for intervention, and the maintenance efficiency is improved. And an abnormal area can be marked in the whole image data through the fault positioning unit, so that the specific position of the equipment abnormity can be conveniently locked.

Description

Technical Field

[0001] The present invention relates to the field of weak current monitoring technology, and in particular to a weak current monitoring system and method based on intelligent Internet of Things. Background Art

[0002] The weak current monitoring system is monitored through front-end equipment such as high-definition cameras and sensors, and uses a series of data to achieve remote monitoring and management. The system has functions such as intelligent analysis and abnormal alarm, can quickly detect and respond to abnormal situations, and effectively prevent and control safety risks. The invention patent with application number 202110427495.6 discloses "a weak current monitoring system based on intelligent Internet of Things. The present invention includes vertical weak current monitoring branches, horizontal weak current monitoring branches, monitoring data transmission branches and monitoring centers; each of the monitoring data transmission branches includes multiple transmission points and at least one data aggregation point. Each transmission point can transmit data along the monitoring data transmission branch and aggregate the data at the data aggregation point. The data aggregation point can transmit the data to the data aggregation points in other adjacent monitoring rings, or to the monitoring center. The data in each data aggregation point is finally transmitted to the monitoring center. The present invention can reduce the construction cost of the weak current monitoring system, make the node allocation of the monitoring system more reasonable, and ensure the monitoring effect, but also save the overall energy consumption of the monitoring system."

[0003] The above-mentioned existing technology solves the problem of increased overall energy consumption caused by unreasonable allocation of monitoring nodes. However, during the use of the system, since infrared image data is not used to analyze the equipment in each area, the monitoring of the equipment often requires the installation of a large number of sensors, which increases the cost of use. In addition, during the monitoring process, it is impossible to locate and mark abnormal areas of the equipment, which requires maintenance personnel to check the lines and equipment one by one, thus affecting work efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a weak current monitoring system and method based on intelligent Internet of Things to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: a weak current monitoring system based on intelligent Internet of Things, comprising a fault location unit;

[0006] A real-time data monitoring unit, which uses the infrared detection equipment carried by the inspection robot to shoot the equipment and lines in the designated area to obtain infrared image data, uses a camera to shoot each numbered area to obtain image data, uses a temperature sensor to monitor each device to obtain temperature data, and transmits these data to a database for storage;

[0007] A device identification unit receives infrared image data, performs preliminary segmentation on the image data according to a set threshold, determines all connected domains in the image data and their proportions, selects the connected domain with the largest proportion as the main area of ​​the image data, determines the grayscale value of the seed point in the main area and the grayscale values ​​of other pixels in the neighborhood of the point, and then selects all remaining seed points in the image data according to the current seed point, determines the image data after secondary segmentation according to the positions of all seed points, extracts the feature vector corresponding to the image data, and inputs it into the device identification model, uses the model to perform analysis, and thus determines the type of the device;

[0008] A fault detection unit receives infrared image data, determines a local threshold using a genetic algorithm, segments the image based on the data, counts the original grayscale mean of each closed area in the image, takes the area with the largest grayscale mean as the abnormal area, determines the current temperature value of a specified detection point in the abnormal area, calculates the relative temperature difference corresponding to the point based on the current temperature value of the detection point and the temperature value in the normal area, and determines the fault state of the current device;

[0009] An early warning prompt unit uses deep learning technology to analyze temperature data, gas data and equipment risk levels. When the data and risk levels exceed preset values, an early warning prompt is issued to the control center, and the data exceeding the preset values ​​is uploaded.

[0010] Preferably, the real-time data monitoring unit includes an infrared image acquisition module, a sensor data acquisition module and a data association module. The infrared image acquisition module uses the infrared detection equipment carried by the inspection robot to shoot the equipment and lines in the designated area, and after obtaining the infrared image data, transmits it to the industrial computer. The sensor data acquisition module numbers each area, uses a camera to shoot each numbered area to obtain image data, uses a temperature sensor to monitor each device to obtain temperature data, monitors each area through a gas sensor to obtain gas content data, and transmits these data to a database for storage. After the data association module determines the equipment corresponding to each area, it numbers it and associates the equipment-related data with the corresponding area data.

