A line following method for a live-line robot

Abstract

The present application relates to the technical field of robots, and in particular to a line detection method for a hot-line work robot, which comprises obtaining a to-be-processed image of a line area of a power distribution network, pre-processing the to-be-processed image to obtain a to-be-detected image; identifying line connection, insulator and connector; determining whether the temperature is abnormal; detecting whether the insulator connector is abnormal; and determining whether the detection is abnormal. By combining a target detection model and temperature diagnosis technology, the present application automatically identifies the line position, insulator and connector, and quickly locates and processes abnormal conditions, with high diagnosis accuracy and strong reliability. Through comprehensive and detailed detection of the line connection, potential equipment hazards can be found in time, preventive maintenance of the power distribution network equipment can be realized, equipment failure risks can be reduced, and the accuracy, autonomy and preventability of the power distribution network equipment detection are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of robotics, and more particularly to a method for detecting the movement of a live-line working robot. Background Technology

[0002] With the continuous development and intelligent upgrading of the power system, the safe operation and maintenance of the power distribution network has become increasingly important. Traditional power line inspection methods mainly rely on manual operations, which have many shortcomings, such as high personnel safety risks, low detection efficiency, and insufficient detection accuracy. With the rapid development of robotics and artificial intelligence, live-line working robots are increasingly widely used in the field of power detection, becoming an effective way to solve the pain points of traditional manual inspection.

[0003] However, existing methods for identifying work lines in live-line working robots still face many challenges. Traditional work line identification methods mainly rely on single features, making it difficult to comprehensively and accurately determine anomalies at work line connections. Furthermore, existing methods lack multi-dimensional and multi-level detection means for work line connections, making it difficult to detect potential abnormalities and defects in a timely manner.

[0004] Patent document CN106595762A discloses a method for testing tension insulators using a live-line working robot. The live-line working robot has a robotic arm mounted on a robot platform. Responding to control data, the robotic arm carries a partial discharge detector, an insulator zero-value tester, a current transformer, as well as a high-definition camera, an infrared camera, and an electronic ultraviolet flaw detector. It moves around the tension insulator and performs data detection. The data processing and control system processes the detection data from each detection device, compares the obtained relevant index values ​​with normal values ​​in the database to determine the working status of the tension insulator, and thus generates a test report.

[0005] This reveals the following problem: In existing technologies, the features for line recognition are too limited, making it difficult to accurately detect defects and resulting in poor accuracy. Summary of the Invention

[0006] To address this issue, the present invention provides a method for detecting power distribution network lines for live-line working robots. This method overcomes the problem of poor accuracy in existing technologies due to the reliance on single line identification features, difficulty in accurately detecting defects, and the use of multi-information fusion combined with deep learning technology to achieve accurate identification and anomaly detection of power distribution network lines.

[0007] To achieve the above objectives, the present invention provides a method for detecting the path of a live-line working robot, comprising:

[0008] Step S1: Obtain the image to be processed in the distribution network line area, and preprocess the image to be processed, including denoising, adjusting pixel values ​​and adjusting brightness, to obtain the image to be detected;

[0009] Step S2: Analyze the image to be detected to identify the two-dimensional coordinates of the row lines, and determine the row line connection, insulator and connector based on the row line position;

[0010] Step S3: Detect the first temperature at the connection point of the line and compare it with the preset first standard temperature. Based on the comparison result, determine whether the temperature is abnormal. If the temperature is abnormal, clean the surface of the connection point and calculate the temperature change after cleaning.

[0011] Step S4: Compare the temperature change with the preset temperature change threshold. Based on the comparison result, determine whether the surface dust is abnormal. If it is determined that the surface dust is not abnormal, check whether the insulator is damaged. If it is damaged, replace it. If it is not damaged, check whether the insulator connector is abnormal.

[0012] Step S5: If the insulator connector is abnormal, the connector is replaced. If the insulator connector is normal, the processing temperature of the next row line connection is detected and compared with the first temperature to determine whether an abnormality is detected. If an abnormality is detected, a second standard temperature is calculated and the second standard temperature is used instead of the first standard temperature to compare and judge the temperature of the remaining row line connections.

