Abnormal Monitoring Method and Device for Boiler Tubes of Thermal Power Stations

Abstract

The present invention relates to the technical field of operation and maintenance of thermal power plant boilers, and particularly to a method and device for abnormal monitoring of boiler tubes in thermal power plants. The method includes: collecting the wall temperatures of each heat exchange node arranged in sequence along the medium flow direction in the boiler tubes of the thermal power plant; inputting all the wall temperatures into a pre-trained first abnormal monitoring model to output a first judgment result; collecting multi-source data of all heat exchange nodes, where the multi-source data includes the physical and chemical characteristics of the boiler tubes, the service life characteristics of the boiler tubes, the static fluctuation characteristics of the boiler tubes, the surface image characteristics of the boiler tubes, and the surface heat distribution characteristics of the boiler tubes; inputting the multi-source data into a pre-trained second abnormal monitoring model to output a second judgment result for further monitoring of abnormal nodes, where the second judgment result includes the scores of all heat exchange nodes under the current working state, and the abnormal nodes are the heat exchange nodes with scores lower than a preset value in the second judgment result. This solution can accurately monitor the abnormal conditions of the boiler tubes.

Description

Technical Field

[0001] The present invention relates to the technical field of operation and maintenance of boilers in thermal power plants, and particularly to a method and device for abnormal monitoring of boiler tubes in thermal power plants. Background Art

[0002] As one of the core equipment in a thermal power plant, the operation efficiency and safety of a boiler are directly related to the stable operation of the entire thermal power plant. However, during actual operation, due to the long-term action of various factors such as high temperature, high pressure, and corrosion, various faults are likely to occur in the boiler tubes, such as leakage, blockage, deformation, etc. These faults not only reduce the thermal efficiency of the boiler, but may also lead to safety accidents in severe cases.

[0003] Existing abnormal monitoring methods for boiler tubes mainly focus on the analysis of temperature data, often relying on threshold judgment or simple statistical analysis, and lacking the comprehensive utilization of multi-source data. For example, when local high temperature of boiler tubes is caused by insufficient combustion in the furnace, if only relying on simple temperature threshold judgment, the abnormal conditions of the boiler tubes cannot be accurately monitored. Summary of the Invention

[0004] In order to accurately monitor the abnormal conditions of boiler tubes, an embodiment of the present invention provides a method and device for abnormal monitoring of boiler tubes in a thermal power plant.

[0005] In a first aspect, an embodiment of the present invention provides a method for abnormal monitoring of boiler tubes in a thermal power plant, including:

[0006] Collecting the wall temperatures of each heat exchange node arranged in sequence according to the medium flow direction in the boiler tubes of the thermal power plant; wherein, the heat exchange node is the node of the boiler tube position before and after the medium in the boiler tube passes through the heat exchange area, and the heat exchange area includes the direct action area of the combustion flame, the heat exchange area between the flue gas and the tube wall, and the hot air preheating area;

[0007] Inputting all the wall temperatures into a pre-trained first abnormal monitoring model, and outputting a first judgment result; wherein, the first judgment result includes the heat exchange nodes with abnormal wall temperatures;

[0008] Collecting multi-source data of all heat exchange nodes with abnormal wall temperatures; wherein, the multi-source data includes the physical and chemical characteristics of the boiler tube, the service life characteristics of the boiler tube, the static fluctuation characteristics of the boiler tube, the surface image characteristics of the boiler tube, and the surface heat distribution characteristics of the boiler tube;

[0009] Inputting the multi-source data into a pre-trained second abnormal monitoring model, and outputting a second judgment result to further monitor the abnormal nodes; wherein, the second judgment result includes the scores of all the heat exchange nodes in the current working state, and the abnormal nodes are the heat exchange nodes with scores lower than a preset score in the second judgment result.

[0010] In a second aspect, the present application also provides an abnormal monitoring device for the boiler tubes of a thermal power plant, including:

[0011] A first acquisition module, configured to acquire the wall temperatures of each heat exchange node arranged in sequence in the boiler tubes of the thermal power plant according to the medium flow direction; wherein, the heat exchange node is the node of the boiler tube position before and after the medium in the boiler tube passes through the heat exchange area, and the heat exchange area includes the direct action area of the combustion flame, the heat exchange area between the flue gas and the tube wall, and the hot air preheating area;

[0012] A first judgment module, configured to input all the wall temperatures into a pre-trained first abnormal monitoring model, and output a first judgment result; wherein, the first judgment result includes the heat exchange nodes with abnormal wall temperatures;

[0013] A second acquisition module, configured to acquire multi-source data of all heat exchange nodes with abnormal wall temperatures; wherein, the multi-source data includes the physical and chemical characteristics of the boiler tubes, the service life characteristics of the boiler tubes, the static fluctuation characteristics of the boiler tubes, the surface image characteristics of the boiler tubes, and the surface heat distribution characteristics of the boiler tubes;

