Method and device for detecting high-temperature corrosion of water cooling wall of coal-fired boiler, equipment and medium

By using the YOLOv8 model of characteristic pyramid network structure and attention mechanism in thermal generator sets to perform high-temperature corrosion detection of water-cooled walls, the problems of time-consuming and labor-consuming and detection blind spots in the existing technology are solved, and automation is achieved. Real-time inspection improves detection efficiency and equipment maintenance accuracy and safety.

CN118941529APending Publication Date: 2024-11-12XIAN THERMAL POWER RES INST CO LTD

Patent Information

Application Number
CN202410993756.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

In existing thermal power generator sets, water-cooled walls are prone to high-temperature corrosion in high-temperature and high-pressure environments, resulting in pipe leakage or bursting. The existing manual detection methods are time-consuming and labor-intensive and difficult to cover all areas inside the furnace, and there are detection blind spots.

Method used

The high-temperature corrosion defect detection model is constructed using the YOLOv8 model based on the characteristic pyramid network structure and attention mechanism. By obtaining the surface image of the water-cooled wall pipe for automatic detection, the types of high-temperature corrosion defects and their severity are identified, and the detection report is generated.

Benefits of technology

It realizes automated and real-time detection of high-temperature corrosion defects in water-cooled walls, reduces time and labor costs, improves detection efficiency, promptly detects potential equipment safety hazards, and reduces operational risks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941529A_ABST
    Figure CN118941529A_ABST
Patent Text Reader

Abstract

The embodiment of the invention relates to the technical field of thermal power generation equipment maintenance, and provides a coal-fired boiler water-cooled wall high-temperature corrosion detection method and device, equipment and a medium, and the method comprises the following steps: obtaining a water-cooled wall pipe surface image of a coal-fired boiler; inputting the surface image of the water wall pipe into a trained high-temperature corrosion defect detection model to obtain a corresponding detection result; wherein the high-temperature corrosion defect detection model is constructed based on a YOLOv8 model introducing a feature pyramid network structure and an attention mechanism; the detection result comprises the type and severity of the high-temperature corrosion defect and the corresponding defect position. According to the embodiment of the invention, the automatic and real-time detection of the high-temperature corrosion defect of the water cooling wall is effectively realized, the time and labor cost is reduced, and the operation risk is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of thermal power generation equipment maintenance, and in particular to a method and device, equipment and medium for detecting high temperature corrosion of a water-cooled wall of a coal-fired boiler. Background Art

[0002] In existing thermal power generating units, the furnace water-cooled wall is a key component of heat exchange. However, the water-cooled wall is exposed to high temperature, high pressure and corrosive gas environment for a long time, which is prone to high-temperature corrosion, leading to pipe leakage or bursting. In order to prevent and reduce the damage caused by high-temperature corrosion of the water-cooled wall, the existing technology usually adopts manual inspection methods to manually visually inspect the actual situation of the water-cooled wall. However, this manual visual inspection method is not only time-consuming and labor-intensive, but also difficult to cover all areas inside the furnace, and there are blind spots in the inspection. Summary of the invention

[0003] The present disclosure aims to solve at least one of the problems existing in the prior art and to provide a method and device, equipment and medium for detecting high temperature corrosion of water-cooled walls of coal-fired boilers.

[0004] In one aspect of the present disclosure, a method for detecting high temperature corrosion of a water-cooled wall of a coal-fired boiler is provided, the method comprising:

[0005] Obtain surface images of water-wall tubes of coal-fired boilers;

[0006] The surface image of the water-cooled wall tube is input into a trained high-temperature corrosion defect detection model to obtain corresponding detection results; wherein the high-temperature corrosion defect detection model is constructed based on the YOLOv8 model that introduces a feature pyramid network structure and an attention mechanism; the detection results include the type of high-temperature corrosion defects, their severity, and the corresponding defect location.

[0007] Optionally, the high temperature corrosion defect detection model is trained according to the following steps:

[0008] Collect sample images of the water-cooled wall tube surface and construct a training data set;

[0009] According to the high-temperature corrosion defect characteristics of water-cooled wall tubes, the feature pyramid network structure is added to the YOLOv8 model open source framework, and the channel attention mechanism is integrated to construct the high-temperature corrosion defect detection model;

[0010] Performing data enhancement processing on the training data set to obtain an enhanced data set;

[0011] The enhanced data set is used to train the high-temperature corrosion defect detection model using a preset training strategy to obtain the trained high-temperature corrosion defect detection model.

