A thermal power plant boiler overhauling method, system, device and storage medium

Abstract

The present application relates to a kind of thermal power plant boiler overhauling method, system, equipment and storage medium, the method includes the following steps: data acquisition robot is arranged in thermal power plant boiler, and boiler state data is collected by data acquisition robot;Boiler defect identification model is constructed, and boiler state data is input into boiler defect identification model, and the boiler defect type and defect position are obtained;Boiler state prediction model is constructed, and boiler state data and historical maintenance data are input into boiler state prediction model, and the boiler predicted state data are obtained;Boiler risk classification model is constructed, and whether current boiler needs maintenance is judged based on boiler predicted state data;If current boiler needs maintenance, then according to the boiler defect type, corresponding maintenance robot is called to repair the corresponding defect position of boiler.

Description

Technical Field

[0001] This invention relates to a method, system, equipment, and storage medium for the maintenance of boilers in thermal power plants, belonging to the field of boiler maintenance technology for thermal power plants. Background Technology

[0002] As the core power equipment for power generation, the operational stability of boilers in thermal power plants directly determines the unit's power generation efficiency, safety level, and economic benefits. With the thermal power industry developing towards higher parameters, larger capacity, and cleaner production, the internal structure of boilers is becoming increasingly complex. Heating surface components are subjected to harsh conditions such as high temperature, high pressure, corrosion, and wear for extended periods, making them prone to potential malfunctions such as cracks, corrosion, ash accumulation, and thinning of the wall.

[0003] Traditional boiler maintenance methods, primarily relying on manual inspections and periodic disassembly and maintenance, suffer from several insurmountable technical bottlenecks: First, maintenance efficiency is low. Manual inspections require entry into enclosed spaces such as the furnace and flue, resulting in harsh working conditions and cumbersome procedures. For a 600MW boiler unit, a single comprehensive inspection requires a team of 8-10 people and takes 5-7 days, failing to meet the needs of rapid unit maintenance. Second, detection accuracy is limited. Relying on manual visual observation and portable instruments, it is difficult to detect early defects such as micro-cracks less than 0.2mm wide and latent corrosion, resulting in a missed detection rate as high as 15%-20%, which can easily lead to minor hidden dangers developing into major equipment failures. Third, safety risks are prominent. Enclosed spaces pose safety hazards such as high-temperature burns, toxic gas poisoning, and falls from heights, making it difficult to guarantee the personal safety of workers. Fourth, maintenance costs are high. Periodic maintenance suffers from "over-maintenance" or "under-maintenance," with ineffective maintenance costs accounting for over 30%, and significant economic losses caused by unplanned shutdowns.

[0004] In recent years, the rapid iteration of artificial intelligence, robotics and Internet of Things technologies has made it possible to upgrade boiler maintenance, but existing technologies still have shortcomings: most existing inspection robots are designed for single functions and lack the ability to adapt to multiple scenarios; data processing is mostly limited to single-dimensional analysis and has not achieved deep integration of multi-source data; fault diagnosis algorithms have insufficient generalization ability and are difficult to adapt to the diverse fault types under the complex operating conditions of boilers. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention proposes a method, system, equipment, and storage medium for boiler maintenance in thermal power plants.

[0006] The technical solution of the present invention is as follows: On the one hand, the present invention provides a method for overhauling a boiler in a thermal power plant, comprising the following steps: Data acquisition robots are installed in the boilers of thermal power plants to collect boiler status data. Construct a boiler defect identification model, input boiler status data into the boiler defect identification model, and obtain the boiler defect type and defect location; Construct a boiler condition prediction model by inputting boiler condition data and historical maintenance data into the boiler condition prediction model to obtain boiler predicted condition data. Construct a boiler risk classification model to determine whether the current boiler needs maintenance based on boiler prediction status data; If the boiler needs maintenance, the corresponding maintenance robot will be called to repair the corresponding defect location based on the boiler defect type.

[0007] Preferably, the data acquisition robot includes a wall-climbing robot and a flying robot.

[0008] Preferably, the wall-climbing robot is installed on the boiler water-cooled wall surface and is equipped with a camera device, a laser rangefinder and an ultrasonic flaw detector to collect real-time image data, wall thickness data and surface defect data of the water-cooled wall surface; The flying robot is equipped with a camera device and an infrared thermal imager to collect surface image data and temperature distribution data of the boiler furnace, superheater, and reheater tube bundle.

