Intelligent maintenance methods and systems for LED displays using maintenance robots

Abstract

This invention discloses an intelligent repair method and system for LED displays using a repair robot. The invention relates to the technical field of repair robots. It determines the surface fault characteristics and model of the corresponding LED display based on image detection of the table surface; determines the internal fault characteristics of the LED display based on the model number; determines an optimized fault procedure list for the LED display based on the surface and internal fault characteristics; determines intelligent repair events for each LED display using the repair robot's repair arm based on the optimized fault procedure list, the LED display model, and the LED display's location; and determines optimization measures for the repair progress based on two repair schedules, the repair robot's task sequence list, and the current time. These optimization measures include batch repair of the final repair procedures for each LED display, coordination with external collaborative equipment, or adjusting the operating efficiency of the repair arm, thereby improving the accuracy of the optimization measures for the repair progress.

Description

Technical Field

[0001] This invention relates to the technical field of maintenance robots, and in particular to an intelligent maintenance method and system for LED displays using maintenance robots. Background Technology

[0002] With the development of technology, LED displays are gradually being applied to people's lives and are also used as industrial equipment. During the production process, LED displays may have some surface or internal faults, requiring them to be placed on an LED display maintenance station. In the current technology, LED displays on the maintenance station are repaired using a single repair device. However, a single repair device can only repair surface faults of LED displays and cannot be compatible with repairing internal faults, nor can it optimize the repair progress for LED displays on different maintenance stations. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides an intelligent repair method and system for LED displays using a repair robot.

[0004] This invention provides an intelligent maintenance method for LED displays using a maintenance robot, comprising:

[0005] The maintenance robot is surrounded by two LED display maintenance stations. The robot's camera captures images of the two stations in a circular motion.

[0006] In two tabletop images, the surface fault characteristics and model of the corresponding LED display screen are determined based on image detection of the tabletop images;

[0007] The internal fault characteristics of the LED display screen are determined by tracing its model; the fault optimization process table for the LED display screen is determined based on its surface and internal fault characteristics.

[0008] Based on the fault optimization process list, the model of the LED display screen and the location of the LED display screen, the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display screen are determined, and the maintenance progress of the two maintenance arms relative to the corresponding LED display screen is collected.

[0009] Based on the two maintenance schedules, the task sequence list of the maintenance robot, and the current time, optimization measures for the maintenance schedule are determined. These optimization measures include batch maintenance of the final maintenance process for each LED display, coordination with external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm.

[0010] This invention provides an intelligent repair system for LED displays using a repair robot. This intelligent repair system is applied to the aforementioned intelligent repair method for LED displays using a repair robot. The intelligent repair system for LED displays includes:

[0011] The tabletop image module is used to capture images of the two LED display maintenance tables around the maintenance robot by taking a circular shot of the two LED display maintenance tables using the maintenance robot's camera.

[0012] The surface fault module is used to determine the surface fault characteristics and model of the corresponding LED display screen based on image detection of the two table surface images;

[0013] The fault optimization process table module is used to determine the internal fault characteristics of the LED display screen based on the model of the LED display screen; and to determine the fault optimization process table of the LED display screen based on the surface fault characteristics and internal fault characteristics of the LED display screen.

[0014] The intelligent maintenance module is used to determine the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display based on the fault optimization process list, the model of the LED display, and the location of the LED display, and to collect the maintenance progress of the two maintenance arms relative to the corresponding LED display.

[0015] The optimization measures module is used to determine optimization measures for the maintenance progress based on the two maintenance progresses, the task sequence list of the maintenance robot, and the current time. These optimization measures include batch maintenance of the final maintenance process of each LED display, coordination of external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm.

[0016] Compared with the prior art, the beneficial effects of the present invention are:

[0017] In this embodiment of the invention, the method involves two LED display maintenance stations surrounding a maintenance robot. Two station images are captured by the robot's camera through a circular motion. Based on image detection in the two station images, the surface fault characteristics and model of the corresponding LED display are determined. The internal fault characteristics of the LED display are determined by tracing its model. A fault optimization process table for the LED display is determined based on its surface and internal fault characteristics. This approach incorporates both surface and internal fault characteristics, achieving a holistic consideration of both and improving the accuracy of the fault optimization process table.

[0018] Therefore, based on the fault optimization process table, the model of the LED display screen, and the location of the LED display screen, the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display screen are determined, and the maintenance progress of the two maintenance arms relative to the corresponding LED display screen is collected. Based on the two maintenance progresses, the task sequence table of the maintenance robot, and the current time, optimization measures for the maintenance progress are determined. These optimization measures include batch maintenance of the final maintenance process of each LED display screen, coordination with external collaborative equipment, or adjustment of the movement efficiency of the maintenance arms. The maintenance progress of the two maintenance arms relative to the corresponding LED display screen is introduced, which is compatible with the overall consideration of the maintenance progress of the two maintenance arms relative to the corresponding LED display screen, the task sequence table of the maintenance robot, and the current time. This improves the accuracy of the optimization measures for the maintenance progress and realizes that one maintenance robot can perform synchronous and intelligent maintenance on LED display screens at different LED display maintenance stations. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the intelligent repair method for LED displays using a repair robot according to an embodiment of the present invention.

[0020] Figure 2 This is a flowchart illustrating step S11 of the intelligent repair method for LED displays using a repair robot in an embodiment of the present invention.

[0021] Figure 3 This is a flowchart illustrating step S12 of the intelligent repair method for LED displays using a repair robot in an embodiment of the present invention.

[0022] Figure 4 This is a flowchart illustrating step S13 of the intelligent repair method for LED displays using a repair robot in an embodiment of the present invention.

[0023] Figure 5 This is a flowchart illustrating step S14 of the intelligent repair method for LED displays using a repair robot in an embodiment of the present invention.

[0024] Figure 6 This is a flowchart illustrating step S15 of the intelligent repair method for LED displays using a repair robot in an embodiment of the present invention.

[0025] Figure 7 This is a schematic diagram of the structural composition of the intelligent maintenance system for LED displays using a maintenance robot, as described in this embodiment of the invention. Detailed Implementation

[0026] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0027] Please see Figures 1 to 7 A method for intelligent maintenance of LED displays using a maintenance robot, applied to intelligent maintenance scenarios of maintenance robots; the method includes:

[0028] Step S11: Two LED display maintenance stations are set up around the maintenance robot. The two station surfaces are captured by the maintenance robot's camera taking a circular shot.

[0029] Step S12: In the two tabletop images, determine the surface fault characteristics and model of the corresponding LED display screen based on image detection of the tabletop images;

[0030] Step S13: Determine the internal fault characteristics of the LED display screen based on the model number; determine the fault optimization process table for the LED display screen based on the surface and internal fault characteristics of the LED display screen;

[0031] Step S14: Based on the fault optimization process list, the model of the LED display screen and the location of the LED display screen, determine the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display screen, and collect the maintenance progress of the two maintenance arms relative to the corresponding LED display screen.

[0032] Step S15: Determine optimization measures for the maintenance progress based on the two maintenance progresses, the task sequence list of the maintenance robot, and the current time; these optimization measures include batch maintenance of the final maintenance process of each LED display, coordination of external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm.

[0033] refer to Figure 2 In step S11, two LED display maintenance stations are set up around the maintenance robot, and images of the two stations are collected by the camera of the maintenance robot taking a circular shot of the two LED display maintenance stations.