[0012] Optionally, the enhanced data set is used to train the high temperature corrosion defect detection model using a preset training strategy to obtain the trained high temperature corrosion defect detection model, including:

[0013] The high temperature corrosion defect detection model is initialized by applying a transfer learning strategy, and hyperparameters of the high temperature corrosion defect detection model are tuned based on the enhanced data set. The optimal hyperparameter combination is selected through cross-validation to obtain the trained high temperature corrosion defect detection model.

[0014] Optionally, performing hyperparameter tuning on the high temperature corrosion defect detection model based on the enhanced data set and selecting the optimal hyperparameter combination through cross validation includes:

[0015] During the hyperparameter tuning process, a dynamic learning rate strategy is used to dynamically adjust the learning rate;

[0016] During the cross-validation process, training is stopped when the validation loss no longer decreases, and the hyperparameter combination at this time is taken as the optimal hyperparameter combination.

[0017] Optionally, the data enhancement processing includes any one or more of random flipping, random rotation, brightness adjustment, and contrast adjustment.

[0018] Optionally, collecting sample images of the surface of water-cooled wall tubes to construct a training data set includes:

[0019] Collecting a plurality of sample images of the water-cooled wall tube surface with different types of high-temperature corrosion defects;

[0020] The high-temperature corrosion defect type and the corresponding corrosion area of ​​each sample image of the water-cooled wall tube surface are respectively marked.

[0021] Optionally, after obtaining the corresponding detection result, the detection method further includes:

[0022] The test results are optimized and analyzed to generate a corresponding test report; the test report includes any one or more of a defect location map, defect area, defect type statistics, and severity analysis.

[0023] Another aspect of the present disclosure provides a coal-fired boiler water-cooled wall high-temperature corrosion detection device, the detection device comprising:

[0024] An acquisition module, used for acquiring a surface image of a water-cooled wall tube of a coal-fired boiler;

[0025] The detection module is used to input the surface image of the water-cooled wall tube into a trained high-temperature corrosion defect detection model to obtain corresponding detection results; wherein the high-temperature corrosion defect detection model is constructed based on the YOLOv8 model that introduces a feature pyramid network structure and an attention mechanism; the detection results include the type of high-temperature corrosion defects, their severity, and the corresponding defect location.

[0026] Optionally, the detection device further includes a training module; the training module is used to train the high temperature corrosion defect detection model according to the following steps:

[0027] Collect sample images of the water-cooled wall tube surface and construct a training data set;

[0028] According to the high-temperature corrosion defect characteristics of water-cooled wall tubes, the feature pyramid network structure is added to the YOLOv8 model open source framework, and the channel attention mechanism is integrated to construct the high-temperature corrosion defect detection model;

[0029] Performing data enhancement processing on the training data set to obtain an enhanced data set;

[0030] The enhanced data set is used to train the high-temperature corrosion defect detection model using a preset training strategy to obtain the trained high-temperature corrosion defect detection model.

[0031] Optionally, the training module is used to adopt the enhanced data set and a preset training strategy to train the high temperature corrosion defect detection model to obtain the trained high temperature corrosion defect detection model, including:

[0032] The training module is used to: apply the transfer learning strategy to initialize the high-temperature corrosion defect detection model, perform hyperparameter tuning on the high-temperature corrosion defect detection model based on the enhanced data set, select the optimal hyperparameter combination through cross-validation, and obtain the trained high-temperature corrosion defect detection model.

[0033] Optionally, the training module is used to perform hyperparameter tuning on the high temperature corrosion defect detection model based on the enhanced data set, and select the optimal hyperparameter combination through cross-validation, including:

[0034] The training module is used to:

[0035] During the hyperparameter tuning process, a dynamic learning rate strategy is used to dynamically adjust the learning rate;

[0036] During the cross-validation process, training is stopped when the validation loss no longer decreases, and the hyperparameter combination at this time is taken as the optimal hyperparameter combination.