[0009] Preferably, a three-dimensional model of the boiler is constructed, and the position coordinates of the wall-climbing robot and the flying robot are mapped in real time relative to the spatial coordinates of the three-dimensional model of the boiler. The coordinates of the climbing robot and the flying robot are set in the 3D model of the boiler. The climbing robot and the flying robot move according to the corresponding coordinates of the climbing robot and collect boiler status data.

[0010] Preferably, the boiler defect identification model is implemented based on an image recognition model constructed using the YOLOv8 algorithm.

[0011] Preferably, the boiler condition prediction model is constructed based on a long short-term memory network.

[0012] Preferably, a boiler risk classification model is constructed based on the random forest algorithm; The boiler risk level is determined by predicting the boiler's status data, and the boiler's need for maintenance and maintenance time limit are determined according to the preset rules for different risk levels.

[0013] On the other hand, the present invention provides a boiler maintenance system for thermal power plants, including a robot data acquisition module, a defect identification module, a status prediction module, a risk classification module, and a maintenance module. The robot data acquisition module is used to control the data acquisition robot installed in the boiler of the thermal power plant to collect boiler status data. The defect identification module is used to construct a boiler defect identification model. Boiler status data is input into the boiler defect identification model to obtain the boiler defect type and defect location. The state prediction module is used to construct a boiler state prediction model. Boiler state data and historical maintenance data are input into the boiler state prediction model to obtain boiler predicted state data. The risk classification module is used to construct a boiler risk classification model and determine whether the current boiler needs maintenance based on the boiler's predicted status data. The maintenance module is used to call the corresponding maintenance robot to repair the corresponding defect location of the boiler according to the boiler defect type.

[0014] In another aspect, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in the present invention.

[0015] In another aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the method described in the present invention.

[0016] The present invention has the following beneficial effects: 1. This invention uses wall-climbing robots and flying robots to collect multi-source data on key components such as water-cooled walls, furnaces, superheaters / reheaters, and automatically outputs the defect type and location by a defect identification model, which significantly reduces the frequency and intensity of manual entry into high-temperature and high-risk spaces and improves maintenance safety and efficiency.

[0017] 2. The boiler status prediction model of this invention integrates current status data and historical maintenance data to output predicted status. Then, the risk classification model determines whether maintenance is needed and the maintenance time limit, so as to realize early warning of deterioration trend and reduce sudden shutdowns and unplanned maintenance. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0021] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0022] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0023] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0024] See Figure 1 In some embodiments, a method for overhauling a thermal power plant boiler is proposed, including the following steps: Data acquisition robots are installed in the boilers of thermal power plants to collect boiler status data. Construct a boiler defect identification model, input boiler status data into the boiler defect identification model, and obtain the boiler defect type and defect location; Construct a boiler condition prediction model by inputting boiler condition data and historical maintenance data into the boiler condition prediction model to obtain boiler predicted condition data. Construct a boiler risk classification model to determine whether the current boiler needs maintenance based on boiler prediction status data; If the boiler needs maintenance, the corresponding maintenance robot will be called to repair the corresponding defect location based on the boiler defect type.

[0025] In some embodiments, the data acquisition robot includes a wall-climbing robot and a flying robot.

[0026] In some embodiments, the wall-climbing robot is installed on the boiler water-cooled wall surface and is equipped with a camera device, a laser rangefinder and an ultrasonic flaw detector to collect real-time image data, wall thickness data and surface defect data of the water-cooled wall surface. The flying robot is equipped with a camera device and an infrared thermal imager to collect surface image data and temperature distribution data of the boiler furnace, superheater, and reheater tube bundle.

[0027] In one specific embodiment, the wall-climbing robot adopts a permanent magnet adsorption tracked chassis, which is adapted to vertical or inclined surfaces such as boiler water-cooled walls. It carries a high-definition industrial camera (resolution ≥1920×1080, frame rate ≥30fps), a high-precision laser rangefinder (measurement accuracy ±0.1mm) and an ultrasonic flaw detector (detection depth 0-50mm), and supports autonomous planning of the climbing path. With a maximum climbing angle of 45° and an adsorption force of ≥200N, it ensures stable operation in high-temperature environments; it calibrates its own position in real time through a magnetic navigation + laser ranging and positioning module, controls the track walking mechanism to adjust the walking speed and direction, and accurately reaches the target area; at the same time, it performs basic dust removal tasks. The flying robot adopts a multi-rotor anti-interference design, is equipped with a high-definition camera and an infrared thermal imager (temperature measurement range -20℃-1200℃, temperature measurement accuracy ±1℃), and is equipped with obstacle avoidance radar and a high-precision positioning module. Relying on the UWB+visual SLAM positioning module, it collects its own pose data in real time, and adjusts its flight attitude and speed autonomously in combination with the preset cruise path. It also controls the onboard detection equipment (high-definition camera, thermal imager) to collect data and transmit it back in real time.