[0013] Further, step S2 includes:

[0014] Step S21: Analyze the image to be detected using the Ground-SAM2 algorithm and the YOLO model respectively to identify the two-dimensional coordinates of the line, calculate the coordinate deviation based on the identification results, and compare the coordinate deviation with the preset standard deviation threshold.

[0015] When the coordinate deviation is greater than the standard deviation threshold, the YOLO model is trained using the image to be detected.

[0016] The identification is determined to be correct when the coordinate deviation is less than the standard deviation threshold.

[0017] Step S22: Determine the line position coordinates based on the two-dimensional coordinates of the line and the point cloud data of the distribution network line area. Based on the line position coordinates, determine the line connection, insulator and connector at the line using a deep learning model.

[0018] Further, step S3 includes:

[0019] Step S31: Detect the first temperature at the connection point of the row lines;

[0020] Step S32: Compare the first temperature with the first standard temperature. If the first temperature is greater than the first standard temperature, then determine that the temperature is abnormal.

[0021] Step S33: When an abnormal temperature is detected, use a cleaning device to clean the surface of the connection and detect the processing temperature of the line connection after cleaning.

[0022] Step S34: Calculate the difference between the first temperature and the processing temperature to obtain the temperature change.

[0023] Furthermore, in step S4, the process of determining whether the surface dust is abnormal based on the comparison results includes:

[0024] If the temperature change is less than the preset temperature change threshold, it is determined that it is not an abnormality of surface dust.

[0025] If the temperature change is greater than the preset temperature change threshold, it is determined that the surface dust is abnormal.

[0026] Furthermore, in step S4, the process of detecting whether the insulator is damaged includes:

[0027] Obtain a surface image of the insulator;

[0028] The surface image is input into a preset convolutional neural network model;

[0029] The surface image is analyzed by the convolutional neural network model to determine whether the insulator is damaged.

[0030] Furthermore, in step S4, the process of detecting whether the connector is abnormal includes:

[0031] Remove the insulator at the line connection to expose the connector:

[0032] The voltage of the connector is detected to obtain the voltage to be measured.

[0033] The voltage to be measured is compared with a preset standard voltage to obtain a comparison result. Based on the comparison result, it is determined whether the connector is abnormal.

[0034] If the voltage to be tested is greater than the standard voltage, then the connector is determined to be abnormal.

[0035] If the voltage to be tested is less than or equal to the standard voltage, then the connector is considered to be normal.

[0036] Further, step S5 includes:

[0037] Step S51, move to the next connection point of the row line;

[0038] Step S52: Detect the second temperature at the next connection point;

[0039] Step S53: Calculate the temperature difference between the first temperature and the second temperature, compare the temperature difference with a preset temperature difference threshold, and determine whether the detection is normal based on the comparison result.

[0040] Step S54: Calculate the second standard temperature when an abnormality is detected.

[0041] Furthermore, the process of comparing the temperature difference with a preset temperature difference threshold in step S53 to determine whether the detection is normal includes:

[0042] If the temperature difference is less than or equal to the temperature difference threshold, then an anomaly is determined.

[0043] If the temperature difference is greater than the temperature difference threshold, the detection is considered normal.

[0044] Furthermore, in step S54, the process of calculating the second standard temperature when an anomaly is detected includes:

[0045] When an anomaly is detected, the average temperature of the first temperature and the second temperature is calculated.

[0046] The average temperature is set as the second standard temperature.

[0047] Further, step S1 includes:

[0048] Step S11: Acquire the image to be processed in the power distribution network line area using a camera;

[0049] Step S12: Perform noise smoothing and denoising on the image to be processed, adjust the pixel values ​​to standard pixel values, and adjust the image brightness to standard brightness to obtain the image to be detected.

[0050] Compared with existing technologies, the beneficial effects of this invention are that by combining target detection models and temperature diagnostic technology, it achieves intelligent and precise detection of power distribution network line areas. It can automatically identify line positions, insulators, and connectors, and quickly locate and handle abnormalities. It can not only detect temperature anomalies, but also further determine whether it is a surface dust problem, and detect abnormalities on the surface of insulators and connectors. It has high diagnostic accuracy and strong reliability. Through comprehensive and detailed detection of line connections, potential equipment hazards can be discovered in a timely manner, enabling preventive maintenance of power distribution network equipment, reducing the risk of equipment failure, and significantly improving the accuracy, autonomy, and preventiveness of power distribution network equipment detection.