[0037] Optionally, the data enhancement processing includes any one or more of random flipping, random rotation, brightness adjustment, and contrast adjustment.

[0038] Optionally, the training module is used to collect sample images of the surface of water-cooled wall tubes to construct a training data set, including:

[0039] The training module is used to:

[0040] Collecting a plurality of sample images of the water-cooled wall tube surface with different types of high-temperature corrosion defects;

[0041] The high-temperature corrosion defect type and the corresponding corrosion area of ​​each sample image of the water-cooled wall tube surface are respectively marked.

[0042] Optionally, the detection device further includes:

[0043] The optimization and analysis module is used to optimize the test results, analyze the test results, and generate a corresponding test report; the test report includes any one or more of the defect location map, defect area, defect type statistics, and severity analysis.

[0044] Another aspect of the present disclosure provides an electronic device, including:

[0045] at least one processor; and,

[0046] a memory communicatively connected to at least one processor; wherein,

[0047] The memory stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor so that the at least one processor can execute the high-temperature corrosion detection method for the water-cooled wall of a coal-fired boiler described above.

[0048] Another aspect of the present disclosure provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for detecting high-temperature corrosion of a water-cooled wall of a coal-fired boiler.

[0049] Compared with the prior art, the present invention effectively realizes the automated and real-time detection of high-temperature corrosion defects of the water-cooled wall by utilizing a trained high-temperature corrosion defect detection model to perform high-temperature corrosion detection of the water-cooled wall based on the surface image of the water-cooled wall tube. This not only reduces time and labor costs, greatly improves detection efficiency, and reduces the burden of manual inspection during shutdown and maintenance of coal-fired units, but also can timely discover potential safety hazards of equipment, improve the accuracy and safety of equipment maintenance work, reduce unexpected shutdowns and economic losses caused by high-temperature corrosion, and thus reduce operational risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplifications do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0051] Figure 1 A flow chart of a method for detecting high temperature corrosion of a water-cooled wall of a coal-fired boiler provided in one embodiment of the present disclosure;

[0052] Figure 2 A schematic structural diagram of a coal-fired boiler water-cooled wall high-temperature corrosion detection device provided by another embodiment of the present disclosure;

[0053] Figure 3 A schematic structural diagram of an electronic device provided in another embodiment of the present disclosure. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical scheme and advantages of the embodiments of the present disclosure clearer, the embodiments of the present disclosure will be described in detail below in conjunction with the accompanying drawings. However, it can be understood by those skilled in the art that in each embodiment of the present disclosure, many technical details are proposed in order to enable readers to better understand the present disclosure. However, even without these technical details and various changes and modifications based on the following embodiments, the technical scheme claimed for protection in the present disclosure can also be implemented. The division of the following embodiments is for the convenience of description and should not constitute any limitation on the specific implementation of the present disclosure. The various embodiments can be combined and referenced with each other without contradiction.

[0055] One embodiment of the present disclosure relates to a method for detecting high temperature corrosion of a water-cooled wall of a coal-fired boiler, the process of which is as follows: Figure 1 As shown, including:

[0056] Step S110, obtaining a surface image of a water-cooled wall tube of a coal-fired boiler.

[0057] Specifically, step S110 can capture the surface image of the water-cooled wall tube of the coal-fired boiler in real time through an image acquisition system, or can obtain the surface image of the water-cooled wall tube of the coal-fired boiler from a pre-stored image, which is not limited in this embodiment.

[0058] Step S120, input the surface image of the water-cooled wall tube into the trained high-temperature corrosion defect detection model to obtain corresponding detection results; wherein the high-temperature corrosion defect detection model is constructed based on the YOLOv8 model that introduces a feature pyramid network structure and an attention mechanism; the detection results include the type of high-temperature corrosion defects, their severity, and the corresponding defect location.

[0059] Specifically, the trained high-temperature corrosion defect detection model can be directly deployed to the edge computing platform so that users can obtain the corresponding detection results through the edge computing platform. Alternatively, the trained high-temperature corrosion defect detection model can also be directly deployed to the image acquisition system to instantly analyze the real-time captured water-cooled wall tube surface image, quickly identify high-temperature corrosion defects, and obtain corresponding detection results.