[0011] Preferably, the device identification unit includes a grayscale value analysis module and a seed point setting module. The grayscale value analysis module receives infrared image data, grayscales it, obtains a histogram corresponding to the image, calculates the grayscale value at each position point according to a segmented transformation algorithm, outputs a new grayscale value, uses these data to count the number of pixels corresponding to each grayscale level, analyzes the pixel ratio of each grayscale level according to the total number of pixels of the image information, thereby obtaining a new histogram, performs equalization on the histogram, and outputs processed image data according to the equalized grayscale distribution. The seed point setting module performs preliminary segmentation on the image data according to a set threshold, determines all connected domains in the image data and their proportions, selects the connected domain with the largest proportion as the main area of ​​the image data, analyzes the main area, obtains the center point position corresponding to the area, and sets the point as a seed point. The segmented transformation algorithm is specifically as follows:

[0012]

[0013] Among them, x1, x2 represent the original grayscale values ​​of the two specified position points before transformation, y1, y2 represent the grayscale values ​​of the two specified position points after transformation, x represents the grayscale value of the current position point, and f(x) represents the grayscale value after transformation.

[0014] Preferably, the device identification unit also includes an image data segmentation module and a device type output module. After the image data segmentation module determines the grayscale value of the seed point and the grayscale values ​​of other pixels in the neighborhood of the point, it calculates the difference between the seed point and the other adjacent pixels. If the difference is less than a preset value, the current pixel is determined to be a seed point and stored in a set. Otherwise, it is marked and the operation is repeated until all pixels in the connected domain are processed. The image data after secondary segmentation is determined according to the positions of all seed points in the set. The device type output module extracts the segmented image data, determines the feature vector corresponding to the image data, and inputs it into the device identification model. The model is used for analysis to determine the type of the device, and the area number and device type corresponding to the image data are output through a visual interface. The device identification model is specifically an identification model built based on machine learning technology.

[0015] Preferably, the fault detection unit includes a closed area generation module and a risk level division module. After receiving the infrared image data, the closed area generation module performs grid processing on it, calculates the variance corresponding to each grid using a grayscale analysis algorithm, analyzes the new grayscale value of each pixel in the grid based on the variance, and determines the local threshold using a genetic algorithm. Then, the image is segmented based on these data to obtain multiple closed areas. The risk level division module divides the equipment risk level into four categories, namely, general level, warning level, danger level and emergency level. The grayscale analysis algorithm is specifically as follows:

[0016]

[0017]

[0018] Among them, f(x,y) represents the gray value of any pixel in the image, μ(i,j) represents the gray mean, σ 2 (i, j) represents the local variance, (x, y) represents the position coordinates of the pixel point, (i, j) represents the position coordinates of the center point of the field, and m, n represent parameters.

[0019] Preferably, the fault detection unit also includes a temperature analysis module and a relative temperature difference determination module. The temperature analysis module counts the original grayscale mean of each closed area, takes the area with the largest grayscale mean as the abnormal area, determines the upper limit and lower limit of the current infrared image temperature field according to the temperature sensor data of the area, and uses these data to calculate the current temperature value of the specified detection point in the abnormal area. After the relative temperature difference determination module extracts the temperature value of the detection point in the normal area, it calculates the relative temperature difference corresponding to the point according to the current temperature value of the detection point and the temperature value in the normal area. If the relative temperature difference is greater than or equal to the preset value, the current device is judged to be an external fault, otherwise, the current device is judged to be a pending fault.

[0020] Preferably, the fault location unit includes a risk assessment module, a feature point marking module and a position output module. If the current device is an external fault, the risk assessment module determines the risk level corresponding to the fault according to the range of the relative temperature difference of the detection point, and outputs these data through a visual interface. If the current device is a pending fault, manual detection is used to determine the risk level. The feature point marking module uses the SURF algorithm to extract feature points of the image data of the abnormal area and the overall image data respectively, determines the relative position of each feature point, and matches the feature points of the abnormal area with the feature points in the overall image data. If the match is successful, the successfully matched feature points are marked in the overall image data and output through a visual interface. If the position output module fails to match, the image data of the abnormal area is converted into a template image, and the template image is used to perform sliding comparison on the overall image data. The similarity between the template image and the area to be detected in the overall image data is calculated by a similarity measurement algorithm, and the area to be detected with the highest similarity is output as the target position.