[0014] A second judgment module, configured to input the multi-source data into a pre-trained second abnormal monitoring model, and output a second judgment result to further monitor the abnormal nodes; wherein, the second judgment result includes the scores of all the heat exchange nodes in the current working state, and the abnormal nodes are the heat exchange nodes with scores lower than a preset value in the second judgment result.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: First, the heat exchange nodes with abnormal wall temperatures are initially judged by the wall temperatures of each heat exchange node, and then the multi-source data of all heat exchange nodes with abnormal wall temperatures collected are input into the second abnormal monitoring model, and a second judgment result is output to further monitor the abnormal nodes. In this way, not only can the specific boiler tube area with abnormal heat exchange be accurately located, rather than the entire boiler or a large area in general, but also the working state of the boiler tubes can be comprehensively evaluated, thereby improving the accuracy and reliability of the monitoring. Description of the Drawings

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a flowchart of an abnormal monitoring method for the boiler tubes of a thermal power plant provided by an embodiment of the present invention;

[0018] Figure 2 It is a hardware architecture diagram of an electronic device provided by an embodiment of the present invention;

[0019] Figure 3 It is a structural diagram of an abnormal monitoring device for the boiler tubes of a thermal power plant provided by an embodiment of the present invention. Specific embodiments

[0020] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] Please refer to Figure 1 , an embodiment of the present invention provides an abnormal monitoring method for the boiler tubes of a thermal power plant, and the method includes:

[0022] Step 100: Collect the wall temperatures of each heat exchange node arranged in sequence in the direction of medium flow in the boiler tubes of the thermal power plant; wherein, the heat exchange node is the node of the position of the boiler tube where the medium in the boiler tube is before and after passing through the heat exchange area, and the heat exchange area includes the direct action area of the combustion flame, the heat exchange area between the flue gas and the tube wall, and the hot air preheating area;

[0023] Step 102: Input all the wall temperatures into a pre-trained first abnormal monitoring model, and output a first judgment result; wherein, the first judgment result includes the heat exchange nodes with abnormal wall temperatures;

[0024] Step 104: Collect multi-source data of all heat exchange nodes with abnormal wall temperatures; wherein, the multi-source data includes the physical and chemical characteristics of the boiler tubes, the service life characteristics of the boiler tubes, the static fluctuation characteristics of the boiler tubes, the surface image characteristics of the boiler tubes, and the surface heat distribution characteristics of the boiler tubes;

[0025] Step 106: Input the multi-source data into a pre-trained second abnormal monitoring model, and output a second judgment result to further monitor the abnormal nodes; wherein, the second judgment result includes the scores of all heat exchange nodes in the current working state, and the abnormal nodes are the heat exchange nodes with scores lower than the preset score in the second judgment result.

[0026] In this embodiment, first, the heat exchange nodes with abnormal tube wall temperatures are preliminarily determined through the tube wall temperatures of each heat exchange node. Then, the multi-source data of all the heat exchange nodes with abnormal tube wall temperatures collected are input into the second anomaly monitoring model, and a second judgment result is output to further monitor the abnormal nodes. In this way, the specific furnace tube area with abnormal heat exchange can be accurately located, rather than the entire boiler or a large area in general, and the working state of the furnace tubes can be comprehensively evaluated, thereby improving the accuracy and reliability of the monitoring.

[0027] The execution methods of the following Figure 1 shown steps are described.

[0028] Regarding step 100:

[0029] In some embodiments, step 100 includes the following details:

[0030] Through the design drawings, technical specifications, and operation manuals of the boiler and furnace tubes, the basic information such as the layout and installation position of the furnace tubes is sorted out to ensure the accuracy of subsequent analysis; analyze the combustion process in the furnace, the heat transfer mechanism, and the medium flow characteristics in the furnace tubes; including combustion efficiency, thermal radiation, convective heat transfer, and thermal resistance analysis, etc., to identify which areas are more prone to heat exchange, so as to lock in potential heat exchange nodes;

[0031] A heat exchange node refers to a point or area where significant heat exchange occurs between the inside and outside of the furnace tube, including but not limited to the direct action area of the combustion flame, the heat exchange section between the flue gas and the tube wall, the hot air preheating area, etc.; clearly distinguish the "before heat exchange" nodes (i.e., the heating points before the medium enters the furnace tube) and the "after heat exchange" nodes (the cooling points when the medium leaves the furnace tube);

[0032] According to the flow path of the medium in the furnace tube, all the identified heat exchange nodes are sorted; by considering the actual flow route of the medium, ensure that the order conforms to the physical law and avoid monitoring deviations caused by misranking; at the same time, use professional software or custom algorithms to convert the sorted heat exchange nodes into a digital sequence that is easy to process, namely the heat exchange node sequence of the boiler furnace tubes; the heat exchange node sequence of the boiler furnace tubes contains key information such as the precise coordinates, expected temperature range, and heat load level of each heat exchange node.

