Robot management system and method for inspection of condenser of power plant

Abstract

The invention relates to the technical field of condenser inspection, and particularly discloses a robot management system and method for power plant condenser inspection, and the system comprises an image segmentation module which is used for collecting current condenser image data through an inspection robot, and segmenting an abnormal image of a condenser according to the condenser image data; the cleaning strategy module is used for calculating gray values and LBP values of pixel points in the condenser abnormal image and selecting a cleaning strategy to clean the condenser according to the gray values and the LBP values; and the defect early warning module is used for acquiring the cleaned abnormal image of the condenser, determining the maintenance grade of the condenser according to the cleaned abnormal image of the condenser, and performing defect early warning on the condenser according to the maintenance grade. The recognition accuracy of the inspection picture is enhanced, the cleaning condition and defects of the condenser can be found in time, and the inspection effect of the inspection robot is improved.

Description

Technical Field

[0001] The present application relates to the technical field of condenser inspection, and more specifically, to a robot management system and method for power plant condenser inspection. Background Art

[0002] With the rapid development of technology, the application of intelligent inspection robots in various industrial fields has gradually become a significant trend. Equipped with various sensors and cameras, these robots can autonomously inspect and monitor power plant equipment, including boilers, generators, and condensers, and detect and analyze the cleanliness and defects of the inspection area in real time.

[0003] However, due to the complex structure of the condenser, the existing inspection robot for the condenser cannot accurately recognize the condenser image, and the accuracy of judging the condenser cleaning condition and defect information during inspection is low, which affects the overall operating status of the condenser. Summary of the Invention

[0004] The present invention provides a robot management system and method for power plant condenser inspection, which is used to solve the problem of low accuracy of inspection robots in judging the cleanliness status and defect information of condensers in the prior art, including: An image segmentation module is used to collect current condenser image data through an inspection robot and segment abnormal images of the condenser based on the condenser image data; The cleaning strategy module is used to calculate the grayscale value and LBP value of the pixel points in the abnormal condenser image, and select a cleaning strategy to clean the condenser according to the grayscale value and LBP value; The defect warning module is used to obtain an abnormal image of the cleaned condenser, determine the maintenance level of the condenser according to the abnormal image of the cleaned condenser, and issue a defect warning to the condenser according to the maintenance level.

[0005] Furthermore, the image segmentation module segments the abnormal image of the condenser according to the condenser image data, including: grayscale processing is performed on the condenser image data to obtain condenser grayscale image data; Performing edge detection on the condenser grayscale image data, and extracting a number of closed edge lines in the condenser grayscale image data according to the edge detection result; The center point of the closed edge line is set as the seed point, the seed points of the condenser grayscale image data are traversed, and the positions of the pixels in the eight-connected region of the seed point are recorded; Traverse the pixel points to be tested in the eight-connected region, and determine whether the similarity between the pixel point to be tested and the seed point is within the preset range according to the similarity calculation formula; If the similarity between the pixel to be tested and the seed point is within the preset range, the pixel to be tested is set as a new growth seed point; With the new growth seed point as the center, continue to detect new pixels to be tested until the area can no longer grow, completing the abnormal image segmentation of the condenser.

[0006] Furthermore, the similarity calculation formula is specifically as follows: , in, is the similarity between the i-th seed point and the j-th pixel to be tested, is the preset grayscale weight, is the gray value of the i-th seed point, is the gray value of the jth pixel to be tested, For preset texture weights, is the LBP value of the i-th seed point, is the LBP value of the j-th pixel to be tested.

[0007] Furthermore, the cleaning strategy module selects a cleaning strategy to clean the condenser according to the grayscale value and the LBP value, including: Obtain the average grayscale value and average LBP value of each abnormal image, and cluster the abnormal images according to the average grayscale value and average LBP value of each abnormal image; Determine the cluster partition to which each abnormal image belongs according to the clustering result of the abnormal image, and set the corresponding weight value according to the cluster partition to which the abnormal image belongs; The area ratio of each abnormal image in the condenser grayscale image data is weightedly summed with the corresponding weight value to obtain the required cleanliness of the condenser, and whether cleaning is required is determined based on the required cleanliness; If it is determined that cleaning is required, a corresponding cleaning strategy is determined based on the abnormal image, and the condenser is cleaned according to the cleaning strategy.