[0028] In one specific embodiment, maintenance tasks are performed by a robotic arm. The robotic arm adopts a lightweight design with degrees of freedom, and its end can be equipped with a dust cleaning brush head, a small parts replacement tool, and other work terminals. It has a load capacity of ≥5kg and a repeatability of ±0.05mm. It can be used in conjunction with a wall-climbing robot or independently to perform maintenance operations such as dust cleaning and replacement of small vulnerable parts. After receiving the repair task and defect parameter instructions, and manually confirming the repair process parameters (welding current, coating thickness, tightening torque), the robotic arm, relying on the fixed coordinates of the base and the joint encoder positioning module, accurately locates the repair point and controls the multi-degree-of-freedom arm to perform precise repair operations such as welding, coating spraying, and bolt tightening according to the preset trajectory. The operation progress and parameters are fed back in real time to ensure that the repair accuracy meets the boiler maintenance standards.

[0029] In some embodiments, a three-dimensional model of the boiler is constructed, and the position coordinates of the wall-climbing robot and the flying robot are mapped in real time relative to the spatial coordinates of the three-dimensional model of the boiler. The coordinates of the climbing robot and the flying robot are set in the 3D model of the boiler. The climbing robot and the flying robot move according to the corresponding coordinates of the climbing robot and collect boiler status data.

[0030] In one specific embodiment, the robot is scheduled using a hierarchical cooperative scheduling algorithm: The flying robot is responsible for large-scale inspections, initial screening of defects, and marking of area coordinates; The wall-climbing robot is responsible for close-range and precise verification of defects, basic dust removal, and collection of re-inspection and acceptance data. The robotic arm is responsible for performing precise repair operations such as welding, coating repair, and fastening for the defects that have been verified. The collaborative relay sequence is as follows: flying robot inspection → uploading suspected defects → triggering wall-climbing robot relay, wall-climbing robot review → completing basic processing as needed → triggering robotic arm relay (when repair is required) or directly entering re-inspection (when no repair is required), robotic arm completes repair → triggering wall-climbing robot + flying robot collaborative re-inspection; Set up conflict resolution strategies. For path conflicts, use pre-planned paths + dynamic obstacle avoidance. Adjust the timing or path according to "core operation priority". For communication conflicts, use time division multiple access + priority sorting (robotic arm > wall climbing robot > flying robot) to ensure command transmission. For resource conflicts, allocate work areas and shared resources according to "detection priority, repair later" to avoid competition.

[0031] In one specific embodiment, the image acquisition card and sensor data acquisition module are used to synchronously acquire and standardize the multi-source data such as images, wall thickness, temperature, and distance obtained by the robot. The sampling frequency can be adaptively adjusted within the range of 10-100Hz according to the detection requirements. The system adopts a dual-mode communication architecture of "5G + wired industrial Ethernet". 5G communication is used for wireless data transmission during robot movement, with a transmission rate of ≥1Gbps and a latency of ≤50ms, ensuring real-time performance in mobile scenarios; wired industrial Ethernet is used for fixed areas or large data transmission, ensuring data transmission stability. By employing noise reduction, filtering, and data completion algorithms, the collected raw data is denoised, outliers are removed, and missing values ​​are filled in, thereby improving data quality and providing reliable input for subsequent analysis.