[0033] In this step, by accurately extracting and arranging the heat exchange nodes, an accurate data basis is provided for subsequent furnace tube anomaly monitoring, which helps to reduce misjudgments caused by inaccurate or missing data and improve the accuracy of monitoring; sorting the heat exchange nodes according to the flow path of the medium in the furnace tube to ensure that the order conforms to the physical law helps to more accurately reflect the heat exchange process in the furnace tube and improve the reliability of monitoring.

[0034] Collect the wall temperatures of each heat transfer node in the heat transfer node sequence of the boiler furnace tube; for example, install temperature sensors at all heat transfer node positions; the temperature sensors use resistance temperature detectors (RTDs), thermocouples or other high-precision temperature measurement devices, which can withstand the high temperature and high pressure environment during boiler operation and have good stability and response speed; during the operation of the boiler, the temperature sensors continuously monitor the wall temperatures of each heat transfer node; the temperature data is transmitted to the central control system or the data acquisition and monitoring control system by wired or wireless means to achieve remote monitoring and data recording.

[0035] Ensure that the collected temperature data corresponds one-to-one with the heat transfer node sequence and the timestamps are accurate; ensure that the temperature state of each heat transfer node at a specific time point can be accurately reflected during analysis; conduct a preliminary quality check on the collected data to eliminate outliers or incorrect readings, which may be caused by factors such as sensor failures and signal interference; use data filtering and verification algorithms to improve the reliability of the data; organize the verified temperature data in the order of the heat transfer node sequence and store it in the database; it is not only used for real-time monitoring but also serves as the basic data for subsequent analysis, and it is necessary to ensure the security and traceability of the data.

[0036] Through the above technical details, the accuracy and availability of the wall temperature data can be ensured, providing a solid data foundation for subsequent anomaly monitoring, helping to improve the accuracy and timeliness of boiler furnace tube anomaly monitoring, and thus ensuring the safe and efficient operation of the thermal power station.

[0037] Regarding Step 102:

[0038] The goal of Step 102 is a machine learning algorithm to identify heat transfer nodes with abnormal temperatures from all the wall temperatures of the boiler furnace tube; the abnormal manifestations indicate problems such as local overheating, decreased heat exchange efficiency or physical damage inside the furnace tube; by comparing historical data, theoretical model prediction values or set performance criteria.

[0039] Furthermore, the construction process of the first anomaly monitoring model is as follows:

[0040] Collect a large amount of historical data, including the wall temperatures of the boiler furnace tube in normal and abnormal states; it can be obtained through sensor monitoring, on-site measurement or from previous operation and maintenance records; the collected raw data may contain noise, missing values or outliers, and preprocessing is required to improve the data quality; including steps such as data cleaning, data transformation and data augmentation;

[0041] Select a supervised learning model, such as including support vector machines, random forests, gradient boosting trees, recurrent neural networks, long short-term memory networks or convolutional neural networks, etc.;

[0042] Use the labeled historical data to train the model; this includes dividing the data into a training set and a validation set, using the training set to train the model, and using the validation set to evaluate the model's performance; by adjusting the model's parameters and hyperparameters, the model's performance can be optimized, such as increasing the accuracy rate and reducing the false alarm rate, etc.;

[0043] Use an independent test set to evaluate the model's performance; the evaluation metrics include accuracy rate, recall rate, F1 score, ROC curve, and AUC value, etc.; if the model's performance is not ideal, it is possible to return to the previous steps for adjustment, such as modifying feature selection, trying different models, or optimizing parameter settings, etc.;

[0044] When the model training is completed and passes the evaluation, deploy it to an actual system for use; at the same time, a monitoring mechanism needs to be set up to ensure that the model remains stable and accurate during operation; this includes regularly updating the model to adapt to new data distributions and changing environmental conditions, as well as setting up an alarm system to promptly detect and handle abnormal situations of the model.