[0034] In the specific implementation of this invention, the specific steps are as follows:

[0035] S111: Collect the location of the LED display maintenance station and determine another LED display maintenance station based on the surrounding detection of the LED display maintenance station. At this time, the two LED display maintenance stations are arranged side by side, and a maintenance station is set between the two LED display maintenance stations. The maintenance robot is located at the maintenance station.

[0036] S112: Construct a corresponding inspection space area based on the maintenance robot and two LED display maintenance stations; the maintenance robot is equipped with a camera, and the circular movement path of the camera is determined based on the spatial position of the camera, the spatial position of the two LED display maintenance stations, and the range of motion of the maintenance robot;

[0037] S113: The camera moves along a circular path driven by the maintenance robot and dynamically captures images of the two LED display maintenance platforms.

[0038] In the embodiments of this application, the maintenance robot needs to be equipped with high-precision sensors to acquire the position of the maintenance station. Typically, LiDAR, ultrasonic sensors, or vision sensors (such as cameras) can be used. The sensors scan the surrounding environment to identify the outline and position of the maintenance station. If a vision sensor is used, image recognition technology can be combined to determine the position of the maintenance station through feature point matching or template matching. The data acquired by the sensors is converted into coordinate information and stored in the robot's control system. Optionally, it is assumed that the maintenance robot is equipped with a LiDAR sensor. LiDAR calculates distance by emitting a laser beam and measuring the time difference of the reflected light. After the robot starts at the maintenance station, the LiDAR begins scanning the surrounding environment; the LiDAR performs a 360-degree scan centered on the maintenance robot; the robot control system analyzes the scan data to identify the outline of the maintenance station. Assuming the maintenance station is a rectangular structure, the LiDAR can detect the coordinates of its four corners; the system converts the coordinates of the four corners of the maintenance station into the coordinates of the center point of a rectangle, for example, the center point coordinates of the first maintenance station are (X1, Y1); simultaneously, the system records the size information of the maintenance station, such as width W1 and length L1.

[0039] After determining the location of the first maintenance station, the robot needs to continue scanning its surrounding area to find the other maintenance station. This can be achieved by expanding the scanning range of the LiDAR or adjusting the camera's shooting angle. Since the two maintenance stations are arranged side-by-side, the location of the second maintenance station usually has a spatial relationship with the first. For example, they may be on the same straight line or at a fixed angle. The location information of the first maintenance station is fused with the detection data of the second maintenance station to ensure the accurate positional relationship between the two stations. Optionally, after determining the location of the first maintenance station, the robot adjusts the scanning direction of the LiDAR and continues scanning the area to the right of the first maintenance station. The LiDAR scans the area to the right of the first maintenance station, detecting the outline of the second maintenance station. The system analyzes the scan data, determines the coordinates of the four corners of the second maintenance station, and calculates its center point coordinates as (X2, Y2). The system records the dimensions of the second maintenance station, such as width W2 and length L2. By comparing the center point coordinates of the two maintenance stations, it confirms that they are arranged side-by-side. For example, X2 = X1 + W1 + spacing.

[0040] The maintenance station is the reference position for the maintenance robot to operate, typically located between two maintenance stations. The location of the maintenance station needs to be pre-defined, and the robot must be able to navigate to that position accurately. The two maintenance stations are arranged side-by-side, with the maintenance station situated between them, ensuring sufficient operating space for the robot. This layout facilitates the robot's operation of both maintenance stations while avoiding collisions with other equipment. Safety devices, such as light curtains or emergency stop buttons, are installed between the maintenance station and the maintenance stations to prevent accidents during robot movement.

[0041] Furthermore, the detection space area refers to the three-dimensional spatial range that the maintenance robot's camera can cover, used to photograph the surfaces of the two LED display maintenance stations. The boundaries of the detection space are determined based on the spatial positions of the maintenance robot and the two maintenance stations. This space needs to encompass the surfaces of both maintenance stations, taking into account the camera's shooting angle and range. A three-dimensional model of the environment, including the outline and height of the maintenance stations and the position of the maintenance robot, is acquired through sensors (such as LiDAR or cameras).

[0042] Determine the camera's mounting position and initial orientation on the maintenance robot. This can be achieved through sensor calibration or mechanical design. Based on the range of motion of the robot's arm, determine the camera's possible movement path. The range of motion needs to consider the joint angle limitations of the arm and collision detection. Based on the detection space area and the camera's range of motion, plan the camera's circular movement path. The path must ensure that the camera can capture images of the surfaces of both maintenance stations from multiple angles.

[0043] Therefore, the camera moves along a circular path driven by the maintenance robot and dynamically captures images of the two LED display maintenance platforms to obtain images of the two platforms.

[0044] At this point, the camera needs to be mounted on the end of the robotic arm or other movable part of the maintenance robot to ensure that it can move on the circular path. Before shooting, the camera needs to be initialized and calibrated, including focal length adjustment, aperture setting, white balance calibration, etc., to ensure the quality of the captured images. If the robot is equipped with other sensors (such as LiDAR or ultrasonic sensors), the camera's shooting data needs to be fused with the data from other sensors to improve the accuracy of positioning and shooting.

[0045] Based on the circular movement path determined in step S112, plan the motion trajectory of the robotic arm. The path needs to ensure that the camera can capture images of the maintenance platform from multiple angles. Through the robot's control system, control the robotic arm to move smoothly along the planned path. The limitations of the robotic arm's movement speed, acceleration, and joint angles need to be considered to ensure the smoothness and accuracy of the movement. During the movement, adjust the camera's shooting parameters (such as focal length, aperture, and exposure time) in real time based on the actual shooting results (such as image clarity and lighting conditions).

[0046] The camera is triggered to capture images as the robotic arm moves to each shooting point. This triggering can be achieved via the robotic arm's position sensor or time control. The captured images are transmitted in real-time to the robot's control system and stored. Image compression techniques can be used to reduce data transmission while ensuring image quality. Preprocessing of the acquired images, including noise reduction, contrast enhancement, and edge detection, improves the accuracy of subsequent image analysis. Optionally, assuming a 30-degree shooting interval at each shooting point, the camera is triggered to capture images as the robotic arm moves to each point. When the robotic arm reaches a designated position, the position sensor detects that the robotic arm has reached the designated position, triggering the camera to capture images. The captured images are transmitted to the robot's control system via a data cable and stored on the hard drive. JPEG compression is used to compress the images to 10% of their original size to reduce storage space and transmission time. Preprocessing of the acquired images includes using Gaussian filtering to remove image noise, adjusting contrast to enhance image details, and using edge detection algorithms to extract the contours of the maintenance platform.

[0047] refer to Figure 3 In step S12, the surface fault characteristics and model of the corresponding LED display screen are determined based on image detection of the two tabletop images.

[0048] In the specific implementation of this invention, the specific steps are as follows:

[0049] S121: Real-time monitoring of two tabletop images, image processing of the two tabletop images, at this time, the image processing of each tabletop image divides into multiple image areas, and the surface image and model identification image of the LED display screen are determined based on the filtering of multiple image areas;

[0050] S122: If the surface image and model identification image of the LED display screen are in the same area of ​​the table surface image, then multi-channel recognition is performed based on the surface image and model identification image of the LED display screen; in the multi-channel recognition of the surface image and model identification image of the LED display screen, the surface shape and model content of the LED display screen are output, and the abnormal surface area is determined according to the detection of the surface shape of the LED display screen.