[0008] Furthermore, clustering the abnormal images according to the average grayscale value and the average LBP value of each abnormal image includes: A sample data set is established based on the average grayscale value and average LBP value of each abnormal image, and k initial cluster centers of the sample data set are randomly selected; Calculate the Euclidean distance between the sample data in the sample data set and the initial cluster center, and divide each abnormal image into the corresponding cluster partition according to the Euclidean distance between the sample data in the sample data set and the initial cluster center; Calculate the mean of the sample data in each cluster partition, and recalculate the cluster center based on the mean of the sample data in each cluster partition; Repeat the above steps until the cluster center no longer changes or the number of iterations reaches the preset maximum number of iterations, and obtain the clustering result of the abnormal image.

[0009] Furthermore, the step of setting corresponding weight values according to the cluster partition to which the abnormal image belongs includes: Obtain a central abnormal image closest to the cluster center in the cluster partition, obtain a preset sliding window, and segment the central abnormal image according to the preset sliding window to obtain several abnormal segmentation images; Extracting the feature vector of each abnormal segmented image, and calculating the cosine similarity between the abnormal segmented image and the preset normal segmented image based on the feature vector of the abnormal segmented image; Establishing a similarity grid of the abnormal image based on the cosine similarity between the abnormal segmented image and the preset normal segmented image, and counting the number of grids in the similarity grid whose cosine similarity is less than a first preset threshold; The ratio of the number of grids whose cosine similarity is less than a first preset threshold to a preset normal number in the similarity grid is calculated to obtain the weight value of the corresponding cluster partition.

[0010] Furthermore, determining a corresponding cleaning strategy based on the abnormal image includes: Obtain historical abnormal images of the condenser and corresponding cleaning strategies, and establish a training sample set based on the historical abnormal images and corresponding cleaning strategies; Establish an initial strategy generation model based on the training sample set and train the initial strategy generation model to obtain a trained strategy generation model; The abnormal image of the current condenser is input into the trained strategy generation model to obtain the corresponding condenser cleaning strategy.

[0011] Furthermore, the defect warning module determines the maintenance level of the condenser according to the abnormal image of the cleaned condenser, including: Obtaining a required cleanliness level of the abnormal image after condenser cleaning, and determining whether the required cleanliness level is less than a second preset threshold; If the required cleanliness level is less than the second preset threshold, setting the preset first maintenance level as the maintenance level of the condenser; If the required cleanliness is greater than or equal to the second preset threshold, determining whether the required cleanliness is less than a third preset threshold; If the required cleanliness level is less than the third preset threshold, setting the preset second maintenance level as the maintenance level of the condenser; If the required cleanliness level is greater than or equal to the third preset threshold, the preset third maintenance level is set as the maintenance level of the condenser.

[0012] Furthermore, the defect warning module performs defect warning on the condenser according to the maintenance level, including: When the maintenance level of the condenser is the third maintenance level, a defect warning is issued for the condenser.

[0013] To achieve the above objectives, the present invention further provides a robot management method for power plant condenser inspection, comprising: An image segmentation module is used to collect current condenser image data through an inspection robot and segment abnormal images of the condenser based on the condenser image data; The cleaning strategy module is used to calculate the grayscale value and LBP value of the pixel points in the abnormal condenser image, and select a cleaning strategy to clean the condenser according to the grayscale value and LBP value; The defect warning module is used to obtain an abnormal image of the cleaned condenser, determine the maintenance level of the condenser according to the abnormal image of the cleaned condenser, and issue a defect warning to the condenser according to the maintenance level.