[0045] Regarding Step 104:

[0046] The specific operations are as follows:

[0047] Collection of physical and chemical characteristics of furnace tubes: The materials used in the manufacturing of each section of furnace tubes have a direct impact on heat resistance and corrosion resistance. Therefore, it is necessary to collect the material reports of furnace tubes, such as material composition, melting point, coefficient of thermal expansion, corrosion resistance, etc., which helps to evaluate the performance of specific materials under the current working conditions and whether they meet the design expectations or are approaching the material limits;

[0048] Analysis of the service life characteristics of furnace tubes: Based on information such as the service time, historical maintenance records, and cumulative working load of furnace tubes, evaluate the degree of aging of furnace tubes; this involves querying the historical database to obtain the detailed maintenance history of each section of furnace tubes and using the service life analysis model to predict the remaining service life;

[0049] Monitoring of static fluctuation characteristics of furnace tubes: Monitor in real time through the tremor sensors installed on the furnace tubes, and use signal processing techniques such as fast Fourier transform to analyze the tremor signals to identify abnormal tremor frequencies and amplitudes;

[0050] Extraction of surface image characteristics of furnace tubes: Regularly take pictures of the furnace tube surface through a high-definition camera, and then use image processing algorithms to analyze the images to identify defects such as cracks, corrosion spots, and deformations that are difficult to detect by the naked eye, which helps to visually and accurately evaluate the appearance condition of the furnace tubes;

[0051] Measurement of surface heat distribution characteristics of furnace tubes: Use an infrared thermal imager to scan the surface of the furnace tubes in the area with abnormal heat exchange to generate a heat distribution map, which is used to reveal areas of local overheating or uneven heat transfer.

[0052] Specifically, for the convenience of being used as the input of the second anomaly monitoring model in step 106, the above-mentioned various features need to be quantified as follows:

[0053] Quantification of the physical and chemical characteristics of the furnace tube: The quantification index of the physical and chemical characteristics of the furnace tube is the temperature resistance index. The calculation of the temperature resistance index needs to combine the melting point, the maximum working temperature and the thermal expansion coefficient of the material. The specific formula is: Temperature resistance index = (Melting point - Maximum working temperature) / Thermal expansion coefficient; Through the above calculation, the temperature resistance performance of the furnace tube under the current working conditions is comprehensively evaluated; The higher the value, the stronger the ability of the material to maintain stability at high temperatures;

[0054] Quantification of the remaining service life characteristics of the furnace tube: The quantification index of the remaining service life characteristics of the furnace tube is the percentage of the remaining service life. Based on the fatigue model of the material, the cumulative working load (such as temperature-time integral, number of pressure cycles) and the historical maintenance records, the theoretical service life of the furnace tube is calculated, and then it is converted into a percentage relative to the design life; For example, if the theoretical calculation shows that the furnace tube can still be safely used for 5 years and the design life is 20 years, then the percentage of the remaining service life is 25%;

[0055] Quantification of the static fluctuation characteristics of the furnace tube: The quantification index of the static fluctuation characteristics of the furnace tube is the vibration intensity index. Through the data collected by the vibration sensor, the vibration amplitude at the main vibration frequency obtained by using the fast Fourier transform analysis, combined with the frequency weight and the square of the velocity or acceleration, is comprehensively calculated to obtain the vibration intensity index VI; The higher the VI value, the more intense the vibration and the more serious the possible problems; The specific calculation formula is as follows:

[0056]

[0057] where A f represents the vibration amplitude at the frequency f, and W f is a weight factor related to the frequency (such as a certain functional relationship between the square of the velocity or acceleration and the frequency);

[0058] Quantification of the surface image characteristics of the furnace tube: The quantification index of the surface image characteristics of the furnace tube is the percentage of the defect area in the surface area of the abnormal furnace tube; The surface image of the furnace tube taken by the high-definition camera is preprocessed, including grayscale conversion, denoising, contrast enhancement, etc., to highlight the defect characteristics such as cracks and corrosion spots; Use advanced image segmentation algorithms to automatically identify and segment the defect areas such as cracks and corrosion spots; Calculate the pixel area of each defect area and accumulate to obtain the total defect area; At the same time, calculate the pixel area of the effective visible area of the entire furnace tube surface; Divide the total defect area by the entire effective visible surface area, multiply by 100%, and get the percentage of the defect area in the surface area of the abnormal furnace tube;

[0059] Quantification of the surface heat distribution characteristics of the furnace tubes: The quantification index of the surface heat distribution characteristics of the furnace tubes is the temperature distribution uniformity; Use an infrared thermal imager to scan the surface of the furnace tubes to obtain high-resolution thermal distribution image data; Convert the thermal distribution image into a temperature matrix, where each pixel point corresponds to a temperature value; Calculate the average temperature, maximum temperature, and minimum temperature of the image; Use the standard deviation or coefficient of variation as the index of temperature distribution uniformity; The standard deviation measures the amplitude of temperature fluctuations, while the coefficient of variation is the ratio of the standard deviation to the average temperature, which can better reflect the degree of fluctuations relative to the average temperature; A lower standard deviation or coefficient of variation indicates a more uniform temperature distribution; Set a threshold or reference range to determine whether the uniformity of the surface heat distribution of the furnace tubes meets the operating requirements; Taking the standard deviation as an example, the calculation formula is as follows:

[0060]

[0061] where T i is the temperature of a single pixel point, is the average temperature, and n is the total number of valid pixels.