[0051] S123: Based on the identification of abnormal surface areas, multiple sub-surface fault features are determined, and the surface fault features of the LED display screen are determined based on the location, shape, and surface fault feature mapping relationship of the multiple sub-surface fault features.

[0052] In the embodiments of this application, the camera of the maintenance robot captures real-time images of the surface of the maintenance platform of the two LED displays. The captured image data is transmitted in real-time to the robot's control system via a high-speed interface (such as USB 3.0 or Ethernet). The captured images are then displayed in real-time in the robot's control system for monitoring and subsequent processing.

[0053] Image processing is performed on the two tabletop images. Assuming the resolution of the captured tabletop images is 1920x1080 pixels, a Gaussian filter is applied to each captured image with a filter kernel size of 5x5 and a standard deviation of 1.5. Histogram equalization is used to adjust the contrast of the images to make the details in the images clearer. The Canny algorithm is used to extract the edge information in the images, with a low threshold of 50 and a high threshold of 150.

[0054] The image of the display surface is divided into multiple equally sized grid regions. For example, the image can be divided into a 4x4 or 8x8 grid, dynamically partitioning the image regions based on edge information or other features. Alternatively, image segmentation algorithms (such as region growing or watershed algorithms) can be used to partition the regions. Optionally, regions can be dynamically partitioned starting from a seed point in the image based on pixel similarity (such as color or grayscale value). Within each partitioned image region, image recognition algorithms (such as template matching or feature point detection) are used to identify the surface and model identifier of the LED display screen. Based on the recognition results, image regions containing the LED display screen surface and model identifier are selected.

[0055] Optionally, assume the maintenance robot captures the following image of the work surface:

[0056] Maintenance console 1: Image resolution: 1920x1080 pixels; Grid division: 8x8 grid, each grid area is 240x135 pixels; Feature recognition results: Display screen surface area: 3rd row and 4th column; Model identification area: 5th row and 6th column;

[0057] Maintenance console 2: Image resolution: 1920x1080 pixels; Grid division: 8x8 grid, each grid area is 240x135 pixels; Feature recognition results: Display screen surface area: 2nd row and 3rd column; Model identification area: 4th row and 5th column.

[0058] Furthermore, if the surface image and model identification image of the LED display screen are located in the same area of ​​the table surface image, multi-channel recognition is performed based on the surface image and model identification image of the LED display screen. In the multi-channel recognition of the surface image and model identification image of the LED display screen, the surface shape and model content of the LED display screen are output. The abnormal surface area is determined based on the detection of the surface shape of the LED display screen, which is compatible with the overall consideration of the detection of the surface shape of the LED display screen and ensures the accuracy of the abnormal surface area.

[0059] At this point, the image area selected in step S121 is examined to determine whether there is an area that simultaneously contains the LED display surface and the model identifier. If the surface image and the model identifier image are in the same area, the location and extent of the area are recorded for subsequent multi-channel recognition. Multi-channel processing is performed on the image in the same area, including color channels, grayscale channels, edge channels, etc. Features such as color distribution, texture features, and edge intensity are extracted from each channel. The features of different channels are fused and analyzed to improve the accuracy and robustness of recognition.

[0060] Optionally, assuming that the surface image and model identification image of the LED display screen are detected simultaneously in the image area of ​​the 3rd row and 4th column of maintenance station 1, check each selected image area to confirm whether there is an area that contains both the surface and model identification, and record the location (3rd row and 4th column) and range (240x135 pixels) of the area; extract information from the three RGB color channels; convert the image to a grayscale image and extract grayscale information; use the Canny algorithm to extract edge information; calculate the histogram distribution of each color channel; use the grayscale co-occurrence matrix (GLCM) to extract texture features; count the number and distribution of edge pixels, and fuse the color, grayscale and edge features to form a comprehensive feature vector.

[0061] Surface morphology recognition: Pre-trained deep learning models (such as convolutional neural networks CNN) are used to identify the surface morphology of the LED display screen, including scratches, stains, and damage. Simultaneously, model information recognition: Optical character recognition (OCR) technology is used to identify the model identification content. The identified surface morphology and model information are output to the control system. Based on the surface morphology detection results, image segmentation algorithms (such as threshold segmentation, region growing, etc.) are used to identify abnormal areas on the surface. The detected abnormal areas are marked, and their positions and sizes are recorded. The abnormal areas are further analyzed to extract their features (such as length, area, shape, etc.).

[0062] Optionally, assuming the surface morphology identified in the image region of row 3 and column 4 is "slight scratch" and the model information is "LED-2023"; using a pre-trained CNN model, the fused feature vector is input, and the output surface morphology is "slight scratch"; using OCR technology, the model information is identified as "LED-2023"; the recognition result is output to the control system, recording the surface morphology as "slight scratch" and the model information as "LED-2023"; using a threshold segmentation algorithm, regions in the grayscale image with grayscale values ​​higher than a certain threshold (e.g., 128) are identified as abnormal regions, the detected abnormal regions are marked, and their positions (e.g., the upper left corner coordinates are (50, 30)) and sizes (e.g., width is 20 pixels, height is 5 pixels) are recorded, and the abnormal regions are further analyzed to extract their features, such as a length of 20 pixels, a width of 5 pixels, and a linear shape.

[0063] Therefore, multiple sub-surface fault features are determined based on the identification of abnormal surface areas. The surface fault features of the LED display screen are determined based on the location, shape, and surface fault feature mapping relationship of multiple sub-surface fault features. This comprehensive approach takes into account the location, shape, and surface fault feature mapping relationship of multiple sub-surface fault features, ensuring the accuracy of the surface fault features of the LED display screen.

[0064] At this point, multiple sub-surface fault features are extracted from the abnormal surface area, such as scratch length, scratch width, stain area, and damage shape. These extracted features are quantified, for example, by measuring the length and width of the scratch and calculating the area of ​​the stain. The sub-surface fault features are then classified according to their type (e.g., scratch, stain, damage). Based on predefined rules, the sub-surface fault features are mapped to specific surface fault features. For example, a scratch length exceeding 10 mm is considered a "severe scratch," and a stain area exceeding 0.5 square centimeters is considered a "severe stain." The extracted sub-surface fault features are matched with predefined rules to determine the final surface fault features. Combining the classification results of all sub-surface fault features, the final surface fault features are determined and output to the control system of the maintenance robot, providing a basis for subsequent maintenance decisions.

[0065] Optionally, suppose that in the image area of ​​the 3rd row and 4th column of maintenance station 1, an abnormal surface area is detected, which is manifested as a scratch and a stain. The length of the scratch is extracted to be 20 pixels and the width is 2 pixels. The area of ​​the stain is calculated to be 100 square pixels. Scratch length: 20 mm (assuming the ratio of pixels to actual size is 1:1), scratch width: 2 mm, stain area: 1 square centimeter (assuming the ratio of pixels to actual size is 1:1). Scratch: length 20 mm, width 2 mm, stain: area 1 square centimeter.