[0021] The weak current monitoring method based on intelligent Internet of Things comprises the following steps:

[0022] S1. Data monitoring and collection: Use the infrared detection equipment carried by the inspection robot to shoot the equipment and lines in the designated area to obtain infrared image data, use the camera to shoot each numbered area to obtain image data, use the temperature sensor to monitor each device to obtain temperature data, and transfer these data to the database for storage;

[0023] S2. Distinguishing device type: receiving infrared image data, performing preliminary segmentation on the image data according to a set threshold, determining all connected domains in the image data and their proportions, selecting the connected domain with the largest proportion as the main area of ​​the image data, determining the grayscale value of the seed point in the main area and the grayscale values ​​of other pixels in the neighborhood of the point, filtering out all remaining seed points in the image data according to the current seed point, determining the image data after secondary segmentation according to the positions of all seed points, extracting the feature vector corresponding to the image data, inputting it into the device recognition model, and using the model to perform analysis to determine the type of the device;

[0024] S3, determine the equipment fault status: receive infrared image data, determine the local threshold value by using genetic algorithm, segment the image according to these data, count the original grayscale mean of each closed area in the image, take the area with the largest grayscale mean as the abnormal area, determine the current temperature value of the specified detection point in the abnormal area, calculate the relative temperature difference corresponding to the point according to the current temperature value of the detection point and the temperature value in the normal area, and determine the current equipment fault status;

[0025] S4. Marking the fault location: If the current device is an external fault, the risk level corresponding to the fault is determined according to the range of the relative temperature difference of the detection point. If the current device is a pending fault, manual detection is used to determine the risk level. The SURF algorithm is used to extract and match the feature points of the image data of the abnormal area and the overall image data respectively, and the successfully matched feature points are marked in the overall image data, and they are output through a visual interface;

[0026] S5. Early warning: Use deep learning technology to analyze temperature data, gas data and equipment risk levels. When the data and risk levels exceed the preset values, an early warning will be sent to the control center and the data exceeding the preset values ​​will be uploaded.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] The present invention collects data on weak current equipment and the area where it is located through a real-time data monitoring unit, so that operators can obtain a series of data in a specified area in a timely manner to monitor changes in the area in real time. The equipment identification unit analyzes all collected infrared image data to obtain the type of equipment. Combined with the specific area corresponding to the image data, the relevant information of the equipment can be further analyzed. At the same time, the fault detection unit is used to determine the fault state corresponding to each equipment when an abnormality is found, so as to ensure that maintenance personnel can take timely measures to intervene. The fault location unit can mark the abnormal area in the overall image data, so that the specific location of the equipment abnormality can be locked in the first time. The early warning prompt unit can send early warning information to the control center in time when abnormal data is generated, so as to ensure the safety of each area. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 Provides a schematic diagram of the overall system flow for an embodiment of the present invention;

[0030] Figure 2 An internal module block diagram of a real-time data monitoring unit provided by an embodiment of the present invention;

[0031] Figure 3 A block diagram of the internal modules of a device identification unit provided in an embodiment of the present invention;

[0032] Figure 4 An internal module block diagram of a fault detection unit provided by an embodiment of the present invention;

[0033] Figure 5 This is a block diagram of the internal modules of the fault location unit provided in an embodiment of the present invention.

[0034] In the figure: 1. Real-time data monitoring unit; 101. Infrared image acquisition module; 102. Sensor data acquisition module; 103. Data association module; 2. Equipment identification unit; 201. Gray value analysis module; 202. Seed point setting module; 203. Image data segmentation module; 204. Equipment type output module; 3. Fault detection unit; 301. Closed area generation module; 302. Risk level classification module; 303. Temperature analysis module; 304. Relative temperature difference determination module; 4. Fault location unit; 401. Risk assessment module; 402. Feature point marking module; 403. Position output module. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] See also Figure 1-5 ,The present invention provides a technical solution: a weak current monitoring system based on intelligent Internet of Things, including a fault location unit 4;

[0037] Real-time data monitoring unit 1, the real-time data monitoring unit 1 uses the infrared detection equipment carried by the inspection robot to shoot the equipment and lines in the specified area to obtain infrared image data, uses a camera to shoot each numbered area to obtain image data, uses a temperature sensor to monitor each device to obtain temperature data, and transmits these data to a database for storage;