[0051] Furthermore, by using the YOLO model fine-tuned from the Ground-SAM2 model to analyze image recognition of line lines, the accuracy and reliability of detection can be improved. The target detection model has a fast computation speed in the field of object detection. This method uses the target detection model to quickly analyze the image to be detected, which helps to improve detection efficiency and real-time performance, and helps to improve the reliability and safety of the power system.

[0052] Furthermore, by accurately detecting the temperature at the connection point of the line and comparing it with the preset standard temperature, abnormal equipment temperature can be quickly and accurately identified, providing a key basis for subsequent fault diagnosis. The temperature anomaly detection and surface cleaning process can promptly detect and handle potential equipment hazards, effectively provide early warning of possible power equipment failures, improve equipment operational reliability, and realize a fully automated process of temperature detection, anomaly judgment and surface cleaning, reducing manual intervention and improving the autonomous maintenance capability of the live-line working robot.

[0053] Furthermore, by comparing temperature change thresholds, it is possible to accurately distinguish whether surface dust is abnormal. By measuring the temperature change before and after cleaning, minute temperature differences can be captured, and even slight surface dust accumulation can be detected in time, improving the sensitivity of detection. The preset temperature change threshold can be adjusted according to the actual environment and equipment characteristics, making the dust abnormality judgment method highly adaptable and universal.

[0054] Furthermore, through deep learning technology, the model can automatically learn and extract features from insulator surface images to detect surface damage. This enables high-precision identification of minute cracks, defects, and damage on the insulator surface, allowing for automatic early warning and precise location of insulator damage. This provides accurate information for subsequent maintenance and replacement, improving detection efficiency and reliability, and preventing equipment failures in advance.

[0055] Furthermore, by directly removing the insulator and performing voltage testing on the connector, a precise diagnosis of the connector's internal condition is achieved. By judging voltage anomalies, potential hidden dangers in the connector can be detected in a timely manner, such as increased contact resistance or insulation damage that may lead to voltage anomalies. This enables accurate identification and early warning in the early stages of fault development, which is beneficial for early detection of potential equipment risks and improving power supply reliability.

[0056] Furthermore, by calculating the temperature difference between adjacent connections and comparing it with a preset threshold, a comprehensive and systematic inspection of the power distribution network lines is achieved. This enables more sensitive detection of abnormal temperature changes, improves the accuracy of fault identification, and avoids potential problems that may be missed by single-point detection. By comparing the temperature difference with the preset threshold, it is possible to more accurately determine whether there are abnormalities at the connection, effectively distinguish between normal temperature fluctuations and actual faults, and reduce the false judgment rate.

[0057] Furthermore, by setting a temperature difference threshold, it is possible to accurately distinguish whether the temperature change at the connection point of the line is within the normal range, making the anomaly detection more objective and scientific. The preset temperature difference threshold can be flexibly adjusted according to the actual application scenario and equipment characteristics, which can improve the sensitivity of detection and avoid false alarms caused by oversensitivity. Through accurate judgment of temperature difference, the accuracy and reliability of detection are improved, and potential equipment problems can be detected early, enabling early warning maintenance of the distribution network lines and reducing the risk of sudden equipment failure.

[0058] Furthermore, by calculating the average of the first and second temperatures and dynamically adjusting the standard temperature, the detection system achieves intelligent self-adaptation to different environmental and equipment temperature characteristics. The dynamic temperature standard adjustment mechanism can effectively eliminate misjudgments caused by factors such as environmental temperature fluctuations and equipment aging. By using the average calculation method, it can more accurately identify real temperature anomalies and reduce false alarm and false negative rates.