[0060] The trained high-temperature corrosion defect detection model can not only locate the defect location, but also classify the types of high-temperature corrosion defects and their severity, thereby providing direct guidance for the subsequent maintenance of the water-cooled wall. For example, the types of high-temperature corrosion defects may include but are not limited to slight corrosion, pitting, spalling, etc. The severity may include but is not limited to a mild level, a moderate level, a severe level, etc. Among them, the severity corresponding to slight corrosion may be a mild level, the severity corresponding to pitting may be a moderate level, and the severity corresponding to spalling may be a severe level. Of course, the types of high-temperature corrosion defects may also include other types, and the severity may also include other levels, which are not limited in this embodiment.

[0061] Compared with the prior art, the method for detecting high-temperature corrosion of the water-cooled wall of a coal-fired boiler provided in the embodiment of the present disclosure performs high-temperature corrosion detection of the water-cooled wall based on the surface image of the water-cooled wall tube by utilizing a trained high-temperature corrosion defect detection model, thereby effectively realizing the automated and real-time detection of high-temperature corrosion defects of the water-cooled wall. This not only reduces time and labor costs, greatly improves detection efficiency, and alleviates the burden of manual inspection during shutdown and maintenance of coal-fired units, but also can timely discover potential safety hazards of equipment, improve the accuracy and safety of equipment maintenance work, and reduce unexpected shutdowns and economic losses caused by high-temperature corrosion, thereby reducing operational risks.

[0062] Exemplarily, the high-temperature corrosion defect detection model is trained according to the following steps: collecting sample images of the surface of water-cooled wall tubes to construct a training data set; adding a feature pyramid network (FPN) structure to the YOLOv8 model open source framework according to the characteristics of high-temperature corrosion defects of water-cooled wall tubes, and integrating a channel attention mechanism to construct a high-temperature corrosion defect detection model; performing data enhancement processing on the training data set to obtain an enhanced data set; using the enhanced data set and a preset training strategy to train the high-temperature corrosion defect detection model to obtain a trained high-temperature corrosion defect detection model.

[0063] Specifically, this embodiment builds a high-temperature corrosion defect detection model by customizing the YOLOv8 model, and uses the constructed data set to train the high-temperature corrosion defect detection model.

[0064] Exemplarily, sample images of the surface of water-cooled wall tubes are collected to construct a training data set, including: collecting multiple sample images of the surface of water-cooled wall tubes with different types of high-temperature corrosion defects; and labeling the high-temperature corrosion defect type and the corresponding corrosion area of ​​each sample image of the surface of water-cooled wall tubes.

[0065] Specifically, when collecting sample images of the surface of the water-cooled wall tube, during the shutdown and maintenance of the coal-fired unit, the sample images of the surface of the water-cooled wall tube can be collected manually by entering the furnace, or by using platforms such as drones. The collected sample images of the surface of the water-cooled wall tube need to cover images in a normal state and images corresponding to various types of high-temperature corrosion defects and various degrees of severity. Among them, the high-temperature corrosion defect type corresponding to the image in a normal state is no corrosion. In addition to no corrosion, high-temperature corrosion defect types can also include slight corrosion, pitting, peeling, etc. The severity levels include mild level, moderate level, severe level, etc.

[0066] When annotating sample images of the water-wall tube surface, professional software can be used to accurately annotate the corrosion areas in each sample image of the water-wall tube surface, including annotating contour boundary boxes and annotations of the type and severity of high-temperature corrosion defects, to form a high-quality data set and provide an accurate reference for model training.

[0067] When building a high-temperature corrosion defect detection model, in order to enhance the recognition ability of defects of different scales, the network structure can be adjusted based on the YOLOv8 model open source framework according to the characteristics of high-temperature corrosion defects of water-cooled wall tubes, and a feature pyramid network structure can be added to the YOLOv8 model open source framework. At the same time, in order to improve the model's attention to key features and enhance the model's feature expression capabilities, a channel attention mechanism can also be integrated into the model, for example, the SE (Squeeze-and-Excitation) block is added to the YOLOv8 model open source framework. Among them, the SE block is a channel attention model that increases the attention to important channels by adjusting the weight of each channel, thereby improving the expressiveness of the model.