[0038] Device identification unit 2, device identification unit 2 receives infrared image data, performs preliminary segmentation on the image data according to a set threshold, determines all connected domains in the image data and their proportions, selects the connected domain with the largest proportion as the main area of ​​the image data, determines the grayscale value of the seed point in the main area and the grayscale values ​​of other pixels in the neighborhood of the point, and then selects all remaining seed points in the image data according to the current seed point, determines the image data after secondary segmentation according to the positions of all seed points, extracts the feature vector corresponding to the image data, and inputs it into the device identification model, uses the model to analyze, and thus determines the type of the device;

[0039] Fault detection unit 3, fault detection unit 3 receives infrared image data, determines local thresholds using genetic algorithms, segments the image based on these data, counts the original grayscale mean values ​​of each closed area in the image, takes the area with the largest grayscale mean value as the abnormal area, determines the current temperature value of a specified detection point in the abnormal area, calculates the relative temperature difference corresponding to the point based on the current temperature value of the detection point and the temperature value in the normal area, and determines the fault state of the current device;

[0040] The early warning prompt unit 5 uses deep learning technology to analyze temperature data, gas data and equipment risk level. When the data and risk level exceed the preset value, an early warning prompt is issued to the control center, and the data exceeding the preset value is uploaded.

[0041] The real-time data monitoring unit 1 includes an infrared image acquisition module 101, a sensor data acquisition module 102 and a data association module 103. The infrared image acquisition module 101 uses the infrared detection equipment carried by the inspection robot to shoot the equipment and lines in the specified area, and after obtaining the infrared image data, transmits it to the industrial computer. The sensor data acquisition module 102 numbers each area, uses a camera to shoot each numbered area to obtain image data, uses a temperature sensor to monitor each device to obtain temperature data, and uses a gas sensor to monitor each area to obtain gas content data. These data are transmitted to the database for storage. After the data association module 103 determines the equipment corresponding to each area, it numbers it and associates the equipment-related data with the corresponding area data;

[0042] The device identification unit 2 includes a grayscale value analysis module 201 and a seed point setting module 202. The grayscale value analysis module 201 receives infrared image data, grayscales it, obtains a histogram corresponding to the image, calculates the grayscale value at each position point according to the segmented transformation algorithm, outputs a new grayscale value, uses these data to count the number of pixels corresponding to each grayscale level, analyzes the pixel ratio of each grayscale level according to the total number of pixels of the image information, thereby obtaining a new histogram, performs equalization on the histogram, and outputs the processed image data according to the equalized grayscale distribution. The seed point setting module 202 performs preliminary segmentation on the image data according to the set threshold, determines all connected domains in the image data and their proportions, selects the connected domain with the largest proportion as the main area of ​​the image data, analyzes the main area, obtains the center point position corresponding to the area, and sets the point as the seed point. The segmented transformation algorithm is specifically as follows:

[0043]

[0044] Among them, x1, x2 represent the original grayscale values ​​of the two specified position points before transformation, y1, y2 represent the grayscale values ​​of the two specified position points after transformation, x represents the grayscale value of the current position point, and f(x) represents the grayscale value after transformation;

[0045] The device identification unit 2 also includes an image data segmentation module 203 and a device type output module 204. After the image data segmentation module 203 determines the gray value of the seed point and the gray values ​​of other pixels in the neighborhood of the point, it calculates the difference between the seed point and other adjacent pixels. If the difference is less than a preset value, the current pixel is determined to be a seed point and stored in a set. Otherwise, it is marked. The operation is repeated until all pixels in the connected domain are processed. The image data after secondary segmentation is determined according to the positions of all seed points in the set. The device type output module 204 extracts the segmented image data, determines the feature vector corresponding to the image data, and inputs it into the device identification model. The model is used for analysis to determine the type of the device, and the area number and device type corresponding to the image data are output through a visual interface.

[0046] The fault detection unit 3 includes a closed area generation module 301 and a risk level division module 302. After receiving the infrared image data, the closed area generation module 301 performs grid processing on it, calculates the variance corresponding to each grid using a grayscale analysis algorithm, analyzes the new grayscale value of each pixel in the grid based on the variance, and determines the local threshold using a genetic algorithm. Then, the image is segmented based on these data to obtain multiple closed areas. The risk level division module 302 divides the equipment risk level into four categories, namely, general level, warning level, danger level and emergency level. The grayscale analysis algorithm is specifically as follows:

[0047]

[0048]