[0014] The beneficial effects of the present invention are: By applying the above technical solution, the present invention can accurately segment the abnormal images that need to be cleaned by segmenting the condenser images collected by the inspection robot, and select the cleaning strategy in a targeted manner according to the grayscale value and LBP value of the abnormal image, thereby greatly improving the cleaning effect. The defect condition of the condenser is judged by the inspection image after cleaning to ensure the normal operation of the condenser. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 A schematic structural diagram of a robot management system for power plant condenser inspection proposed in an embodiment of the present invention is shown; Figure 2 The figure shows an overall flow chart of a robot management method for power plant condenser inspection proposed in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0018] The embodiment of the present application provides a robot management system for power plant condenser inspection, such as Figure 1 As shown, including: The image segmentation module is used to collect the current condenser image data through the inspection robot and segment the abnormal image of the condenser based on the condenser image data; the cleaning strategy module is used to calculate the grayscale value and LBP value of the pixel points in the abnormal condenser image, and select the cleaning strategy to clean the condenser based on the grayscale value and LBP value; the defect warning module is used to obtain the abnormal image of the condenser after cleaning, determine the maintenance level of the condenser based on the abnormal image of the condenser after cleaning, and issue a defect warning for the condenser according to the maintenance level.

[0019] In this embodiment, the power plant condenser is inspected by an inspection robot and inspection images are collected. The abnormal images in the inspection images are segmented out by the image segmentation module, so that the cleaning strategy module determines the cleaning condition of the condenser according to the grayscale value and LBP value of the abnormal image, and selects the corresponding cleaning strategy according to the cleaning condition. After the cleaning is completed, the inspection robot collects a secondary inspection image, and the maintenance level of the condenser is determined according to the abnormal image at the corresponding position of the primary inspection image in the secondary inspection image. Therefore, it is judged whether the condenser has defects based on the maintenance level and a defect warning is issued.

[0020] In some embodiments of the present application, the image segmentation module segments an abnormal image of the condenser based on the condenser image data, including: grayscale processing the condenser image data to obtain condenser grayscale image data; edge detection on the condenser grayscale image data, and extracting several closed edge lines in the condenser grayscale image data based on the edge detection results; setting the center point of the closed edge line as a seed point, traversing the seed points of the condenser grayscale image data, and recording the positions of the pixels in the eight-connected area of the seed point; traversing the pixel points to be tested in the eight-connected area, and judging whether the similarity between the pixel points to be tested and the seed points is within a preset interval according to a similarity calculation formula; if the similarity between the pixel points to be tested and the seed points is within the preset interval, setting the pixel points to be tested as a new growth seed point; with the new growth seed point as the center, continue to detect new pixel points to be tested until the area can no longer grow, thereby completing the segmentation of the abnormal image of the condenser.

[0021] In some embodiments of the present application, the similarity calculation formula is specifically: , in, is the similarity between the i-th seed point and the j-th pixel to be tested, is the preset grayscale weight, is the gray value of the i-th seed point, is the gray value of the jth pixel to be tested, For preset texture weights, is the LBP value of the i-th seed point, is the LBP value of the j-th pixel to be tested.

[0022] In this embodiment, the closed edge lines inside the condenser grayscale image data are detected based on the canny edge detection algorithm, and the center point of the closed area formed by the closed edge line is set as the seed point. The similarity calculation formula is established through the grayscale value and the LBP value to perform regional growth on the seed point. The abnormal image is segmented by combining the grayscale characteristics and texture characteristics of the condenser grayscale image data. The grayscale values of the neighboring pixels and the seed points are compared through the LBP operator to obtain a binary mode and then converted into decimal to obtain the LBP value of the corresponding seed point.

[0023] In some embodiments of the present application, the cleaning strategy module selects a cleaning strategy to clean the condenser based on the grayscale value and the LBP value, including: obtaining the average grayscale value and the average LBP value of each abnormal image, and clustering each abnormal image according to the average grayscale value and the average LBP value of each abnormal image; determining the cluster partition to which each abnormal image belongs based on the clustering result of the abnormal image, and setting a corresponding weight value based on the cluster partition to which the abnormal image belongs; performing a weighted summation of the area ratio of each abnormal image in the condenser grayscale image data and the corresponding weight value to obtain the required cleanliness of the condenser, and judging whether cleaning is needed based on the required cleanliness; if it is judged that cleaning is needed, determining the corresponding cleaning strategy based on the abnormal image, and cleaning the condenser according to the cleaning strategy.

[0024] In this embodiment, abnormal images are clustered to achieve classification of each abnormal image. The normality of the abnormal image is determined by the classification result of each abnormal image, and the corresponding weight value is set. The required cleanliness is calculated by taking the weighted sum of the area ratio of the abnormal image to the condenser grayscale image data and the corresponding weight value, so as to achieve an accurate assessment of the condenser cleaning condition. When it is determined that cleaning is needed, the corresponding cleaning strategy is selected according to the image features of all abnormal images to clean the condenser.