[0059] Furthermore, through noise smoothing and denoising technology, random noise and interference signals in the image are effectively suppressed, significantly improving the image clarity and contrast, providing a high-quality image foundation for subsequent target recognition. Through standardized preprocessing, the consistency and stability of images in different environments are enhanced, and the detection accuracy of the live-line working robot in different working scenarios is improved. Attached Figure Description

[0060] Figure 1 This is a flowchart of a line detection method for a live-line working robot according to an embodiment of the present invention;

[0061] Figure 2 This is a flowchart illustrating the detection of temperature changes according to an embodiment of the present invention;

[0062] Figure 3 This is a flowchart illustrating how to determine whether to adjust the first standard temperature according to an embodiment of the present invention.

[0063] Figure 4 This is a flowchart illustrating the process of obtaining the image to be detected in an embodiment of the present invention. Detailed Implementation

[0064] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0065] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0066] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate the direction or positional relationship, are based on the direction or positional relationship shown in the drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0067] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0068] Please see Figure 1 ,like Figure 1 As shown, it is a flowchart of the line detection method for a live-line working robot according to an embodiment of the present invention;

[0069] Specifically, embodiments of the present invention provide a method for detecting the path of a live-line working robot, comprising:

[0070] Step S1: Obtain the image to be processed in the distribution network line area, and preprocess the image to be processed, including denoising, adjusting pixel values ​​and adjusting brightness, to obtain the image to be detected;

[0071] Step S2: Analyze the image to be detected to identify the two-dimensional coordinates of the row lines, and determine the row line connection, insulator and connector based on the row line position;

[0072] Step S3: Detect the first temperature at the connection point of the line and compare it with the preset first standard temperature. Based on the comparison result, determine whether the temperature is abnormal. If the temperature is abnormal, clean the surface of the connection point and calculate the temperature change after cleaning.

[0073] Step S4: Compare the temperature change with the preset temperature change threshold. Based on the comparison result, determine whether the surface dust is abnormal. If it is determined that the surface dust is not abnormal, check whether the insulator is damaged. If it is damaged, replace it. If it is not damaged, check whether the insulator connector is abnormal.

[0074] Step S5: If the insulator connector is abnormal, the connector is replaced. If the insulator connector is normal, the processing temperature of the next row line connection is detected and compared with the first temperature to determine whether an abnormality is detected. If an abnormality is detected, a second standard temperature is calculated and the second standard temperature is used instead of the first standard temperature to compare and judge the temperature of the remaining row line connections.

[0075] Specifically, at the power distribution network, a live-line robot acquires images of the network and uses the YOLO model and Ground-SAM2 algorithm to detect and identify the network lines. The robot then moves to the lines to identify insulators and connectors. Insulators are electrical devices in the power distribution network that support conductors and isolate them from supports, primarily providing electrical insulation and mechanical support. Connectors are interface components used to connect conductors and cables. Temperature checks are performed at the line connections to determine if there are any abnormalities in the transmission. If an abnormality is detected, the surface of the line is cleaned of dust and other substances, and the temperature is checked again to verify if the problem is due to dust accumulation. If the temperature remains abnormal after cleaning, the insulators and connectors are checked to determine if there is a problem. If no problems are found, the temperature of other connections is used to verify this connection. If the temperature of other connections is also abnormal, it indicates a large power transmission volume, and the first standard temperature is replaced to avoid misjudgment. If the temperatures of other connections are normal, it indicates a potential problem with this particular line.

[0076] Specifically, by combining target detection models and temperature diagnostic technology, intelligent and precise detection of power distribution network line areas is achieved. It can automatically identify line locations, insulators, and connectors, and quickly locate and handle abnormalities. It can not only detect temperature anomalies but also further determine whether the problem is due to surface dust, and detect abnormalities on the surface of insulators and connectors. The diagnostic accuracy is high and the reliability is strong. Through comprehensive and detailed detection of line connections, potential equipment hazards can be discovered in a timely manner, enabling preventive maintenance of power distribution network equipment, reducing the risk of equipment failure, and significantly improving the accuracy, autonomy, and preventiveness of power distribution network equipment detection.

[0077] Specifically, step S2 includes:

[0078] Step S21: Analyze the image to be detected using the Ground-SAM2 algorithm and the YOLO model respectively to identify the two-dimensional coordinates of the line, calculate the coordinate deviation based on the identification results, and compare the coordinate deviation with the preset standard deviation threshold.