[0068] In order to simulate images under different illumination and viewing angles and enhance the generalization ability of the model, it is necessary to perform data augmentation processing on the training data set. Exemplarily, the data augmentation processing may include any one or more of random flipping, random rotation, brightness adjustment, and contrast adjustment.

[0069] Exemplarily, an enhanced data set is used and a preset training strategy is utilized to train a high-temperature corrosion defect detection model to obtain a trained high-temperature corrosion defect detection model, including: applying a transfer learning strategy to initialize the high-temperature corrosion defect detection model, performing hyperparameter tuning on the high-temperature corrosion defect detection model based on the enhanced data set, and selecting the optimal hyperparameter combination through cross-validation to obtain a trained high-temperature corrosion defect detection model.

[0070] Specifically, when the high-temperature corrosion defect detection model is trained using a preset training strategy, an enhanced data set is used for training, a transfer learning strategy is applied to initialize the high-temperature corrosion defect detection model, and then fine-tuning is performed on a custom data set to optimize the classification performance and positioning performance of the model. The custom data set may be a data set consisting of labeled water-cooled wall tube surface sample images selected from the enhanced data set.

[0071] Exemplarily, hyperparameters of a high-temperature corrosion defect detection model are tuned based on an enhanced data set, and an optimal hyperparameter combination is selected through cross-validation, including: during the hyperparameter tuning process, dynamically adjusting the learning rate using a dynamic learning rate strategy; during the cross-validation process, training is stopped when the validation loss no longer decreases, and the hyperparameter combination at this time is used as the optimal hyperparameter combination.

[0072] Specifically, this embodiment adopts a dynamic learning rate strategy and early stopping method to ensure training efficiency and model performance. The hyperparameter combination includes a combination of learning rate, batch size, loss function weight, etc. By cross-validating and selecting the optimal hyperparameter combination, the model training process can be ensured to be efficient and the results to be optimal.

[0073] Exemplarily, after obtaining the corresponding test results, the high-temperature corrosion detection method for the water-cooled wall of a coal-fired boiler also includes: optimizing the test results, analyzing the test results, and generating a corresponding test report; the test report includes any one or more of a defect location map, defect area, defect type statistics, and severity analysis.

[0074] Specifically, this embodiment does not limit the specific method of optimizing the detection results. For example, when optimizing the detection results, morphological operations can be used to remove noise in the detection results, and the position and size of the bounding box corresponding to the defect position can be optimized, thereby improving the accuracy and aesthetics of the final detection results.

[0075] When analyzing the test results, the automated report generation system can be used to organize the test results into an easy-to-understand format, generate a test report including defect location maps, defect areas, defect type statistics, severity analysis, etc., and give corresponding equipment maintenance recommendations to facilitate technicians to make quick decisions.

[0076] Another embodiment of the present disclosure relates to a high temperature corrosion detection device for water-cooled wall of a coal-fired boiler, such as Figure 2 As shown, it includes an acquisition module 210 and a detection module 220.

[0077] The acquisition module 210 is used to acquire the surface image of the water-cooled wall tube of the coal-fired boiler.

[0078] The detection module 220 is used to input the surface image of the water-cooled wall tube into the trained high-temperature corrosion defect detection model to obtain the corresponding detection results; wherein, the high-temperature corrosion defect detection model is constructed based on the YOLOv8 model that introduces the feature pyramid network structure and attention mechanism; the detection results include the type of high-temperature corrosion defects, their severity and the corresponding defect location.

[0079] Exemplarily, the high-temperature corrosion detection device for the water-cooled wall of a coal-fired boiler also includes a training module; the training module is used to train the high-temperature corrosion defect detection model according to the following steps: collect sample images of the surface of the water-cooled wall tube to construct a training data set; according to the high-temperature corrosion defect characteristics of the water-cooled wall tube, add a feature pyramid network structure to the YOLOv8 model open source framework, and integrate a channel attention mechanism to construct a high-temperature corrosion defect detection model; perform data enhancement processing on the training data set to obtain an enhanced data set; use the enhanced data set and a preset training strategy to train the high-temperature corrosion defect detection model to obtain a trained high-temperature corrosion defect detection model.