[0062] In this step, by collecting multi-source data affected by the tube wall, including physical properties (material characteristics, vibration characteristics), status assessment (service life characteristics), direct observation (surface image characteristics), and thermal performance analysis (surface heat distribution characteristics), a comprehensive assessment of the working state of the furnace tubes is achieved; Conducting detailed quantification processing on each characteristic can not only more accurately identify problems, such as directly evaluating the stability of materials under high-temperature conditions through the temperature resistance index, but also can early warn of potential failure risks through the percentage of remaining service life, ensuring timely maintenance measures are taken; Through the quantitative analysis of vibration intensity, the proportion of surface defect area, and temperature distribution uniformity, potential problems can be discovered and solved before a failure occurs, reducing unplanned downtime, extending the service life of the equipment, and reducing operating costs; Integrating all the quantitative data provides a solid data foundation for the subsequent second anomaly monitoring model; This makes the decision-making based on data analysis more scientific and reasonable, avoiding misjudgments that may be caused by relying solely on experience.

[0063] Regarding step 106:

[0064] Step 106 integrates the data on abnormal furnace tube areas collected from multiple sources and comprehensively evaluates the current working state of these furnace tubes through a pre-trained second anomaly monitoring model; The second anomaly monitoring model uses a weighted calculation method. By assigning a weight to each characteristic affecting the tube wall state, then multiplying these characteristic values by their corresponding weights, and finally summing up to obtain a comprehensive score; It can reflect the different contributions of different characteristics to the health state of the furnace tubes; The specific calculation formula is:

[0065]

[0066] P1 = ω M M + ω S S + ω V V + ω I I + ω H H;

[0067] P2 = ω MS (M S) + ω SV (S V) + ω SI (S I) + ω IH (I H);

[0068] Wherein, E represents the score of each heat exchange node in the current working state, P1 represents the linear combination of basic features, P2 represents the linear combination of interaction terms, M represents the standard quantization value of the physical and chemical characteristics of the furnace tube, S represents the standard quantization value of the service life characteristics of the furnace tube, V represents the standard quantization value of the static fluctuation characteristics of the furnace tube, I represents the standard quantization value of the surface image characteristics of the furnace tube, and H represents the standard quantization value of the surface heat distribution characteristics of the furnace tube;

[0069] ω M 、ω S 、ω V 、ω I and ω H respectively represent the weight coefficients of the physical and chemical characteristics of the furnace tube, the service life characteristics of the furnace tube, the static fluctuation characteristics of the furnace tube, the surface image characteristics of the furnace tube, and the surface heat distribution characteristics of the furnace tube;

[0070] ω MS 、ω SV 、ω SI and ω IH respectively represent the weight coefficients of the interaction terms between the physical and chemical characteristics of the furnace tube and the service life characteristics of the furnace tube, the interaction terms between the service life characteristics of the furnace tube and the static fluctuation characteristics of the furnace tube, the interaction terms between the service life characteristics of the furnace tube and the surface image characteristics of the furnace tube, and the interaction terms between the surface image characteristics of the furnace tube and the surface heat distribution characteristics of the furnace tube.

[0071] Specifically, by considering the interaction between the physical and chemical characteristics of the furnace tube and the service life characteristics of the furnace tube, the process and trend of furnace tube aging can be predicted more accurately; different physical and chemical characteristics of the furnace tube will have different performance manifestations at different service life stages, and the interaction terms can help evaluate the durability of the physical and chemical characteristics of the furnace tube at a specific service life;

[0072] As the service life decreases, the static fluctuation of the furnace tube will intensify accordingly. By considering the interaction between the service life characteristics of the furnace tube and the static fluctuation characteristics of the furnace tube, it can help predict the increase in failure risk and highlight the static fluctuation problem caused by the decrease in service life when evaluating the operating state of the furnace tube;

[0073] As the service life decreases, the surface defects of the furnace tubes will gradually increase or expand. By considering the interaction between the service life characteristics of the furnace tubes and the surface image characteristics of the furnace tubes, the problem of surface defects of the furnace tubes caused by the decrease in service life can be highlighted when evaluating the operating status of the furnace tubes;

[0074] Since the surface defects of the furnace tubes are likely to cause local overheating of the furnace tubes, by considering the interaction between the surface image characteristics of the furnace tubes and the surface heat distribution characteristics of the furnace tubes, the problem of local overheating of the furnace tubes caused by the surface defects of the furnace tubes can be highlighted when evaluating the operating status of the furnace tubes.