[0079] When the coordinate deviation is greater than the standard deviation threshold, the YOLO model is trained using the image to be detected.

[0080] The identification is determined to be correct when the coordinate deviation is less than the standard deviation threshold.

[0081] Step S22: Determine the line position coordinates based on the two-dimensional coordinates of the line and the point cloud data of the distribution network line area. Based on the line position coordinates, determine the line connection, insulator and connector at the line using a deep learning model.

[0082] Specifically, a YOLO model is loaded using PyTorch, and the YOLO model and the Ground-SAM2 algorithm are used to identify the two-dimensional coordinates of the line. The position coordinates are the centerline coordinates of the identified line. The coordinate deviation is determined by calculating the offset of the centerline coordinates identified by the YOLO model and the Ground-SAM2 algorithm. If the deviation exceeds a preset standard deviation threshold, it indicates that the YOLO model's recognition is not accurate enough. The YOLO model is then trained using the image to be detected to ensure recognition accuracy. After accurately identifying the line coordinates, the robot is controlled to move to the line and perform... The system uses a deep learning model to identify line connections, insulators, and connectors. In practice, the preset standard deviation threshold is 3. The YOLO model identifies the line center coordinates as (100, 200), while the Ground-SAM2 algorithm identifies the center coordinates as (102, 198). The calculation process involves taking the square root of the sum of the squares of the differences between the horizontal and vertical coordinates of the two datasets, resulting in a coordinate deviation of 2.83, which is less than the standard deviation threshold, confirming accurate identification. The system then combines the point cloud data of the power distribution network line area to determine the line position coordinates and controls the robot to reach the identified line position to perform the operation.

[0083] Specifically, by using a YOLO model fine-tuned from the Ground-SAM2 model to analyze image recognition of line lines, the accuracy and reliability of detection can be improved. The target detection model has a fast computation speed in the field of object detection. This method uses the target detection model to quickly analyze the image to be detected, which helps to improve detection efficiency and real-time performance, and helps to improve the reliability and safety of the power system.

[0084] Please continue reading. Figure 2 ,like Figure 2 As shown, it is a flowchart of the temperature change detection in an embodiment of the present invention;

[0085] Specifically, step S3 includes:

[0086] Step S31: Detect the first temperature at the connection point of the row lines;

[0087] Step S32: Compare the first temperature with the first standard temperature. If the first temperature is greater than the first standard temperature, then determine that the temperature is abnormal.

[0088] Step S33: When an abnormal temperature is detected, use a cleaning device to clean the surface of the connection and detect the processing temperature of the line connection after cleaning.

[0089] Step S34: Calculate the difference between the first temperature and the processing temperature to obtain the temperature change.

[0090] In the specific implementation process, the first standard temperature is set to 75°C. The first temperature is detected to be 85°C. Since the first temperature is higher than the first standard temperature, it is determined that the temperature is abnormal. The robot uses a cleaning device to clean the surface of the connection. After cleaning, the temperature is detected to be 82°C.

[0091] Specifically, by accurately detecting the temperature at the connection point of the line and comparing it with the preset standard temperature, abnormal equipment temperature can be quickly and accurately identified, providing a key basis for subsequent fault diagnosis. The temperature anomaly detection and surface cleaning process can promptly detect and handle potential equipment hazards, effectively provide early warning of possible power equipment failures, improve equipment operational reliability, and realize a fully automated process of temperature detection, anomaly judgment and surface cleaning, reducing manual intervention and improving the autonomous maintenance capability of live-line working robots.

[0092] Specifically, in step S4, the process of determining whether the surface dust is abnormal based on the comparison results includes:

[0093] If the temperature change is less than the preset temperature change threshold, it is determined that it is not an abnormality of surface dust.

[0094] If the temperature change is greater than the preset temperature change threshold, it is determined that the surface dust is abnormal.

[0095] In the specific implementation process, the preset temperature change threshold is 10°C. The first temperature detected is 85°C. The temperature after surface cleaning is 82°C. The temperature change is 3°C, which is less than the temperature change threshold of 10°C. Therefore, it is determined that the surface dust is not abnormal.