[0080] Exemplarily, the training module is used to train the high temperature corrosion defect detection model using an enhanced data set and a preset training strategy to obtain a trained high temperature corrosion defect detection model, including:

[0081] The training module is used to: apply the transfer learning strategy to initialize the high-temperature corrosion defect detection model, perform hyperparameter tuning on the high-temperature corrosion defect detection model based on the enhanced data set, select the optimal hyperparameter combination through cross-validation, and obtain a trained high-temperature corrosion defect detection model.

[0082] Exemplarily, the training module is used to perform hyperparameter tuning on the high temperature corrosion defect detection model based on the enhanced data set, and select the optimal hyperparameter combination through cross validation, including:

[0083] The training module is used to: During the hyperparameter tuning process, a dynamic learning rate strategy is used to dynamically adjust the learning rate; during the cross-validation process, training is stopped when the validation loss no longer decreases, and the hyperparameter combination at this time is used as the optimal hyperparameter combination.

[0084] Exemplarily, the data enhancement processing includes any one or more of random flipping, random rotation, brightness adjustment, and contrast adjustment.

[0085] Exemplarily, the training module is used to collect sample images of the surface of water-cooled wall tubes and construct a training data set, including:

[0086] The training module is used to: collect multiple sample images of the surface of water-cooled wall tubes with different types of high-temperature corrosion defects; and respectively mark the type of high-temperature corrosion defects and the corresponding corrosion area of ​​each sample image of the surface of water-cooled wall tubes.

[0087] Exemplarily, the coal-fired boiler water-cooled wall high-temperature corrosion detection device also includes an optimization analysis module. The optimization analysis module is used to optimize the detection results, analyze the detection results, and generate a corresponding detection report; the detection report includes any one or more of the defect location map, defect area, defect type statistics, and severity analysis.

[0088] The specific implementation method of the high-temperature corrosion detection device for the water-cooled wall of a coal-fired boiler provided in the embodiment of the present disclosure can be referred to the high-temperature corrosion detection method for the water-cooled wall of a coal-fired boiler provided in the embodiment of the present disclosure, and will not be repeated here.

[0089] Compared with the prior art, the high-temperature corrosion detection device for the water-cooled wall of a coal-fired boiler provided in the embodiment of the present invention performs high-temperature corrosion detection of the water-cooled wall based on the surface image of the water-cooled wall tube by utilizing a trained high-temperature corrosion defect detection model, thereby effectively realizing the automated and real-time detection of high-temperature corrosion defects of the water-cooled wall. This not only reduces time and labor costs, greatly improves detection efficiency, and alleviates the burden of manual inspection during shutdown and maintenance of coal-fired units, but also can promptly discover potential safety hazards of equipment, improve the accuracy and safety of equipment maintenance work, and reduce unexpected shutdowns and economic losses caused by high-temperature corrosion, thereby reducing operational risks.

[0090] Another embodiment of the present disclosure relates to an electronic device, such as Figure 3 As shown, including:

[0091] at least one processor 301; and,

[0092] A memory 302 is communicatively connected to at least one processor 301; wherein,

[0093] The memory 302 stores instructions that can be executed by at least one processor 301, and the instructions are executed by at least one processor 301 so that the at least one processor 301 can execute the high-temperature corrosion detection method for the water-cooled wall of a coal-fired boiler described in the above embodiment.

[0094] Among them, the memory and the processor are connected in a bus manner, and the bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors and memories together. The bus can also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be one element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices on a transmission medium. The data processed by the processor is transmitted on a wireless medium via an antenna, and further, the antenna also receives data and transmits the data to the processor.

[0095] The processor is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory can be used to store data used by the processor when performing operations.

[0096] Another embodiment of the present disclosure relates to a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method for detecting high-temperature corrosion of a water-cooled wall of a coal-fired boiler described in the above embodiment.

[0097] That is, those skilled in the art can understand that all or part of the steps in the method described in the above embodiments can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including several instructions for making a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) perform all or part of the steps of the method described in each embodiment of the present disclosure. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0098] Those skilled in the art will appreciate that the above-mentioned embodiments are specific embodiments for implementing the present disclosure, and in actual applications, various changes may be made thereto in form and detail without departing from the spirit and scope of the present disclosure.