[0066] Suppose the predefined mapping rules are as follows: Scratch: length < 10 mm: minor scratch; length ≥ 10 mm: severe scratch; Stains: area < 0.5 square centimeters: minor stain; area ≥ 0.5 square centimeters: severe stain; In this case, the scratch is 20 mm long, which is judged as a "severe scratch" according to the rules, and the stain is 1 square centimeter in area, which is judged as a "severe stain" according to the rules.

[0067] In the image area of ​​the 3rd row and 4th column of maintenance station 1, after feature extraction and rule matching, the sub-surface fault features are determined as follows: Scratches: severe scratches (length 20 mm, width 2 mm); Stains: severe stains (area 1 square centimeter).

[0068] refer to Figure 4 In step S13, the internal fault characteristics of the LED display screen are determined by tracing the model of the LED display screen; and the fault optimization process table of the LED display screen is determined by tracing the surface fault characteristics and internal fault characteristics of the LED display screen.

[0069] In the specific implementation of this invention, the specific steps are as follows:

[0070] S131: Collect the model number of the LED display screen, determine the corresponding traceability path based on the model number and processing path of the LED display screen, and determine the past processing events of the LED display screen based on the traceability path, the processing database of the LED display screen and the model number of the LED display screen.

[0071] S132: Based on the identification of past processing events of the LED display screen, determine the corresponding processing abnormal events, and determine the internal fault characteristics of the LED display screen based on the process corresponding to the processing abnormal event, the processing process list of the LED display screen, and the abnormal data corresponding to the LED display screen.

[0072] S133: Determine a first fault process table based on past processing events of the LED display screen and surface fault characteristics of the LED display screen; determine a second fault process table based on past processing events of the LED display screen and internal fault characteristics of the LED display screen; and determine a fault optimization process table for the LED display screen by combining the first fault process table and the second fault process table.

[0073] In the embodiments of this application, the model of the LED display screen is collected, and the corresponding traceability path is determined according to the model of the LED display screen and the processing path of the LED display screen. Based on the traceability path, the processing database of the LED display screen and the model of the LED display screen, the past processing events of the LED display screen are determined. This approach takes into account the overall consideration of the traceability path, the processing database of the LED display screen and the model of the LED display screen, thus ensuring the accuracy of the past processing events of the LED display screen.

[0074] At this point, the model information is extracted from the model identification image of the display screen using optical character recognition (OCR) technology. The collected model information is verified to ensure its accuracy and completeness. The collected model information is stored in the control system of the maintenance robot for later use. Based on the collected model, the standard processing path of the LED display screen of that model is queried. Combined with the standard processing path, the traceability path of the display screen is generated to query its processing history, verify the completeness and accuracy of the traceability path, and ensure that all relevant processing events can be traced.

[0075] Optionally, assuming that the model identifier "LED-2023" is identified by OCR in the image area of ​​the 3rd row and 4th column of maintenance station 1, the model identifier image is identified using OCR technology to extract the model "LED-2023". The extracted model information is verified to conform to the predefined format (such as "LED-XXXX"). The model "LED-2023" is stored in the control system of the maintenance robot. The standard processing path of model "LED-2023" is queried to obtain "Production Line A -> Process 1 -> Process 2 -> Process 3". A traceability path is generated to query the processing history of the display screen on production line A. The completeness of the traceability path is verified to ensure that detailed information of each process can be traced. The standard processing path of model "LED-2023" is "Production Line A -> Process 1 -> Process 2 -> Process 3".

[0076] Access the LED display screen's processing database and query its processing history according to the traceability path. Extract the processing events for each process, including normal and abnormal events, from the database. Organize the extracted events to form a complete processing history. Optionally, assuming the database records the processing events for model "LED-2023" on production line A, access the database and enter the traceability path "Production line A -> Process 1 -> Process 2 -> Process 3". Extract the processing events for model "LED-2023" for each process: Process 1: Completed normally; Process 2: Welding temperature abnormality, recorded on June 20, 2025; Process 3: Completed normally. Organize the extracted events into a complete processing history: Process 1: Completed normally; Process 2: Welding temperature abnormality (June 20, 2025); Process 3: Completed normally.

[0077] Furthermore, based on the identification of past processing events of the LED display screen, corresponding processing abnormal events are identified. Based on the process corresponding to the processing abnormal event, the processing process list of the LED display screen, and the abnormal data corresponding to the LED display screen, the internal fault characteristics of the LED display screen are determined. This comprehensive consideration of the process corresponding to the processing abnormal event, the processing process list of the LED display screen, and the abnormal data corresponding to the LED display screen ensures the accuracy of the internal fault characteristics of the LED display screen.

[0078] At this point, abnormal processing events are identified from past processing incidents. This can be achieved by analyzing abnormal records in the processing data (such as abnormal temperatures, welding failures, etc.). The identified abnormal events are categorized, for example, welding abnormalities, temperature abnormalities, mechanical damage, etc. Specific information for each abnormal event is recorded, including the time of occurrence, process, abnormality type, and related data. The processing process list for this model of LED display is then queried to understand the specific content of each process and possible failure points. The abnormal processing events are matched with the processing process list to determine the corresponding process. Combining the specific information of the abnormal events with the processing process list, the internal fault characteristics of the LED display are deduced. For example, abnormal welding temperature may lead to incomplete solder joints or insufficient weld strength.

[0079] Optionally, suppose that in the image area of ​​row 3, column 4 of maintenance station 1, the following past processing events for model "LED-2023" are found: Process 1: Completed normally; Process 2: Abnormal welding temperature (June 20, 2025); Process 3: Completed normally. In this case, the abnormal welding temperature event in Process 2 is identified from the processing events, classified as "Welding Abnormality," and the specific information of the abnormal event is recorded: Process: Process 2; Abnormality type: Abnormal welding temperature; Occurrence time: June 20, 2025; Relevant data: Welding temperature exceeds the standard range (e.g., standard temperature is 250°C, actual temperature is 280°C).

[0080] Assume the processing steps for model "LED-2023" are as follows: Step 1: PCB board assembly; Step 2: LED chip soldering; Step 3: Functional testing. Then, query the processing steps for model "LED-2023" to determine the specific content of each step. Match the abnormal soldering temperature event in Step 2 with the processing steps to determine that the abnormal event occurred in the LED chip soldering step. Based on the specific information of the abnormal soldering temperature (temperature exceeding the standard range), deduce the internal fault characteristics as "poor solder joint" or "insufficient solder joint strength".

[0081] Therefore, a first fault procedure table is determined based on past processing events and surface fault characteristics of the LED display screen, and a second fault procedure table is determined based on past processing events and internal fault characteristics of the LED display screen. A fault optimization procedure table for the LED display screen is then determined by combining the first and second fault procedure tables. This approach incorporates the overall consideration of combining the first and second fault procedure tables, ensuring the accuracy of the fault optimization procedure table for the LED display screen. Furthermore, by introducing both surface and internal fault characteristics of the LED display screen, a comprehensive consideration of both surface and internal fault characteristics is achieved, further improving the accuracy of the fault optimization procedure table for the LED display screen.

[0082] At this point, the past processing events of the LED display screen are analyzed to identify processing steps that may affect surface fault characteristics. The surface fault characteristics are associated with the processing events to determine the specific processing steps that cause the surface fault. Based on the association results, a first fault procedure table for surface faults is generated. The first fault procedure table is generated based on the past processing events and surface fault characteristics of the LED display screen and is used to record the processing steps related to the surface faults and their corresponding repair procedures.