[0049] Among them, f(x,y) represents the gray value of any pixel in the image, μ(i,j) represents the gray mean, σ 2 (i, j) represents the local variance, (x, y) represents the position coordinates of the pixel point, (i, j) represents the position coordinates of the center point of the field, and m, n represent parameters;

[0050] The fault detection unit 3 also includes a temperature analysis module 303 and a relative temperature difference determination module 304. The temperature analysis module 303 counts the original grayscale mean values ​​of each closed area, takes the area with the largest grayscale mean value as the abnormal area, determines the upper limit and lower limit values ​​of the current infrared image temperature field according to the temperature sensor data of the area, and uses these data to calculate the current temperature value of the specified detection point in the abnormal area. After the relative temperature difference determination module 304 extracts the temperature value of the detection point in the normal area, it calculates the relative temperature difference corresponding to the point according to the current temperature value of the detection point and the temperature value in the normal area. If the relative temperature difference is greater than or equal to the preset value, it is judged that the current device is an external fault, otherwise it is judged that the current device is a pending fault;

[0051] The fault location unit 4 includes a risk assessment module 401, a feature point marking module 402 and a position output module 403. If the current device is an external fault, the risk assessment module 401 determines the risk level corresponding to the fault according to the range of the relative temperature difference of the detection point, and outputs these data through a visual interface. If the current device is a pending fault, it is detected manually and the risk level is determined. The feature point marking module 402 uses the SURF algorithm to extract the feature points of the image data of the abnormal area and the overall image data respectively, determines the relative position of each feature point, matches the feature points of the abnormal area with the feature points in the overall image data, and if the match is successful, marks the successfully matched feature points in the overall image data and outputs them through a visual interface. If the position output module 403 fails to match, the image data of the abnormal area is converted into a template image, and the template image is used to perform sliding comparison on the overall image data. The similarity between the template image and the area to be detected in the overall image data is calculated by a similarity measurement algorithm, and the area to be detected with the highest similarity is output as the target position. The similarity measurement algorithm is specifically as follows:

[0052]

[0053] Among them, D(i,j) represents the similarity, B i,j (m,n) represents the content of the area to be detected, A(m,n) represents the content of the template image, M and N represent the length and width of the original image data, and i,j represent the pixel coordinates of the image;

[0054] A weak current monitoring method based on intelligent Internet of Things includes the following steps:

[0055] S1. Data monitoring and collection: Use the infrared detection equipment carried by the inspection robot to shoot the equipment and lines in the designated area to obtain infrared image data, use the camera to shoot each numbered area to obtain image data, use the temperature sensor to monitor each device to obtain temperature data, and transfer these data to the database for storage;

[0056] S2. Distinguishing device type: receiving infrared image data, performing preliminary segmentation on the image data according to a set threshold, determining all connected domains in the image data and their proportions, selecting the connected domain with the largest proportion as the main area of ​​the image data, determining the grayscale value of the seed point in the main area and the grayscale values ​​of other pixels in the neighborhood of the point, filtering out all remaining seed points in the image data according to the current seed point, determining the image data after secondary segmentation according to the positions of all seed points, extracting the feature vector corresponding to the image data, inputting it into the device recognition model, and using the model to perform analysis to determine the type of the device;

[0057] S3, determine the equipment fault status: receive infrared image data, determine the local threshold value by using genetic algorithm, segment the image according to these data, count the original grayscale mean of each closed area in the image, take the area with the largest grayscale mean as the abnormal area, determine the current temperature value of the specified detection point in the abnormal area, calculate the relative temperature difference corresponding to the point according to the current temperature value of the detection point and the temperature value in the normal area, and determine the current equipment fault status;

[0058] S4. Marking the fault location: If the current device is an external fault, the risk level corresponding to the fault is determined according to the range of the relative temperature difference of the detection point. If the current device is a pending fault, manual detection is used to determine the risk level. The SURF algorithm is used to extract and match the feature points of the image data of the abnormal area and the overall image data respectively, and the successfully matched feature points are marked in the overall image data, and they are output through a visual interface;

[0059] S5. Early warning: Use deep learning technology to analyze temperature data, gas data and equipment risk levels. When the data and risk levels exceed the preset values, an early warning will be sent to the control center and the data exceeding the preset values ​​will be uploaded.