[0025] In some embodiments of the present application, clustering each abnormal image according to the average grayscale value and the average LBP value of each abnormal image includes: establishing a sample data set according to the average grayscale value and the average LBP value of each abnormal image, and randomly selecting k initial cluster centers of the sample data set; calculating the Euclidean distance from the sample data in the sample data set to the initial cluster center, and dividing each abnormal image into a corresponding cluster partition according to the Euclidean distance from the sample data in the sample data set to the initial cluster center; calculating the mean of the sample data in each cluster partition, and recalculating the cluster center according to the mean of the sample data in each cluster partition; repeating the above steps until the cluster center no longer changes or the number of iterations reaches a preset maximum number of iterations, thereby obtaining a clustering result of the abnormal image.

[0026] In this embodiment, each abnormal image is clustered based on the average grayscale value and average LBP value of the abnormal image based on the k-means clustering algorithm to achieve classification of the abnormal image. The k value of this embodiment is selected according to the number of clustered images. The larger the number, the higher the corresponding k value.

[0027] In some embodiments of the present application, the setting of corresponding weight values according to the cluster partition to which the abnormal image belongs includes: obtaining a central abnormal image closest to the cluster center in the cluster partition, obtaining a preset sliding window, and segmenting the central abnormal image according to the preset sliding window to obtain a plurality of abnormal segmented images; extracting feature vectors of each abnormal segmented image, and calculating the cosine similarity between the abnormal segmented image and the preset normal segmented image according to the feature vector of the abnormal segmented image; establishing a similarity grid of the abnormal image according to the cosine similarity between the abnormal segmented image and the preset normal segmented image, and counting the number of grids in the similarity grid whose cosine similarity is less than a first preset threshold; calculating the ratio of the number of grids in the similarity grid whose cosine similarity is less than the first preset threshold to the preset normal number, and obtaining the weight value of the corresponding cluster partition.

[0028] In this embodiment, the feature vector of the abnormal segmented image is extracted by combining structural features and color features, and then a similarity grid is established through the cosine similarity between the abnormal segmented image and the preset normal segmented image. The ratio of the number of grids whose cosine similarity is less than a first preset threshold to the preset normal number is set as the weight value of the corresponding abnormal image. The more grids there are, the higher the weight value of the required cleanliness of the abnormal image.

[0029] In some embodiments of the present application, determining the corresponding cleaning strategy based on the abnormal image includes: obtaining historical abnormal images of the condenser and the corresponding cleaning strategies, and establishing a training sample set based on the historical abnormal images and the corresponding cleaning strategies; establishing an initial strategy generation model based on the training sample set and training the initial strategy generation model to obtain a trained strategy generation model; inputting the abnormal image of the current condenser into the trained strategy generation model to obtain the corresponding condenser cleaning strategy.

[0030] In this embodiment, a training sample set is established by using abnormal images of the condenser during historical cleaning and the corresponding cleaning strategies to train an initial strategy generation neural network model, and the corresponding condenser cleaning strategy is obtained by inputting the abnormal image of the current condenser into the trained strategy generation model.

[0031] In some embodiments of the present application, the defect warning module determines the maintenance level of the condenser based on the abnormal image of the condenser after cleaning, including: obtaining the required cleanliness of the abnormal image of the condenser after cleaning, and judging whether the required cleanliness is less than a second preset threshold; if the required cleanliness is less than the second preset threshold, setting the preset first maintenance level as the maintenance level of the condenser; if the required cleanliness is greater than or equal to the second preset threshold, judging whether the required cleanliness is less than a third preset threshold; if the required cleanliness is less than the third preset threshold, setting the preset second maintenance level as the maintenance level of the condenser; if the required cleanliness is greater than or equal to the third preset threshold, setting the preset third maintenance level as the maintenance level of the condenser.

[0032] In some embodiments of the present application, the defect warning module performs defect warning on the condenser according to the maintenance level, including: when the maintenance level of the condenser is the third maintenance level, performing defect warning on the condenser.