[0083] Optionally, assuming the LED display model is "LED-2023", previous processing events included module assembly, soldering, and testing. The surface fault characteristic is "module not lighting up". In this case, for the processing event, module assembly: check if the module's power socket is loose; soldering: check if the solder joints are cold solder joints; testing: check if the power supply is normal during the testing process. The surface fault characteristic "module not lighting up" may be related to a loose power socket in the module assembly process. A first fault procedure table is introduced, as shown in Table 1:

[0084] Table 1: First Fault Procedure Table

[0085]

[0086] Analyze past processing events of the LED display screen to identify processing steps that may affect internal fault characteristics. Associate the internal fault characteristics with processing events to determine the specific processing steps that cause the internal fault. Based on the association results, generate a second fault procedure table for the internal fault. The second fault procedure table is generated based on past processing events and internal fault characteristics of the LED display screen and is used to record the processing steps related to the internal fault and their corresponding repair procedures.

[0087] Optionally, assuming the internal fault characteristic of the LED display is "drive circuit failure," for the processing event, the module assembly involves checking whether the drive circuit components of the module are installed correctly; soldering involves checking the soldering quality of the drive circuit components; and testing involves checking whether the circuit functions normally during the testing process. Since the internal fault characteristic "drive circuit failure" may be related to poor soldering of the drive circuit components during the soldering process, a second fault procedure table is introduced, as shown in Table 2.

[0088] Table 2 Second Fault Procedure Table

[0089]

[0090] The first fault procedure table and the second fault procedure table are merged to form a complete fault optimization procedure table. The procedure table is optimized based on the severity of the fault characteristics and the difficulty of repair to ensure repair efficiency. Weights are assigned to each repair procedure, and the priority of the procedures is determined based on these weights. The fault optimization procedure table is generated by integrating the surface and internal fault characteristics of the LED display screen. It is an optimized procedure table used to guide repair personnel to complete repair tasks efficiently and accurately. It integrates the first fault procedure table (for surface faults) and the second fault procedure table (for internal faults), and optimizes the sequencing of procedures based on the severity of the fault, repair priority, and feasibility of actual operation to ensure the efficiency and accuracy of the repair process. Optionally, the fault optimization procedure table can be collected, as shown in Table 3.

[0091] Table 3 Fault Optimization Procedure Table

[0092]

[0093] refer to Figure 5 In step S14, based on the fault optimization process list, the model of the LED display screen and the location of the LED display screen, the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display screen are determined, and the maintenance progress of the two maintenance arms relative to the corresponding LED display screen is collected.

[0094] In the specific implementation of this invention, the specific steps are as follows:

[0095] S141: Collect the fault optimization process table. Each LED display screen is matched with a fault optimization process table. Based on the parsing of the fault optimization process table, determine multiple fault optimization items. Based on the multiple fault optimization items and the corresponding LED display screen positions, determine the maintenance path of the maintenance robot's maintenance arm for each fault optimization item.

[0096] S142: Determine the corresponding fault optimization content based on multiple fault optimization projects and the corresponding LED display model, and determine the corresponding maintenance content based on the mapping relationship between fault optimization content and maintenance type; determine the maintenance action of the maintenance robot's maintenance arm based on the maintenance content;

[0097] S143: Based on the maintenance path and maintenance actions of the maintenance arm for each fault optimization project, determine the intelligent maintenance events of the maintenance robot's maintenance arm for each LED display screen, mark the action sequence of each maintenance action and the retesting procedure of each fault optimization project in the completion stage, and monitor the maintenance process of the two maintenance arms relative to the corresponding LED display screen in real time, and record the maintenance progress of the two maintenance arms relative to the corresponding LED display screen.

[0098] In the embodiments of this application, the fault optimization process table is collected, and each LED display screen is matched with a fault optimization process table. Multiple fault optimization items are determined based on the parsing of the fault optimization process table. The maintenance path of the maintenance robot's maintenance arm for each fault optimization item is determined based on the multiple fault optimization items and the corresponding LED display screen positions. This takes into account the overall consideration of multiple fault optimization items and the corresponding LED display screen positions, ensuring the accuracy of the maintenance path of the maintenance robot's maintenance arm for each fault optimization item.

[0099] At this point, the fault optimization procedure table is retrieved from the maintenance robot's control system. This table typically contains information such as fault characteristics, repair procedures, weights, and scores. The completeness and accuracy of the fault optimization procedure table are verified to ensure that all necessary information is included. The fault optimization procedure table is then stored in the maintenance robot's control system for subsequent processing. Each fault optimization item, including fault characteristics and corresponding repair procedures, is extracted from the fault optimization procedure table. The fault optimization items are then sorted according to their weights or scores, with priority given to items with higher weights or scores. The extracted fault optimization items are then recorded in the maintenance robot's control system.

[0100] Obtain the fault optimization procedure table from the control system of the maintenance robot. Check if the table contains all necessary information, such as fault characteristics, repair procedures, weights, and scores. Store the fault optimization procedure table in the control system. Extract fault optimization items from the fault optimization procedure table: Item 1: Welding point cold weld, repair procedure is welding point re-welding; Item 2: Severe scratch, repair procedure is scratch repair. Sort the items according to their weights: Item 1 (welding point cold weld, weight 0.9); Item 2 (severe scratch, weight 0.8). Record the extracted fault optimization items in the control system.

[0101] Obtain the specific location information of the LED display screen, such as its coordinates on the maintenance platform. Based on the fault optimization project and the location of the display screen, plan the maintenance path of the maintenance arm. Path planning needs to consider the range of motion and movement limitations of the robotic arm, optimize the maintenance path, and ensure that the path is shortest and avoids collisions. Path planning algorithms (such as A* algorithm or Dijkstra's algorithm) can be used for optimization. Optionally, assuming the location of the LED display screen is in the 3rd row and 4th column of maintenance platform 1, with coordinates (50, 30), obtain the location information of the LED display screen, coordinates (50, 30); based on the fault optimization project and the location of the display screen, plan the maintenance path: Start point: maintenance station (coordinates (0, 0)); Path 1: Move to the welding point (coordinates (50, 30)); Path 2: Move to the scratch location (coordinates (80, 60)); End point: Return to the maintenance station (coordinates (0, 0)); Use the A* algorithm to optimize the path, ensuring that the path is shortest and avoids collisions.

[0102] Furthermore, based on multiple fault optimization projects and the corresponding LED display models, the corresponding fault optimization content is determined, and the corresponding maintenance content is determined based on the mapping relationship between the fault optimization content and maintenance types. The maintenance actions of the maintenance robot's maintenance arm are determined based on the maintenance content, which takes into account the overall consideration of the mapping relationship between fault optimization content and maintenance types, and ensures the accuracy of the corresponding maintenance content.

[0103] At this point, based on the LED display model, fault characteristics and repair procedures related to that model are extracted from the fault optimization project. Combining the fault optimization project and model information, specific fault optimization content is determined and recorded in the maintenance robot's control system. A maintenance type mapping table is defined to record the correspondence between fault optimization content and specific maintenance content. Based on the fault optimization content, the corresponding maintenance content is searched from the mapping table and recorded in the maintenance robot's control system. Based on the maintenance content, the corresponding maintenance action is searched from the mapping table. Based on the maintenance action, the specific operation steps of the maintenance arm are planned, including actions such as moving, gripping, welding, and repairing. The maintenance actions are recorded in the maintenance robot's control system.