[0096] Specifically, by comparing temperature change thresholds, it is possible to accurately distinguish whether surface dust is abnormal. By measuring the temperature change before and after cleaning, it is possible to capture minute temperature differences and detect even slight surface dust accumulation in a timely manner, thus improving the sensitivity of detection. The preset temperature change threshold can be adjusted according to the actual environment and equipment characteristics, making the dust abnormality judgment method highly adaptable and versatile.

[0097] Specifically, in step S4, the process of detecting whether the insulator is damaged includes:

[0098] Obtain a surface image of the insulator;

[0099] The surface image is input into a preset convolutional neural network model;

[0100] The surface image is analyzed by the convolutional neural network model to determine whether the insulator is damaged.

[0101] Specifically, a pre-set convolutional neural network is used to analyze surface images to identify whether the insulator is damaged.

[0102] Specifically, through deep learning technology, the model can automatically learn and extract features from insulator surface images to detect surface damage. This enables high-precision identification of minute cracks, defects, and damage on the insulator surface, allowing for automatic early warning and precise location of insulator damage. This provides accurate information for subsequent maintenance and replacement, improving detection efficiency and reliability, and preventing equipment failures in advance.

[0103] Specifically, in step S4, the process of detecting whether the connector is abnormal includes:

[0104] Remove the insulator at the line connection to expose the connector:

[0105] The voltage of the connector is detected to obtain the voltage to be measured.

[0106] The voltage to be measured is compared with a preset standard voltage to obtain a comparison result. Based on the comparison result, it is determined whether the connector is abnormal.

[0107] If the voltage to be tested is greater than the standard voltage, then the connector is determined to be abnormal.

[0108] If the voltage to be tested is less than or equal to the standard voltage, then the connector is considered to be normal.

[0109] In the specific implementation process, the preset standard voltage is 10kV. The robot detects the voltage of the connector through a voltage detector. If the voltage to be tested is detected to be 12kV, then the voltage to be tested 12kV is greater than the standard voltage 10kV, and the connector is judged to be abnormal.

[0110] Specifically, by directly removing the insulator and performing voltage testing on the connector, a precise diagnosis of the connector's internal condition is achieved. By judging voltage anomalies, potential hidden dangers in the connector can be detected in a timely manner, such as increased contact resistance or insulation damage that may lead to voltage anomalies. This enables accurate identification and early warning in the early stages of fault development, which is beneficial for early detection of potential equipment risks and improving power supply reliability.

[0111] Please continue reading. Figure 3 ,like Figure 3 As shown, it is a flowchart of an embodiment of the present invention for determining whether to adjust the first standard temperature;

[0112] Specifically, step S5 includes:

[0113] Step S51, move to the next connection point of the row line;

[0114] Step S52: Detect the second temperature at the next connection point;

[0115] Step S53: Calculate the temperature difference between the first temperature and the second temperature, compare the temperature difference with a preset temperature difference threshold, and determine whether the detection is normal based on the comparison result.

[0116] Step S54: Calculate the second standard temperature when an abnormality is detected.

[0117] Specifically, after verifying that there are no problems with the insulators and connectors, the robot moves to the next connection point to perform temperature detection. Based on the difference between the detected second temperature and the first temperature, it is determined whether there is a problem with this row of lines or whether the temperature rise is normal due to excessive power consumption at this time.

[0118] Specifically, by calculating the temperature difference between adjacent connections and comparing it with a preset threshold, a comprehensive and systematic inspection of the power distribution network lines is achieved. This enables more sensitive detection of abnormal temperature changes, improves the accuracy of fault identification, and avoids potential problems that may be missed by single-point detection. By comparing the temperature difference with the preset threshold, it is possible to more accurately determine whether there are abnormalities at the connection, effectively distinguish between normal temperature fluctuations and actual faults, and reduce the false judgment rate.

[0119] Specifically, the process of comparing the temperature difference with a preset temperature difference threshold to determine whether the detection is normal in step S53 includes:

[0120] If the temperature difference is less than or equal to the temperature difference threshold, then an anomaly is determined.

[0121] If the temperature difference is greater than the temperature difference threshold, the detection is considered normal.