[0075] As Figure 2 、 Figure 3 shown, an embodiment of the present invention provides an abnormal monitoring device for furnace tubes of a thermal power plant boiler. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. In terms of the hardware level, as Figure 2 shown, it is a hardware architecture diagram of an electronic device where an abnormal monitoring device for furnace tubes of a thermal power plant boiler provided by an embodiment of the present invention is located. In addition to Figure 2 the processor, memory, network interface, and non-volatile memory shown, the electronic device where the device is located in the embodiment usually may further include other hardware, such as a forwarding chip responsible for processing packets, etc. Taking software implementation as an example, as Figure 3 shown, as a device in a logical sense, it is formed by the CPU of its corresponding electronic device reading the corresponding computer program in the non-volatile memory into the memory and running it.

[0076] As Figure 3 shown, an abnormal monitoring device for furnace tubes of a thermal power plant boiler provided in this embodiment includes:

[0077] A first acquisition module, configured to acquire the wall temperatures of each heat exchange node arranged in sequence according to the medium flow direction in the furnace tubes of the thermal power plant boiler; wherein, the heat exchange node is a furnace tube position node before and after the medium in the furnace tube passes through the heat exchange area, and the heat exchange area includes a combustion flame direct action area, a heat exchange area between flue gas and the tube wall, and a hot air preheating area;

[0078] A first judgment module, configured to input all the wall temperatures into a pre-trained first abnormal monitoring model, and output a first judgment result; wherein, the first judgment result includes the heat exchange nodes with abnormal wall temperatures;

[0079] A second acquisition module, configured to acquire multi-source data of all heat exchange nodes with abnormal wall temperatures; wherein, the multi-source data includes the physical and chemical characteristics of the furnace tubes, the service life characteristics of the furnace tubes, the static fluctuation characteristics of the furnace tubes, the surface image characteristics of the furnace tubes, and the surface heat distribution characteristics of the furnace tubes;

[0080] A second judgment module, configured to input the multi-source data into a pre-trained second anomaly monitoring model, and output a second judgment result to further detect anomaly nodes; wherein, the second judgment result includes the scores of all the heat exchange nodes in the current working state, and the anomaly nodes are the heat exchange nodes with scores lower than a preset score in the second judgment result.

[0081] In some embodiments, the quantification index of the physical and chemical characteristics of the furnace tubes is the temperature resistance index, the quantification index of the service life characteristics of the furnace tubes is the remaining service life percentage, the quantification index of the static fluctuation characteristics of the furnace tubes is the vibration intensity index, the quantification index of the surface image characteristics of the furnace tubes is the percentage of the defect area in the surface area of the abnormal furnace tubes, and the quantification index of the surface heat distribution characteristics of the furnace tubes is the temperature distribution uniformity.

[0082] In some embodiments, the quantification method of the surface heat distribution characteristics of the furnace tubes includes:

[0083] Use an infrared thermal imager to scan the surface of the furnace tubes to obtain thermal distribution image data;

[0084] Convert the thermal distribution image data into a temperature matrix, and each pixel point corresponds to a temperature value;

[0085] Calculate the surface heat distribution uniformity index of the furnace tubes; the calculation formula is:

[0086]

[0087] wherein, SD represents the quantification value of the surface heat distribution characteristics of the furnace tubes, T i is the temperature of a single pixel point, is the average temperature, and n is the total number of valid pixels in the thermal distribution image data.

[0088] In some embodiments, the quantification method of the surface image characteristics of the furnace tubes includes:

[0089] Preprocess the surface image of the furnace tubes captured by a high-definition camera, including grayscale conversion, denoising, and contrast enhancement;

[0090] Use an image segmentation algorithm to automatically identify and segment the defect areas in the surface image of the furnace tubes, and the defect areas include cracks and corrosion spots;

[0091] Calculate the pixel area of each defect area, and accumulate to obtain the total defect area;

[0092] Calculate the surface area of the effective visible area of the abnormal heat exchange furnace tube area;

[0093] Divide the total defect area by the effective visible surface area to obtain the percentage of the defect area in the surface area of the abnormal furnace tubes.