[0104] Optionally, a preset repair type mapping table can be collected, as shown in Table 4:

[0105] Table 4 Repair Type Mapping Table

[0106]

[0107] Based on the fault optimization content "welding point re-welding" and "scratch repair", find the corresponding repair content from the mapping table: welding point re-welding → welding repair; scratch repair → surface repair; record the repair content in the control system. Based on the repair content "welding repair" and "surface repair", find the corresponding repair action from the mapping table: welding repair → welding action; surface repair → scratch repair action; plan the specific operation steps for repairing the outrigger: move to the welding point position and perform the welding operation; move to the scratch position and perform the scratch repair operation; record the repair actions in the control system.

[0108] Therefore, based on the maintenance path and maintenance actions of the maintenance arm for each fault optimization project, the intelligent maintenance events of the maintenance robot's maintenance arm for each LED display are determined. The action sequence of each maintenance action and the retesting procedures of each fault optimization project in the completion stage are marked. At the same time, the maintenance process of the two maintenance arms relative to the corresponding LED display is monitored in real time, and the maintenance progress of the two maintenance arms relative to the corresponding LED display is recorded. This takes into account the overall consideration of the maintenance path and maintenance actions of the maintenance arm for each fault optimization project, ensuring the accuracy of the intelligent maintenance events of the maintenance robot's maintenance arm for each LED display.

[0109] At this point, based on the maintenance path and maintenance actions, intelligent maintenance events are defined for each fault optimization project. Each maintenance action is assigned a sequential number to ensure the orderly nature of the maintenance process. After each fault optimization project is completed, the retesting procedures that need to be performed are marked. A real-time monitoring system is set up to monitor the movement and operation status of the maintenance arm, record the start and end times of each maintenance action, calculate the maintenance progress, detect possible abnormalities during the maintenance process, such as robotic arm collisions or operation failures, store the maintenance progress records in the control system of the maintenance robot, generate a maintenance progress report for maintenance personnel to refer to, and provide feedback on the maintenance progress to the maintenance personnel so that the maintenance plan can be adjusted in a timely manner.

[0110] Optionally, assume the repair path and repair actions are as follows: Move to the welding point (coordinates (50, 30)) and perform the welding operation; move to the scratch location (coordinates (80, 60)) and perform the scratch repair operation. In this case, define the intelligent repair events: Event 1: Welding Repair (Welding Point Re-welding); Event 2: Scratch Repair (Surface Repair); assign sequence numbers to each repair action: Event 1 (Welding Repair): Action Sequence 1; Event 2 (Scratch Repair): Action Sequence 2; mark the retest process: Event 1 (Welding Repair): Retest process is welding point strength testing; Event 2 (Scratch Repair): Retest process is surface quality inspection; collect the intelligent repair event matching table, as shown in Table 5:

[0111] Table 5 Intelligent Maintenance Event Matching Table

[0112]

[0113] Assume the maintenance process is as follows: Event 1 (Welding Repair) Start Time: June 24, 2025, 10:30 AM; Event 1 (Welding Repair) End Time: June 24, 2025, 10:35 AM; Event 2 (Scratch Repair) Start Time: June 24, 2025, 10:35 AM; Event 2 (Scratch Repair) End Time: June 24, 2025, 10:40 AM; Activate the real-time monitoring system to monitor the movement and operation status of the maintenance arm, and record the start and end times of each maintenance action: Event 1 (Welding Repair): Start Time 10:30 AM, End Time 10:35 AM; Event 2 (Scratch Repair): Start Time 10:35 AM, End Time 10:40 AM; Monitor for any abnormal situations during the maintenance process, such as robotic arm collisions or operational failures.

[0114] The maintenance progress records are stored in the control system of the maintenance robot to generate maintenance progress reports for maintenance personnel to refer to, and the maintenance progress is fed back to the maintenance personnel so that the maintenance plan can be adjusted in a timely manner.

[0115] Optionally, the repair progress records can be stored in the control system to generate a repair progress report: Event 1 (Welding Repair): Completion time 10:35, Repair time 5 minutes; Event 2 (Scratch Repair): Completion time 10:40, Repair time 5 minutes. The repair progress will be fed back to the repair personnel so that the repair plan can be adjusted in a timely manner. The repair progress report is shown in Table 6.

[0116] Table 6 Repair Progress Report

[0117]

[0118] refer to Figure 6 In step S15, optimization measures for the maintenance progress are determined based on the two maintenance progresses, the task sequence list of the maintenance robot, and the current time. These optimization measures include batch maintenance of the final maintenance process of each LED display screen, coordination of external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm.

[0119] In the specific implementation of this invention, the specific steps are as follows:

[0120] S151: Collect the task sequence table of the maintenance robot, determine the current maintenance task and the remaining maintenance task based on the parsing of the task sequence table of the maintenance robot, mark the trigger time of the remaining maintenance task, and determine the remaining time of the current maintenance task based on the trigger time of the remaining maintenance task and the current time.

[0121] S152: Determine the first optimization coefficient based on the remaining time of the current maintenance task and the two maintenance progresses, and determine the second optimization coefficient based on the two maintenance progresses and the task sequence table of the maintenance robot;

[0122] S153: Based on the mapping relationship between the first optimization coefficient, the second optimization coefficient, and the optimization measures, determine the optimization measures for the maintenance progress, and accelerate the maintenance progress based on the implementation of the optimization measures. At this time, the optimization measures cover the batch maintenance of the final maintenance process of each LED display screen, the coordination of external collaborative equipment, or the adjustment of the action efficiency of the maintenance support arm.

[0123] In the embodiments of this application, the task sequence list of the maintenance robot is collected, the current maintenance task and the remaining maintenance tasks are determined by parsing the task sequence list of the maintenance robot, and the trigger time of the remaining maintenance tasks is marked. The remaining time of the current maintenance task is determined based on the trigger time of the remaining maintenance tasks and the current time, which takes into account the overall consideration of the trigger time of the remaining maintenance tasks and the current time, and ensures the accuracy of the remaining time of the current maintenance task.

[0124] At this point, the task sequence list is retrieved from the maintenance robot's control system. This list typically contains detailed information about all maintenance tasks, such as task name, trigger time, and estimated completion time. The completeness and accuracy of the task sequence list are verified, ensuring all necessary information is included. The task sequence list is then stored in the maintenance robot's control system for subsequent processing. The task sequence list is parsed, and based on the current time, the currently executing maintenance task and any remaining tasks that have not yet started are determined. The trigger times of the remaining maintenance tasks are marked for subsequent calculation of the remaining time. Based on the current time and the trigger times of the remaining maintenance tasks, the remaining time for the current maintenance task is calculated.