[0033] In this embodiment, the maintenance level of the condenser is determined by the required cleanliness of the secondary inspection image after cleaning. When the maintenance level is too low, it is determined that the condenser has defects and a defect warning is issued.

[0034] Based on the same technical concept, such as Figure 2 As shown, the present invention also provides a robot management method for power plant condenser inspection, comprising: S101, an image segmentation module, configured to collect current condenser image data through an inspection robot, and segment abnormal images of the condenser based on the condenser image data; S102, a cleaning strategy module, for calculating the grayscale value and LBP value of the pixel points in the abnormal condenser image, and selecting a cleaning strategy to clean the condenser according to the grayscale value and LBP value; S103, a defect warning module is used to obtain an abnormal image of the cleaned condenser, determine the maintenance level of the condenser according to the abnormal image of the cleaned condenser, and issue a defect warning to the condenser according to the maintenance level.

[0035] By applying the above technical solution, the present invention includes an image segmentation module for collecting current condenser image data via an inspection robot and segmenting abnormal condenser images based on the condenser image data; a cleaning strategy module for calculating the grayscale value and LBP value of pixels in the abnormal condenser image and selecting a cleaning strategy for cleaning the condenser based on the grayscale value and LBP value; and a defect warning module for obtaining abnormal condenser images after cleaning, determining the maintenance level of the condenser based on the cleaned abnormal condenser images, and issuing defect warnings for the condenser based on the maintenance level. This invention enhances the recognition accuracy of inspection images, enables timely detection of condenser cleaning conditions and defects, and improves the inspection effectiveness of the inspection robot.

[0036] Through the above description of the embodiments, those skilled in the art will clearly understand that the present invention can be implemented via hardware or via software combined with a necessary general-purpose hardware platform. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product. This software product can be stored on a non-volatile storage medium (such as a CD-ROM, USB flash drive, or external hard drive) and includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute the methods described in various implementation scenarios of the present invention.

[0037] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A robot management system for inspection of condensers in power plants, characterized in that: include: An image segmentation module is used to collect current condenser image data through an inspection robot and segment abnormal images of the condenser based on the condenser image data; The cleaning strategy module is used to calculate the grayscale value and LBP value of the pixel points in the abnormal condenser image, and select a cleaning strategy to clean the condenser according to the grayscale value and LBP value; The defect warning module is used to obtain an abnormal image of the cleaned condenser, determine the maintenance level of the condenser according to the abnormal image of the cleaned condenser, and issue a defect warning to the condenser according to the maintenance level.

2. The robot management system for power plant condenser inspection according to claim 1, characterized in that: The image segmentation module segments an abnormal image of the condenser according to the condenser image data, including: grayscale processing is performed on the condenser image data to obtain condenser grayscale image data; Performing edge detection on the condenser grayscale image data, and extracting a number of closed edge lines in the condenser grayscale image data according to the edge detection result; The center point of the closed edge line is set as the seed point, the seed points of the condenser grayscale image data are traversed, and the positions of the pixels in the eight-connected region of the seed point are recorded; Traverse the pixel points to be tested in the eight-connected region, and determine whether the similarity between the pixel point to be tested and the seed point is within the preset range according to the similarity calculation formula; If the similarity between the pixel to be tested and the seed point is within the preset range, the pixel to be tested is set as a new growth seed point; With the new growth seed point as the center, continue to detect new pixels to be tested until the area can no longer grow, completing the abnormal image segmentation of the condenser.

3. The robot management system for power plant condenser inspection according to claim 2, characterized in that: The similarity calculation formula is specifically: ， in, is the similarity between the i-th seed point and the j-th pixel to be tested, is the preset grayscale weight, is the gray value of the i-th seed point, is the gray value of the jth pixel to be tested, For preset texture weights, is the LBP value of the i-th seed point, is the LBP value of the j-th pixel to be tested.