[0122] In the specific implementation process, the preset temperature difference threshold is 6℃, the first temperature is 85℃, and the second temperature at the next connection point is 83℃. The calculated temperature difference is 2℃, which is less than the preset temperature difference threshold. Therefore, the detection is judged to be abnormal. At this time, the high power consumption and high temperature are normal.

[0123] Specifically, by setting a temperature difference threshold, it is possible to accurately distinguish whether the temperature change at the connection point of the line is within the normal range, making the anomaly detection more objective and scientific. The preset temperature difference threshold can be flexibly adjusted according to the actual application scenario and equipment characteristics, which can improve the sensitivity of detection and avoid false alarms caused by oversensitivity. Through accurate judgment of temperature difference, the accuracy and reliability of detection are improved, and potential equipment problems can be detected early, enabling early warning maintenance of the distribution network lines and reducing the risk of sudden equipment failure.

[0124] Specifically, in step S54, the process of calculating the second standard temperature when an anomaly is detected includes:

[0125] When an anomaly is detected, the average temperature of the first temperature and the second temperature is calculated.

[0126] The average temperature is set as the second standard temperature.

[0127] In the specific implementation process, the first temperature is 85℃, and the second temperature at the next connection point is 83℃. At this point, an abnormality is determined, and the average temperature of the first and second temperatures is calculated to be 84℃, which is used as the second standard temperature for comparison and verification of the remaining connections.

[0128] Specifically, by calculating the average of the first and second temperatures and dynamically adjusting the standard temperature, the detection system achieves intelligent self-adaptation to different environmental and equipment temperature characteristics. The dynamic temperature standard adjustment mechanism can effectively eliminate misjudgments caused by factors such as environmental temperature fluctuations and equipment aging. By using the average calculation method, it can more accurately identify real temperature anomalies and reduce false alarm and false negative rates.

[0129] Please continue reading. Figure 4 ,like Figure 4 As shown, it is a flowchart of obtaining the image to be detected according to an embodiment of the present invention;

[0130] Specifically, step S1 includes:

[0131] Step S11: Acquire the image to be processed in the power distribution network line area using a camera;

[0132] Step S12: Perform noise smoothing and denoising on the image to be processed, adjust the pixel values ​​to standard pixel values, and adjust the image brightness to standard brightness to obtain the image to be detected.

[0133] Specifically, the camera installed on the live-line working robot takes pictures of the power distribution network's overhead lines. Image processing algorithms such as Gaussian blur and median filtering are used to denoise the acquired images, adjusting the pixel values ​​and brightness to the corresponding standard pixel values ​​and brightness to reduce errors caused by environmental interference when analyzing the images.

[0134] Specifically, noise smoothing and denoising technology effectively suppresses random noise and interference signals in the image, significantly improving image clarity and contrast, providing a high-quality image foundation for subsequent target recognition. Standardized preprocessing enhances the consistency and stability of images in different environments, improving the detection accuracy of the live-line working robot in different working scenarios.

[0135] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0136] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A line following method for an electrically live robot, characterized in that, The method comprises the following steps: Step S1, obtaining a to-be-processed image of a power distribution network line area, and preprocessing the to-be-processed image, including denoising, adjusting pixel values and adjusting brightness, to obtain a to-be-detected image; Step S2, analyzing the to-be-detected image to identify line two-dimensional coordinates, and determining line connection points, insulators and connecting heads based on line position; Step S3, detecting a first temperature of the line connection point, comparing the first temperature with a preset first standard temperature, and determining whether the temperature is abnormal based on the comparison result; when it is determined that the temperature is abnormal, the surface of the connection point is cleaned, and the temperature change after cleaning is calculated; Step S4, comparing the temperature change with a preset temperature change threshold, and determining whether the surface dust is abnormal based on the comparison result; when it is determined that the surface dust is not abnormal, it is detected whether the insulator is damaged, and if the insulator is damaged, the insulator is replaced; if the insulator is not damaged, it is detected whether the insulator connecting head is abnormal; Step S5, if the insulator connecting head is abnormal, the connecting head is replaced; if the insulator connecting head is normal, a second temperature of the next line connection point is detected, and compared with the first temperature to determine whether the detection is abnormal; and when it is determined that the detection is abnormal, a second standard temperature is calculated, and the first standard temperature is replaced by the second standard temperature to compare the temperatures of the remaining line connection points.