[0032] In one specific embodiment, the construction of the three-dimensional model is based on boiler CAD, and the high-precision three-dimensional laser scanning data from the site is integrated to complete the calibration. The model is equipped with attribute information such as equipment parameters and maintenance records to construct an integrated "geometry + attribute" model. The system adopts a dynamic update method that combines regular full updates with real-time incremental updates. During scheduled maintenance, full reconstruction and calibration are completed through laser scanning. During routine maintenance, the robot transmits data to update incremental information such as model defects and component status in real time. Select the fixed structural components of the boiler as physical calibration points, register them with the model reference points, and establish a global mapping relationship between the on-site physical coordinate system and the model digital coordinate system; The robot converts the on-site pose data into model coordinates using a global coordinate transformation algorithm, transmits the data back in real time, and renders the position, pose, and path in the model. The 3D model supports scaling, rotation, and sectioning operations, providing a visual representation of the boiler's interior. Set up a data statistics visualization interface to centrally display key information such as detection data, defect statistics, fault warnings, and maintenance progress. It supports data export and historical traceability, and generates periodic maintenance reports. It supports access from multiple terminals including PCs, tablets, and mobile devices, allowing technicians to remotely monitor the maintenance process in real time, issue control commands, and achieve cross-regional collaborative maintenance. It also features access control to ensure data security and operational standards.

[0033] In some embodiments, the boiler defect identification model is implemented based on an image recognition model constructed using the YOLOv8 algorithm.

[0034] In one specific embodiment, boiler status data is encoded by an encoder to obtain feature data, which is then input into an image recognition model based on the YOLOv8 algorithm. The boiler defect identification model outputs eight types of defects: cracks, corrosion, ash accumulation, deformation, wear, weld defects, foreign matter attachment, and loose components. The image recognition model built by the YOLOv8 algorithm is optimized for small target detection, including lightweight adjustment of the network structure, loss function optimization (to improve defect localization accuracy), and targeted data augmentation (to adapt to complex boiler operating conditions). At the same time, the small target feature extraction branch is strengthened to meet the needs of micro-crack detection.

[0035] In some embodiments, the boiler condition prediction model is constructed based on a long short-term memory network; In a specific embodiment, the boiler condition prediction model predicts the wall thickness reduction trend, thermal deviation trend, and crack propagation probability, while outputting a binary classification judgment of "fault status" to comprehensively characterize the equipment fault development trend. To improve prediction accuracy and address the issue of sample imbalance, a combination of oversampling (for a few faulty samples) and undersampling (for normal samples) is used to ensure the fairness of the model's predictions. The model can provide early warnings of potential fault risks 7-30 days in advance, and the warning duration can be dynamically adjusted according to the equipment's operating conditions, allowing sufficient time for maintenance decisions.

[0036] In some embodiments, a boiler risk classification model is constructed based on the random forest algorithm; The boiler risk level is determined by predicting the boiler's status data, and the boiler's need for maintenance and maintenance time limit are determined according to the preset rules for different risk levels.

[0037] In one specific embodiment, a random forest algorithm is used to classify and grade the fault type and severity (divided into four levels: minor, moderate, severe, and urgent). Each level has clearly defined and quantifiable conditions to provide a basis for maintenance decisions. Minor faults are defined as dust accumulation thickness < 5 mm, component loosening torque deviation ≤ 10%, no visible cracks, and corrosion area < 5 cm², which do not affect the normal operation of the equipment. Moderate faults are defined as dust accumulation thickness 5-15 mm, crack width 0.1-0.3 mm and length < 50 mm, and corrosion area 5-20 cm², which do not affect short-term operation but require time-limited treatment. Severe faults are defined as dust accumulation thickness > 15 mm, crack width 0.3-0.5 mm and length 50-100 mm, corrosion area > 20 cm², and wall thickness reduction 5%-10%, which affect the operating efficiency of the equipment and require immediate treatment. Urgent faults are defined as crack width > 0.5 mm and length ≥ 100 mm, wall thickness reduction > 10%, valve leakage > 5%, and obvious deformation, which pose a safety hazard and require immediate shutdown. The robot automatically performs defect location, light / moderate dust removal, and data collection and preliminary judgment for re-inspection and acceptance. After manual confirmation, the robot performs general fault repair welding / coating repair and bolt tightening. For serious / emergency faults, the robot performs repair welding / coating repair, pipe / valve replacement, and final judgment and archiving for re-inspection and acceptance.

[0038] Based on the fault type, severity, equipment operating status, and historical maintenance data, targeted maintenance suggestions are automatically generated, including maintenance priorities, work procedures, required tools, and safety precautions. For minor faults (such as slight dust accumulation), robot-automated execution instructions are directly generated; for general faults (such as micro-cracks), robot repair and manual verification plans are generated; for serious faults (such as large-area corrosion and through-cracks), detailed manual maintenance plans and technical parameters are generated; and for emergency faults (such as through-cracks and significant wall thickness reduction), emergency shutdown maintenance plans are generated and emergency response measures are specified.