[0060] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0061] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A weak current monitoring system based on intelligent Internet of Things, comprising a fault location unit (4), characterized in that: A real-time data monitoring unit (1), wherein the real-time data monitoring unit (1) uses an infrared detection device carried by the inspection robot to photograph the equipment and lines in the designated area to obtain infrared image data, uses a camera to photograph each numbered area to obtain image data, uses a temperature sensor to monitor each device to obtain temperature data, and transmits the data to a database for storage; A device identification unit (2), wherein the device identification unit (2) receives infrared image data, performs preliminary segmentation on the image data according to a set threshold value, determines all connected domains in the image data and their proportions, selects the connected domain with the largest proportion as the main area of ​​the image data, determines the grayscale value of the seed point in the main area and the grayscale values ​​of other pixels in the neighborhood of the point, selects all remaining seed points in the image data according to the current seed point, determines the image data after secondary segmentation according to the positions of all seed points, extracts the feature vector corresponding to the image data, inputs it into the device identification model, and uses the model to perform analysis to determine the type of the device; A fault detection unit (3), wherein the fault detection unit (3) receives infrared image data, determines a local threshold value using a genetic algorithm, segments the image based on the data, calculates the original grayscale mean of each closed area in the image, takes the area with the largest grayscale mean as the abnormal area, determines the current temperature value of a specified detection point in the abnormal area, calculates the relative temperature difference corresponding to the point based on the current temperature value of the detection point and the temperature value in the normal area, and determines the fault state of the current device; An early warning prompt unit (5), wherein the early warning prompt unit (5) uses deep learning technology to analyze temperature data, gas data and equipment risk level, and when the data and risk level exceed preset values, a early warning prompt is issued to the control center, and the data exceeding the preset value is uploaded.

2. According to claim 1, a weak current monitoring system based on intelligent Internet of Things is characterized by: The real-time data monitoring unit (1) comprises an infrared image acquisition module (101), a sensor data acquisition module (102) and a data association module (103). The infrared image acquisition module (101) uses the infrared detection equipment carried by the inspection robot to photograph the equipment and lines in the designated area, obtains infrared image data, and transmits the infrared image data to the industrial computer. The sensor data acquisition module (102) numbers each area, uses a camera to photograph each numbered area to obtain image data, uses a temperature sensor to monitor each equipment to obtain temperature data, uses a gas sensor to monitor each area to obtain gas content data, and transmits the data to a database for storage. The data association module (103) determines the equipment corresponding to each area, numbers it, and associates the equipment-related data with the corresponding area data.

3. According to claim 1, a weak current monitoring system based on intelligent Internet of Things is characterized by: The device identification unit (2) comprises a grayscale value analysis module (201) and a seed point setting module (202). The grayscale value analysis module (201) receives infrared image data, performs grayscale processing on the data, obtains a histogram corresponding to the image, calculates the grayscale value at each position point according to a segmented transformation algorithm, outputs a new grayscale value, uses these data to count the number of pixels corresponding to each grayscale level, analyzes the pixel ratio of each grayscale level according to the total number of pixel points of the image information, thereby obtaining a new histogram, performs equalization processing on the histogram, and outputs processed image data according to the equalized grayscale distribution. The seed point setting module (202) performs preliminary segmentation on the image data according to a set threshold value, determines all connected domains in the image data and their proportions, selects the connected domain with the largest proportion as the main area of ​​the image data, analyzes the main area, obtains the center point position corresponding to the area, and sets the point as the seed point.

4. According to claim 3, a weak current monitoring system based on intelligent Internet of Things is characterized in that: The device identification unit (2) further comprises an image data segmentation module (203) and a device type output module (204). The image data segmentation module (203) determines the gray value of the seed point and the gray values ​​of other pixel points in the neighborhood of the point, and then calculates the difference between the seed point and other adjacent pixel points. If the difference is less than a preset value, the current pixel point is determined to be a seed point and stored in a set. Otherwise, it is marked. The operation is repeated until all pixel points in the connected domain are processed. The image data after secondary segmentation is determined based on the positions of all seed points in the set. The device type output module (204) extracts the segmented image data, determines the feature vector corresponding to the image data, and then inputs it into the device identification model. The model is used for analysis to determine the type of the device, and the area number and device type corresponding to the image data are output through a visual interface.