2. The live-working robot row line detection method of claim 1, wherein In the step S2, the method comprises the following steps: Step S21, analyzing the to-be-detected image by using a Ground-SAM2 algorithm and a YOLO model respectively to identify line two-dimensional coordinates, calculating coordinate deviations based on the identification result, and comparing the coordinate deviations with a preset standard deviation threshold; When the coordinate deviation is greater than the standard deviation threshold, the YOLO model is feedback trained by using the to-be-detected image; When the coordinate deviation is less than the standard deviation threshold, it is determined that the identification is correct; Step S22, determining line position coordinates based on the line two-dimensional coordinates and point cloud data of the power distribution network line area, and determining the line connection point, the insulator and the connecting head at the line by using a deep learning model based on the line position coordinates.

3. The live-working robot row line detection method of claim 1, wherein In the step S3, the method comprises the following steps: Step S31, detecting the first temperature of the line connection point; Step S32, comparing the first temperature with the first standard temperature; if the first temperature is greater than the first standard temperature, it is determined that the temperature is abnormal; Step S33, when it is determined that the temperature is abnormal, the surface of the connection point is cleaned by using a cleaning device, and the processing temperature of the line connection point after cleaning is detected; Step S34, calculating the difference between the first temperature and the processing temperature to obtain the temperature change.

4. The live-working robot row line detection method of claim 1, wherein In the step S4, the process of determining whether the surface dust is abnormal based on the comparison result comprises: If the temperature change is less than the preset temperature change threshold, it is determined that the surface dust is not abnormal; If the temperature change is greater than the preset temperature change threshold, it is determined that the surface dust is abnormal.

5. The live working robot line detection method according to claim 4, characterized in that, In the step S4, the process of detecting whether the insulator is damaged comprises: Obtaining a surface image of the insulator; Inputting the surface image into a preset convolutional neural network model; Identifying and analyzing the surface image by using the convolutional neural network model to obtain a result of whether the insulator is damaged.

6. The live-working robot row line detection method of claim 5, wherein, In the step S4, the process of detecting whether the connector is abnormal includes: removing the insulator at the connection of the line to expose the connector; performing voltage detection on the connector to obtain a to-be-detected voltage; comparing the to-be-detected voltage with a preset standard voltage to obtain a comparison result, and determining whether the connector is abnormal based on the comparison result, wherein if the to-be-detected voltage is greater than the standard voltage, it is determined that the connector is abnormal; if the to-be-detected voltage is less than or equal to the standard voltage, it is determined that the connector is normal.

7. The live-working robot row line detection method of claim 1, wherein, In the step S5, it includes: step S51, moving to the next connection of the line; step S52, detecting a second temperature at the next connection; step S53, calculating a temperature difference between the first temperature and the second temperature, comparing the temperature difference with a preset temperature difference threshold, and determining whether the detection is abnormal according to the comparison result; step S54, calculating the second standard temperature when it is determined that the detection is abnormal.

8. The live working robot line detection method according to claim 7, characterized in that, The process of comparing the temperature difference with the preset temperature difference threshold to determine whether the detection is normal in the step S53 includes: if the temperature difference is less than or equal to the temperature difference threshold, it is determined that the detection is abnormal; if the temperature difference is greater than the temperature difference threshold, it is determined that the detection is normal.

9. The live-working robot row line detection method of claim 7, wherein, The process of calculating the second standard temperature when it is determined that the detection is abnormal in the step S54 includes: when it is determined that the detection is abnormal, calculating a temperature average of the first temperature and the second temperature; setting the temperature average as the second standard temperature.

10. The live-working robot row line detection method of claim 1, wherein, In the step S1, it includes: step S11, obtaining the to-be-processed image of the line area of the power distribution network through a camera; step S12, performing noise smoothing denoising on the to-be-processed image, adjusting the pixel value to a standard pixel value, and adjusting the image brightness to a standard brightness to obtain the to-be-detected image.

Images