[0039] In some embodiments, a boiler maintenance system for thermal power plants is proposed, including a robot data acquisition module, a defect identification module, a status prediction module, a risk classification module, and a maintenance module. The robot data acquisition module is used to control the data acquisition robot installed in the boiler of the thermal power plant to collect boiler status data. The defect identification module is used to construct a boiler defect identification model. Boiler status data is input into the boiler defect identification model to obtain the boiler defect type and defect location. The state prediction module is used to construct a boiler state prediction model. Boiler state data and historical maintenance data are input into the boiler state prediction model to obtain boiler predicted state data. The risk classification module is used to construct a boiler risk classification model and determine whether the current boiler needs maintenance based on the boiler's predicted status data. The maintenance module is used to call the corresponding maintenance robot to repair the corresponding defect location of the boiler according to the boiler defect type.

[0040] In some embodiments, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method as described in any embodiment of the present invention.

[0041] In some embodiments, a computer-readable storage medium is provided on which a computer program is stored, which, when executed by a processor, implements the method as described in any embodiment of the present invention.

[0042] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, A and B simultaneously, or B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0043] Those skilled in the art will recognize that the units and algorithm steps described in the embodiments disclosed herein can be implemented using electronic hardware, computer software, or a combination of electronic hardware and software. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0044] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0045] In the several embodiments provided in this application, any function, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0046] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for overhauling a boiler in a thermal power plant, characterized in that, Includes the following steps: Data acquisition robots are installed in the boilers of thermal power plants to collect boiler status data. Construct a boiler defect identification model, input boiler status data into the boiler defect identification model, and obtain the boiler defect type and defect location; Construct a boiler condition prediction model by inputting boiler condition data and historical maintenance data into the boiler condition prediction model to obtain boiler predicted condition data. Construct a boiler risk classification model to determine whether the current boiler needs maintenance based on boiler prediction status data; If the boiler needs maintenance, the corresponding maintenance robot will be called to repair the corresponding defect location based on the boiler defect type.

2. The method for overhauling a thermal power plant boiler according to claim 1, characterized in that, The data acquisition robots include wall-climbing robots and flying robots.

3. A method for overhauling a thermal power plant boiler according to claim 2, characterized in that, The wall-climbing robot is installed on the boiler water-cooled wall surface and is equipped with a camera device, a laser rangefinder and an ultrasonic flaw detector to collect real-time image data, wall thickness data and surface defect data of the water-cooled wall surface. The flying robot is equipped with a camera device and an infrared thermal imager to collect surface image data and temperature distribution data of the boiler furnace, superheater, and reheater tube bundle.

4. A method for overhauling a thermal power plant boiler according to claim 3, characterized in that, Construct a 3D model of the boiler and map the position coordinates of the wall-climbing robot and the flying robot to the spatial coordinates of the boiler 3D model in real time. The coordinates of the climbing robot and the flying robot are set in the 3D model of the boiler. The climbing robot and the flying robot move according to the corresponding coordinates of the climbing robot and collect boiler status data.

5. A method for overhauling a thermal power plant boiler according to claim 1, characterized in that, The boiler defect identification model is implemented based on an image recognition model built using the YOLOv8 algorithm.

6. A method for overhauling a thermal power plant boiler according to claim 1, characterized in that, The boiler condition prediction model is constructed based on a long short-term memory network.

7. A method for overhauling a thermal power plant boiler according to claim 1, characterized in that, A boiler risk classification model was constructed based on the random forest algorithm. The boiler risk level is determined by predicting the boiler's status data, and the boiler's need for maintenance and maintenance time limit are determined according to the preset rules for different risk levels.

8. A boiler maintenance system for thermal power plants, characterized in that, It includes a robot data acquisition module, a defect identification module, a status prediction module, a risk classification module, and a maintenance module; The robot data acquisition module is used to control the data acquisition robot installed in the boiler of the thermal power plant to collect boiler status data. The defect identification module is used to construct a boiler defect identification model. Boiler status data is input into the boiler defect identification model to obtain the boiler defect type and defect location. The state prediction module is used to construct a boiler state prediction model. Boiler state data and historical maintenance data are input into the boiler state prediction model to obtain boiler predicted state data. The risk classification module is used to construct a boiler risk classification model and determine whether the current boiler needs maintenance based on the boiler's predicted status data. The maintenance module is used to call the corresponding maintenance robot to repair the corresponding defect location of the boiler according to the boiler defect type.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 7.

Images