5. According to claim 1, a weak current monitoring system based on intelligent Internet of Things is characterized by: The fault detection unit (3) comprises a closed area generation module (301) and a risk level division module (302). After receiving the infrared image data, the closed area generation module (301) performs grid processing on the data, calculates the variance corresponding to each grid using a grayscale analysis algorithm, obtains a new grayscale value of each pixel in the grid based on the variance, determines a local threshold using a genetic algorithm, and then segments the image based on these data to obtain a plurality of closed areas. The risk level division module (302) divides the equipment risk level into four categories, namely, a general level, a warning level, a dangerous level, and an emergency level.

6. According to claim 5, a weak current monitoring system based on intelligent Internet of Things is characterized in that: The fault detection unit (3) also includes a temperature analysis module (303) and a relative temperature difference determination module (304). The temperature analysis module (303) counts the original grayscale mean values ​​of each closed area, takes the area with the largest grayscale mean value as the abnormal area, determines the upper limit value and the lower limit value of the current infrared image temperature field according to the temperature sensor data of the area, and uses these data to calculate the current temperature value of the specified detection point in the abnormal area. The relative temperature difference determination module (304) extracts the temperature value of the detection point in the normal area, and calculates the relative temperature difference corresponding to the point according to the current temperature value of the detection point and the temperature value in the normal area. If the relative temperature difference is greater than or equal to a preset value, it is determined that the current device is an external fault, otherwise it is determined that the current device is a pending fault.

7. According to claim 1, a weak current monitoring system based on intelligent Internet of Things is characterized by: The fault location unit (4) comprises a risk assessment module (401), a feature point marking module (402) and a position output module (403). If the current device is an external fault, the risk assessment module (401) determines the risk level corresponding to the fault according to the range of the relative temperature difference of the detection point, and outputs these data through a visual interface. If the current device is a pending fault, manual detection is performed and the risk level is determined. The feature point marking module (402) uses the SURF algorithm to extract feature points from the image data of the abnormal area and the overall image data respectively, determines the relative position of each feature point, matches the feature points of the abnormal area with the feature points in the overall image data, and if the match is successful, marks the successfully matched feature points in the overall image data and outputs them through a visual interface. If the match fails, the position output module (403) converts the image data of the abnormal area into a template image, performs sliding comparison on the overall image data using the template image, calculates the similarity between the template image and the area to be detected in the overall image data through a similarity determination algorithm, and outputs the area to be detected with the highest similarity as the target position.

8. A weak current monitoring method based on intelligent Internet of Things, characterized in that: The weak current monitoring method is applicable to a weak current monitoring system based on intelligent Internet of Things as described in any one of claims 1 to 7, and comprises the following steps: S1. Data monitoring and collection: Use the infrared detection equipment carried by the inspection robot to shoot the equipment and lines in the designated area to obtain infrared image data, use the camera to shoot each numbered area to obtain image data, use the temperature sensor to monitor each device to obtain temperature data, and transfer these data to the database for storage; S2. Distinguishing device type: receiving infrared image data, performing preliminary segmentation on the image data according to a set threshold, determining all connected domains in the image data and their proportions, selecting the connected domain with the largest proportion as the main area of ​​the image data, determining the grayscale value of the seed point in the main area and the grayscale values ​​of other pixels in the neighborhood of the point, filtering out all remaining seed points in the image data according to the current seed point, determining the image data after secondary segmentation according to the positions of all seed points, extracting the feature vector corresponding to the image data, inputting it into the device recognition model, and using the model to perform analysis to determine the type of the device; S3, determine the equipment fault status: receive infrared image data, determine the local threshold value by using genetic algorithm, segment the image according to these data, count the original grayscale mean of each closed area in the image, take the area with the largest grayscale mean as the abnormal area, determine the current temperature value of the specified detection point in the abnormal area, calculate the relative temperature difference corresponding to the point according to the current temperature value of the detection point and the temperature value in the normal area, and determine the current equipment fault status; S4. Marking the fault location: If the current device is an external fault, the risk level corresponding to the fault is determined according to the range of the relative temperature difference of the detection point. If the current device is a pending fault, manual detection is used to determine the risk level. The SURF algorithm is used to extract and match the feature points of the image data of the abnormal area and the overall image data respectively, and the successfully matched feature points are marked in the overall image data, and they are output through a visual interface; S5. Early warning: Use deep learning technology to analyze temperature data, gas data and equipment risk levels. When the data and risk levels exceed the preset values, an early warning will be sent to the control center and the data exceeding the preset values ​​will be uploaded.

Images