[0094] In some embodiments, the calculation formula of the second anomaly monitoring model is as follows:

[0095]

[0096] P1 = ω M M + ω S S + ω V V + ω I I + ω H H;

[0097] P2 = ω MS (MS) + ω SV (SV) + ω SI (SI) + ω IH (IH);

[0098] Wherein, E represents the score of each heat exchange node under the current working state, P1 represents the linear combination of basic features, P2 represents the linear combination of interaction terms, M represents the standard quantization value of the physical and chemical characteristics of the furnace tubes, S represents the standard quantization value of the service life characteristics of the furnace tubes, V represents the standard quantization value of the static fluctuation characteristics of the furnace tubes, I represents the standard quantization value of the surface image characteristics of the furnace tubes, and H represents the standard quantization value of the surface heat distribution characteristics of the furnace tubes;

[0099] ω M 、ω S 、ω V 、ω I and ω H respectively represent the weight coefficients of the physical and chemical characteristics of the furnace tubes, the service life characteristics of the furnace tubes, the static fluctuation characteristics of the furnace tubes, the surface image characteristics of the furnace tubes, and the surface heat distribution characteristics of the furnace tubes;

[0100] ω MS 、ω SV 、ω SI and ω IH respectively represent the weight coefficients of the interaction terms between the physical and chemical characteristics of the furnace tubes and the service life characteristics of the furnace tubes, the interaction terms between the service life characteristics of the furnace tubes and the static fluctuation characteristics of the furnace tubes, the interaction terms between the service life characteristics of the furnace tubes and the surface image characteristics of the furnace tubes, and the interaction terms between the surface image characteristics of the furnace tubes and the surface heat distribution characteristics of the furnace tubes.

[0101] It can be understood that the structure illustrated in the embodiments of the present invention does not constitute a specific limitation on an abnormal monitoring device for boiler furnace tubes in a thermal power station. In other embodiments of the present invention, an abnormal monitoring device for boiler furnace tubes in a thermal power station may include more or fewer components than those shown in the figures, or combine certain components, or split certain components, or have different component arrangements. The illustrated components can be implemented in hardware, software, or a combination of software and hardware.

[0102] For the information interaction, execution process, etc. among the modules in the above device, since they are based on the same concept as the method embodiments of the present invention, the specific content can be referred to the description in the method embodiments of the present invention and will not be elaborated here.

[0103] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0104] Those of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps including the above method embodiments; and the foregoing storage medium includes: various media such as ROM, RAM, magnetic disk or optical disc that can store program codes.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring abnormality of boiler tubes in a thermal power station, characterized in that: include: Collect the tube wall temperature of each heat exchange node arranged in sequence according to the medium flow direction in the boiler tube of the thermal power station; wherein the heat exchange node is the tube position node of the medium in the tube before and after passing through the heat exchange area, and the heat exchange area includes the direct action area of ​​the combustion flame, the heat exchange area between the flue gas and the tube wall, and the hot air preheating area; Input all pipe wall temperatures into a pre-trained first abnormality monitoring model, and output a first judgment result; wherein the first judgment result includes a heat exchange node with abnormal pipe wall temperature; Collect multi-source data of all heat exchange nodes with abnormal tube wall temperature; wherein the multi-source data includes physical and chemical characteristics of furnace tubes, service life characteristics of furnace tubes, static fluctuation characteristics of furnace tubes, image characteristics of furnace tube surfaces, and thermal distribution characteristics of furnace tube surfaces; Input the multi-source data into a pre-trained second abnormal monitoring model, and output a second judgment result to further monitor abnormal nodes; wherein the second judgment result includes scores of all the heat exchange nodes in the current working state, and the abnormal node is a heat exchange node whose score in the second judgment result is lower than a preset score; The physicochemical characteristics of the furnace tube include material composition, melting point, thermal expansion coefficient, and corrosion resistance. The quantitative index of the physicochemical characteristics of the furnace tube is the temperature resistance index, and the temperature resistance index = (melting point - maximum operating temperature) / thermal expansion coefficient; the service life characteristics of the furnace tube include the service time of the furnace tube, historical maintenance records, and accumulated workload. The quantitative index of the service life characteristics of the furnace tube is the percentage of the remaining service life; the static fluctuation characteristics of the furnace tube include abnormal vibration frequency and amplitude. The quantitative index of the static fluctuation characteristics of the furnace tube is the vibration intensity index. The quantitative index of the surface image characteristics of the furnace tube is the percentage of the defect area to the abnormal furnace tube surface area. The quantitative index of the thermal distribution characteristics of the furnace tube surface is the temperature distribution uniformity. The calculation formula of the second abnormality monitoring model is: P1=ω M M+ω S S+ω V V+ω I I+ω H H; P2=ω MS (MS)+ω SV (SV)+ω SI (SI)+ω IH (IH); Wherein, E represents the score of each heat exchange node in the current working state, P1 represents the linear combination of basic characteristics, P2 represents the linear combination of interaction terms, M represents the standard quantitative value of the physical and chemical characteristics of the furnace tube, S represents the standard quantitative value of the service life characteristics of the furnace tube, V represents the standard quantitative value of the static fluctuation characteristics of the furnace tube, I represents the standard quantitative value of the surface image characteristics of the furnace tube, and H represents the standard quantitative value of the thermal distribution characteristics of the surface of the furnace tube; ω M ,ω S ,ω V ,ω I and ω H They respectively represent the weight coefficients of the physicochemical characteristics of the furnace tube, the service life characteristics of the furnace tube, the static fluctuation characteristics of the furnace tube, the surface image characteristics of the furnace tube, and the thermal distribution characteristics of the furnace tube surface; ω MS ,ω SV ,ω SI and ω IH They respectively represent the weight coefficients of the interaction term between the physicochemical characteristics of the furnace tube and the service life characteristics of the furnace tube, the weight coefficients of the interaction term between the service life characteristics of the furnace tube and the static fluctuation characteristics of the furnace tube, the weight coefficients of the interaction term between the service life characteristics of the furnace tube and the image characteristics of the furnace tube surface, and the weight coefficients of the interaction term between the image characteristics of the furnace tube surface and the thermal distribution characteristics of the furnace tube surface.