4. The robot management system for power plant condenser inspection according to claim 1, characterized in that: The cleaning strategy module selects a cleaning strategy to clean the condenser according to the grayscale value and the LBP value, including: Obtain the average grayscale value and average LBP value of each abnormal image, and cluster the abnormal images according to the average grayscale value and average LBP value of each abnormal image; Determine the cluster partition to which each abnormal image belongs according to the clustering result of the abnormal image, and set the corresponding weight value according to the cluster partition to which the abnormal image belongs; The area ratio of each abnormal image in the condenser grayscale image data is weightedly summed with the corresponding weight value to obtain the required cleanliness of the condenser, and whether cleaning is required is determined based on the required cleanliness; If it is determined that cleaning is required, a corresponding cleaning strategy is determined based on the abnormal image, and the condenser is cleaned according to the cleaning strategy.

5. The robot management system for power plant condenser inspection according to claim 4, characterized in that: Clustering each abnormal image according to the average grayscale value and the average LBP value of each abnormal image includes: A sample data set is established based on the average grayscale value and average LBP value of each abnormal image, and k initial cluster centers of the sample data set are randomly selected; Calculate the Euclidean distance between the sample data in the sample data set and the initial cluster center, and divide each abnormal image into the corresponding cluster partition according to the Euclidean distance between the sample data in the sample data set and the initial cluster center; Calculate the mean of the sample data in each cluster partition, and recalculate the cluster center based on the mean of the sample data in each cluster partition; Repeat the above steps until the cluster center no longer changes or the number of iterations reaches the preset maximum number of iterations, and obtain the clustering result of the abnormal image.

6. The robot management system for power plant condenser inspection according to claim 4, characterized in that: The step of setting a corresponding weight value according to the cluster partition to which the abnormal image belongs includes: Obtain a central abnormal image closest to the cluster center in the cluster partition, obtain a preset sliding window, and segment the central abnormal image according to the preset sliding window to obtain several abnormal segmentation images; Extracting the feature vector of each abnormal segmented image, and calculating the cosine similarity between the abnormal segmented image and the preset normal segmented image based on the feature vector of the abnormal segmented image; Establishing a similarity grid of the abnormal image based on the cosine similarity between the abnormal segmented image and the preset normal segmented image, and counting the number of grids in the similarity grid whose cosine similarity is less than a first preset threshold; The ratio of the number of grids whose cosine similarity is less than a first preset threshold to a preset normal number in the similarity grid is calculated to obtain the weight value of the corresponding cluster partition.

7. The robot management system for power plant condenser inspection according to claim 4, characterized in that: Determining a corresponding cleaning strategy based on the abnormal image includes: Obtain historical abnormal images of the condenser and corresponding cleaning strategies, and establish a training sample set based on the historical abnormal images and corresponding cleaning strategies; Establish an initial strategy generation model based on the training sample set and train the initial strategy generation model to obtain a trained strategy generation model; The abnormal image of the current condenser is input into the trained strategy generation model to obtain the corresponding condenser cleaning strategy.

8. The robot management system for power plant condenser inspection according to claim 7, characterized in that: The defect warning module determines the maintenance level of the condenser according to the abnormal image of the cleaned condenser, including: Obtaining the required cleanliness level of the abnormal image after condenser cleaning, and determining whether the required cleanliness level is less than a second preset threshold; If the required cleanliness level is less than the second preset threshold, setting the preset first maintenance level as the maintenance level of the condenser; If the required cleanliness is greater than or equal to the second preset threshold, determining whether the required cleanliness is less than a third preset threshold; If the required cleanliness level is less than the third preset threshold, setting the preset second maintenance level as the maintenance level of the condenser; If the required cleanliness level is greater than or equal to the third preset threshold, the preset third maintenance level is set as the maintenance level of the condenser.

9. The robot management system for power plant condenser inspection according to claim 8, characterized in that: The defect warning module performs defect warning on the condenser according to the maintenance level, including: When the maintenance level of the condenser is the third maintenance level, a defect warning is issued for the condenser.

10. A robot management method for power plant condenser inspection, characterized in that: include: An image segmentation module is used to collect current condenser image data through an inspection robot and segment abnormal images of the condenser based on the condenser image data; The cleaning strategy module is used to calculate the grayscale value and LBP value of the pixel points in the abnormal condenser image, and select a cleaning strategy to clean the condenser according to the grayscale value and LBP value; The defect warning module is used to obtain an abnormal image of the cleaned condenser, determine the maintenance level of the condenser according to the abnormal image of the cleaned condenser, and issue a defect warning to the condenser according to the maintenance level.

Images