[0125] Optionally, the maintenance task sequence list is shown in Table 7:

[0126] Table 7 Maintenance Task Sequence Table

[0127]

[0128] Obtain the maintenance task sequence list from the control system of the maintenance robot. Check if the task sequence list contains all necessary information, such as task name, trigger time, and estimated completion time. Store the task sequence list in the control system. Assuming the current time is 2025-06-24 10:32, the current maintenance task is: Welding Repair (trigger time 10:30, estimated completion time 10:35). The remaining maintenance tasks are: Scratch Repair (trigger time 10:35, estimated completion time 10:40); Surface Quality Inspection (trigger time 10:40, estimated completion time 10:45). Simultaneously, the trigger time for Scratch Repair is 10:35; the trigger time for Surface Quality Inspection is 10:40; the remaining time for the current maintenance task (Welding Repair) is 3 minutes (10:35 - 10:32). Current maintenance task: Welding Repair; Remaining time: 3 minutes; Remaining maintenance tasks: Scratch Repair (trigger time 10:35); Surface Quality Inspection (trigger time 10:40).

[0129] Furthermore, a first optimization coefficient is determined based on the remaining time of the current maintenance task and the two maintenance progresses, and a second optimization coefficient is determined based on the two maintenance progresses and the task sequence table of the maintenance robot. This takes into account the overall consideration of the two maintenance progresses and the task sequence table of the maintenance robot, ensuring the accuracy of the second optimization coefficient.

[0130] At this point, based on the trigger time of the current maintenance task and the current time, the remaining time is calculated, and the progress information of the two maintenance tasks is obtained, including the actual completion time and the estimated completion time. Based on the remaining time and the maintenance progress, the first optimization coefficient is calculated. This coefficient reflects the urgency of the current maintenance task.

[0131] Optionally, assume the current time is 2025-06-24 10:32; the current maintenance task is "welding repair", the trigger time is 10:30, and the estimated completion time is 10:35; the progress information of the two maintenance tasks is as follows: Welding repair: actual start time 10:30, estimated completion time 10:35; Scratch repair: actual start time 10:35, estimated completion time 10:40.

[0132] At this moment, the remaining time for the current repair task (welding repair) is 3 minutes (10:35 - 10:32); the actual progress of welding repair is: 2 minutes completed, 3 minutes remaining; the actual progress of scratch repair is: not yet started; the first optimization coefficient = 1 / remaining time = 1 / 3 ≈ 0.33; the first optimization coefficient is 0.33.

[0133] Summarize the actual and projected progress of the two maintenance tasks, analyze the task sequence list, determine the average progress of all tasks, and calculate the second optimization coefficient based on the summarized maintenance progress and task sequence list. This coefficient reflects the efficiency of the overall maintenance progress.

[0134] Optional, collect the task sequence list, the actual progress of welding repair: 2 minutes completed, 3 minutes remaining, the actual progress of scratch repair: not yet started, the actual progress of surface quality inspection: not yet started; the average progress of all tasks: (5 minutes + 5 minutes + 5 minutes) / 3 = 5 minutes; the second optimization coefficient = 1 / average progress = 1 / 5 = 0.2; the second optimization coefficient: 0.2.

[0135] Therefore, based on the mapping relationship between the first optimization coefficient, the second optimization coefficient, and the optimization measures, optimization measures for maintenance progress are determined, and maintenance progress is accelerated based on the execution of optimization measures. At this time, the optimization measures cover the batch maintenance of the final maintenance process of each LED display, the coordination of external collaborative equipment, or the adjustment of the action efficiency of the maintenance support arm. It is compatible with the overall consideration of the mapping relationship between the first optimization coefficient, the second optimization coefficient, and the optimization measures, ensuring the accuracy of the optimization measures for maintenance progress. At the same time, the maintenance progress of two maintenance support arms relative to the corresponding LED display is introduced, which is compatible with the overall consideration of the maintenance progress of two maintenance support arms relative to the corresponding LED display, the task sequence list of the maintenance robot, and the current time, improving the accuracy of the optimization measures for maintenance progress. It realizes that a maintenance robot can perform synchronous and intelligent maintenance on LED display maintenance stations in different locations.

[0136] At this point, an optimization measure mapping table is defined to record the optimization measures corresponding to different optimization coefficient ranges. Based on the first and second optimization coefficients, appropriate optimization measures are selected from the mapping table. Optimization measures may include batch repair, coordination with external collaborative equipment, or adjusting the motion efficiency of the maintenance arm. If the optimization measure includes batch repair, the final repair process of multiple LED displays is scheduled to be carried out simultaneously. If the optimization measure includes coordination with external collaborative equipment, the external equipment is notified to prepare for collaborative operation. If the optimization measure includes adjusting the motion efficiency of the maintenance arm, the motion parameters of the robotic arm are optimized to improve efficiency. The execution effect of the optimization measures is monitored in real time to ensure that the repair progress is accelerated. Meanwhile, the first optimization coefficient is an indicator calculated based on the remaining time and repair progress of the current repair task, reflecting the urgency of the current repair task. It helps the maintenance robot assess the priority of the current task so that optimization measures can be taken when necessary. The second optimization coefficient is an indicator calculated based on the average progress of all repair tasks and the task sequence table, reflecting the efficiency of the overall repair progress. It helps the maintenance robot assess the execution of the overall repair task so that optimization measures can be taken when necessary.

[0137] Optionally, an optimization measure mapping table can be collected, as shown in Table 8:

[0138] Table 8: Mapping Table of Optimization Measures

[0139]

[0140] At this point, an optimization measure mapping table is created to record the optimization measures corresponding to different optimization coefficient ranges. Based on the first optimization coefficient (0.33) and the second optimization coefficient (0.2), a suitable optimization measure is selected from the mapping table. Based on the selected optimization measure, the specific execution content is described. First optimization coefficient: 0.33; Second optimization coefficient: 0.2; Selected optimization measure: Collaboration of external cooperative equipment.

[0141] Please see Figure 7 , Figure 7 This is a schematic diagram of the structural composition of the intelligent maintenance system for LED displays using a maintenance robot, as described in this embodiment of the invention. The intelligent maintenance system for LED displays using a maintenance robot includes:

[0142] The tabletop image module 21 is used to collect two tabletop images based on the circular shooting of the two LED display maintenance tables around the maintenance robot.

[0143] Surface fault module 22 is used to determine the surface fault characteristics and model of the corresponding LED display screen based on image detection of the two tabletop images;

[0144] The fault optimization process table module 23 is used to determine the internal fault characteristics of the LED display screen based on the model of the LED display screen; and to determine the fault optimization process table of the LED display screen based on the surface fault characteristics and internal fault characteristics of the LED display screen.

[0145] The intelligent maintenance module 24 is used to determine the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display based on the fault optimization process list, the model of the LED display, and the location of the LED display, and to collect the maintenance progress of the two maintenance arms relative to the corresponding LED display.

[0146] The optimization measures module 25 is used to determine optimization measures for the maintenance progress based on the two maintenance progresses, the task sequence list of the maintenance robot, and the current time. These optimization measures include batch maintenance of the final maintenance process of each LED display, coordination of external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm.