Claims

1. A method for detecting high temperature corrosion of water-cooled walls of coal-fired boilers, characterized in that: The detection method comprises: Obtain surface images of water-wall tubes of coal-fired boilers; The surface image of the water-cooled wall tube is input into a trained high-temperature corrosion defect detection model to obtain corresponding detection results; wherein the high-temperature corrosion defect detection model is constructed based on the YOLOv8 model that introduces a feature pyramid network structure and an attention mechanism; the detection results include the type of high-temperature corrosion defects, their severity, and the corresponding defect location.

2. The detection method according to claim 1, characterized in that: The high temperature corrosion defect detection model is trained according to the following steps: Collect sample images of the water-cooled wall tube surface and construct a training data set; According to the high-temperature corrosion defect characteristics of water-cooled wall tubes, the feature pyramid network structure is added to the YOLOv8 model open source framework, and the channel attention mechanism is integrated to construct the high-temperature corrosion defect detection model; Performing data enhancement processing on the training data set to obtain an enhanced data set; The enhanced data set is adopted and a preset training strategy is used to train the high-temperature corrosion defect detection model to obtain the trained high-temperature corrosion defect detection model.

3. The detection method according to claim 2, characterized in that: The enhanced data set is adopted and a preset training strategy is used to train the high temperature corrosion defect detection model to obtain the trained high temperature corrosion defect detection model, including: The high temperature corrosion defect detection model is initialized by applying a transfer learning strategy, and hyperparameters of the high temperature corrosion defect detection model are tuned based on the enhanced data set. The optimal hyperparameter combination is selected through cross-validation to obtain the trained high temperature corrosion defect detection model.

4. The detection method according to claim 3, characterized in that: The hyperparameter tuning of the high temperature corrosion defect detection model based on the enhanced data set and selecting the optimal hyperparameter combination through cross validation include: During the hyperparameter tuning process, a dynamic learning rate strategy is used to dynamically adjust the learning rate; During the cross-validation process, training is stopped when the validation loss no longer decreases, and the hyperparameter combination at this time is taken as the optimal hyperparameter combination.

5. The detection method according to any one of claims 2 to 4, characterized in that: The data enhancement processing includes any one or more of random flipping, random rotation, brightness adjustment, and contrast adjustment.

6. The detection method according to claim 5, characterized in that: The collecting of sample images of the water wall tube surface and constructing a training data set includes: Collecting a plurality of sample images of the water-cooled wall tube surface with different types of high-temperature corrosion defects; The high-temperature corrosion defect type and the corresponding corrosion area of ​​each sample image of the water-cooled wall tube surface are respectively marked.

7. The detection method according to claim 1, characterized in that: After obtaining the corresponding detection result, the detection method further includes: The test results are optimized and analyzed to generate a corresponding test report; the test report includes any one or more of a defect location map, defect area, defect type statistics, and severity analysis.

8. A coal-fired boiler water wall high temperature corrosion detection device, characterized in that: The detection device comprises: An acquisition module, used for acquiring a surface image of a water-cooled wall tube of a coal-fired boiler; The detection module is used to input the surface image of the water-cooled wall tube into a trained high-temperature corrosion defect detection model to obtain corresponding detection results; wherein the high-temperature corrosion defect detection model is constructed based on the YOLOv8 model that introduces a feature pyramid network structure and an attention mechanism; the detection results include the type of high-temperature corrosion defects, their severity, and the corresponding defect location.

9. An electronic device, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the high-temperature corrosion detection method for the water-cooled wall of a coal-fired boiler according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for detecting high-temperature corrosion of water-cooled walls of a coal-fired boiler according to any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Boiler water wall tube defect identification system and method and storage medium

    CN115661429A

  • Defect detection method, device and equipment for water cooling wall of power station boiler and medium

    CN117197113A

  • AI-based boiler furnace water wall tube wall image acquisition and processing system

    CN117745945A

  • Method for segmentation of underground drainage pipeline defects based on full convolutional neural network

    US20210319265A1

Cited By

  • Steel structure surface corrosion identification method and device, medium and equipment

    CN120235852A

  • Anti-corrosion control system and method for pipeline operation robot

    CN120287314A