[0147] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A method for intelligent maintenance of an LED display screen by a maintenance robot, characterized in that, include: The maintenance robot is surrounded by two LED display maintenance stations. The robot's camera captures images of the two stations in a circular motion. In two tabletop images, the surface fault characteristics and model of the corresponding LED display screen are determined based on image detection of the tabletop images; The internal fault characteristics of the LED display screen are determined by tracing its model number. Based on the surface and internal fault characteristics of the LED display screen, a fault optimization process table is determined, including: collecting the LED display screen model; determining the corresponding traceability path based on the LED display screen model and processing path; identifying past processing events of the LED display screen based on the traceability path, the LED display screen processing database, and the LED display screen model; identifying corresponding processing abnormal events based on the identification of past processing events, including welding abnormalities, temperature abnormalities, and mechanical damage; recording specific information for each processing abnormal event, including occurrence time, process, abnormality type, and related data; and determining the LED display screen's fault optimization process based on the process corresponding to the processing abnormal event, the LED display screen's processing process list, and the corresponding abnormal data. The internal fault characteristics are as follows: internal fault characteristics include poor soldering or insufficient soldering strength; a first fault procedure table is determined based on past processing events and surface fault characteristics of the LED display screen; a second fault procedure table is determined based on past processing events and internal fault characteristics of the LED display screen; and a fault optimization procedure table is determined by combining the first and second fault procedure tables. The first fault procedure table records the processing steps related to surface faults and their corresponding repair steps; the second fault procedure table records the processing steps related to internal faults and their corresponding repair steps; the fault optimization procedure table is an optimized procedure table generated by combining the surface and internal fault characteristics of the LED display screen, used to guide maintenance personnel to complete maintenance tasks efficiently and accurately. Based on the fault optimization process list, the model of the LED display screen and the location of the LED display screen, the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display screen are determined, and the maintenance progress of the two maintenance arms relative to the corresponding LED display screen is collected. Based on the two maintenance schedules, the task sequence list of the maintenance robot, and the current time, optimization measures for the maintenance schedule are determined. These optimization measures include batch maintenance of the final maintenance process for each LED display, coordination with external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm. 2.The method of claim 1, wherein, The maintenance robot is surrounded by two LED display maintenance stations. Images of the two stations are captured by the maintenance robot's camera through a circular camera view, including: The location of the LED display maintenance station is collected, and another LED display maintenance station is determined based on the surrounding detection of the LED display maintenance station. At this time, the two LED display maintenance stations are arranged side by side, and a maintenance station is set between the two LED display maintenance stations. The maintenance robot is located at the maintenance station. A corresponding inspection space area is constructed based on a maintenance robot and two LED display maintenance stations; the maintenance robot is equipped with a camera, and the circular movement path of the camera is determined based on the spatial position of the camera, the spatial position of the two LED display maintenance stations, and the range of motion of the maintenance robot. Driven by the maintenance robot, the camera moves along a circular path and dynamically captures images of the two LED display maintenance platforms. 3.The method of claim 1, wherein, The step of determining the surface fault characteristics and model of the corresponding LED display screen based on image detection in two tabletop images includes: The two platform images are monitored in real time, and image processing is performed on the two platform images. At this time, multiple image areas are divided under the image processing of each platform image. The surface image and model identification image of the LED display screen are determined based on the filtering of multiple image areas. If the surface image and model identification image of the LED display screen are located in the same area of ​​the table surface image, then multi-channel recognition is performed based on the surface image and model identification image of the LED display screen; in the multi-channel recognition of the surface image and model identification image of the LED display screen, the surface shape and model content of the LED display screen are output, and the abnormal surface area is determined based on the detection of the surface shape of the LED display screen. Multiple sub-surface fault features are determined based on the identification of abnormal surface areas, and the surface fault features of the LED display screen are determined based on the location, shape, and surface fault feature mapping relationship of the multiple sub-surface fault features. 4.The method of claim 1, wherein, The process involves determining the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display based on the fault optimization process schedule, the model of the LED display, and the location of the LED display, and collecting the maintenance progress of the two maintenance arms relative to the corresponding LED display, including: The fault optimization process table is collected, and each LED display screen is matched with a fault optimization process table. Multiple fault optimization items are determined based on the parsing of the fault optimization process table. Based on the multiple fault optimization items and the corresponding LED display screen positions, the maintenance path of the maintenance robot's maintenance arm for each fault optimization item is determined. Based on multiple fault optimization projects and the corresponding LED display models, the corresponding fault optimization content is determined, and the corresponding maintenance content is determined based on the mapping relationship between the fault optimization content and maintenance types; the maintenance actions of the maintenance robot's maintenance arm are determined based on the maintenance content. 5.The method of claim 4, wherein, The process of determining the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display based on the fault optimization process list, the model of the LED display, and the location of the LED display, and collecting the maintenance progress of the two maintenance arms relative to the corresponding LED display, also includes: Based on the maintenance path and maintenance actions of the maintenance arms for each fault optimization project, the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display screen are determined, and the action sequence of each maintenance action and the retesting procedures of each fault optimization project in the completion stage are marked. At the same time, the maintenance process of the two maintenance arms relative to the corresponding LED display screen is monitored in real time, and the maintenance progress of the two maintenance arms relative to the corresponding LED display screen is recorded. 6.The method of claim 1, wherein, The optimization measures for determining the maintenance progress based on two maintenance progresses, the task sequence list of the maintenance robot, and the current time include: batch maintenance of the final maintenance process for each LED display screen, coordination with external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm. Collect the task sequence list of the maintenance robot, determine the current maintenance task and the remaining maintenance tasks based on the parsing of the task sequence list, mark the trigger time of the remaining maintenance tasks, and determine the remaining time of the current maintenance task based on the trigger time of the remaining maintenance tasks and the current time. 7.The method of claim 6, wherein, The optimization measures for determining the maintenance progress based on two maintenance progresses, the task sequence list of the maintenance robot, and the current time include: batch maintenance of the final maintenance process for each LED display screen, coordination with external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm; and also include: The first optimization coefficient is determined based on the remaining time of the current maintenance task and the two maintenance progresses, and the second optimization coefficient is determined based on the two maintenance progresses and the task sequence list of the maintenance robot. The optimization measures for maintenance progress are determined based on the mapping relationship between the first optimization coefficient, the second optimization coefficient, and the optimization measures. The maintenance progress is accelerated based on the implementation of the optimization measures. At this time, the optimization measures cover the batch maintenance of the final maintenance process of each LED display, the coordination of external collaborative equipment, or the adjustment of the operation efficiency of the maintenance support arm.

8. An intelligent maintenance system of a maintenance robot for an LED display screen, characterized by, The intelligent maintenance system for LED displays using the maintenance robot is applied to the intelligent maintenance method for LED displays using a maintenance robot as described in any one of claims 1-7, wherein the intelligent maintenance system for LED displays using the maintenance robot includes: The tabletop image module is used to capture images of the two LED display maintenance tables around the maintenance robot by taking a circular shot of the two LED display maintenance tables using the maintenance robot's camera. The surface fault module is used to determine the surface fault characteristics and model of the corresponding LED display screen based on image detection of the two table surface images; The fault optimization process table module is used to determine the internal fault characteristics of the LED display screen based on the model of the LED display screen; and to determine the fault optimization process table of the LED display screen based on the surface fault characteristics and internal fault characteristics of the LED display screen. The intelligent maintenance module is used to determine the intelligent maintenance events of the maintenance robot's maintenance arms for each LED display based on the fault optimization process list, the model of the LED display, and the location of the LED display, and to collect the maintenance progress of the two maintenance arms relative to the corresponding LED display. The optimization measures module is used to determine optimization measures for the maintenance progress based on the two maintenance progresses, the task sequence list of the maintenance robot, and the current time. These optimization measures include batch maintenance of the final maintenance process of each LED display, coordination of external collaborative equipment, or adjustment of the movement efficiency of the maintenance arm.

Images