2. The method according to claim 1, characterized in that The method for quantifying the heat distribution characteristics of the furnace tube surface includes: Use an infrared thermal imager to scan the furnace tube surface to obtain thermal distribution image data; Convert the thermal distribution image data into a temperature matrix, where each pixel corresponds to a temperature value; Calculate the heat distribution uniformity index of the furnace tube surface; the calculation formula is: Among them, SD represents the quantitative value of the heat distribution characteristics of the furnace tube surface, T i is the temperature of a single pixel, is the average temperature, and n is the total number of valid pixels in the thermal distribution image data.

3. The method according to claim 2, characterized in that The method for quantifying the furnace tube surface image features includes: Preprocess the furnace tube surface image taken by the high-definition camera, including grayscale, denoising and contrast enhancement; Automatically identify and segment defective areas in the furnace tube surface image using an image segmentation algorithm, wherein the defective areas include cracks and corrosion spots; Calculate the pixel area of ​​each defect area and add them up to get the total defect area; Calculate the effective visible area of ​​the surface of the furnace tube area with abnormal heat exchange; Divide the total defect area by the effective visible surface area to obtain the percentage of defect area to the abnormal furnace tube surface area.

4. An abnormality monitoring device for a boiler tube in a thermal power station, characterized in that: include: The first acquisition module is used to collect the tube wall temperature of each heat exchange node arranged in sequence according to the flow direction of the medium in the boiler tube of the thermal power station; wherein the heat exchange node is the tube position node of the medium in the tube before and after passing through the heat exchange area, and the heat exchange area includes the direct action area of ​​the combustion flame, the heat exchange area between the flue gas and the tube wall, and the hot air preheating area; A first judgment module is used to input all pipe wall temperatures into a pre-trained first abnormality monitoring model, and output a first judgment result; wherein the first judgment result includes a heat exchange node with abnormal pipe wall temperature; The second acquisition module is used to collect multi-source data of all heat exchange nodes with abnormal tube wall temperature; wherein the multi-source data includes physical and chemical characteristics of furnace tubes, service life characteristics of furnace tubes, static fluctuation characteristics of furnace tubes, image characteristics of furnace tube surfaces, and thermal distribution characteristics of furnace tube surfaces; A second judgment module is used to input the multi-source data into a pre-trained second abnormal monitoring model, and output a second judgment result to further monitor abnormal nodes; wherein the second judgment result includes scores of all the heat exchange nodes in the current working state, and the abnormal node is a heat exchange node whose score in the second judgment result is lower than a preset score; The physicochemical characteristics of the furnace tube include material composition, melting point, thermal expansion coefficient, and corrosion resistance. The quantitative index of the physicochemical characteristics of the furnace tube is the temperature resistance index, and the temperature resistance index = (melting point - maximum operating temperature) / thermal expansion coefficient; the service life characteristics of the furnace tube include the service time of the furnace tube, historical maintenance records, and accumulated workload. The quantitative index of the service life characteristics of the furnace tube is the percentage of the remaining service life; the static fluctuation characteristics of the furnace tube include abnormal vibration frequency and amplitude. The quantitative index of the static fluctuation characteristics of the furnace tube is the vibration intensity index. The quantitative index of the surface image characteristics of the furnace tube is the percentage of the defect area to the abnormal furnace tube surface area. The quantitative index of the thermal distribution characteristics of the furnace tube surface is the temperature distribution uniformity. The calculation formula of the second abnormality monitoring model is: P1=ω M M+ω S S+ω V V+ω I I+ω H H; P2=ω MS (MS)+ω SV (SV)+ω SI (SI)+ω IH (IH); Wherein, E represents the score of each heat exchange node in the current working state, P1 represents the linear combination of basic characteristics, P2 represents the linear combination of interaction terms, M represents the standard quantitative value of the physical and chemical characteristics of the furnace tube, S represents the standard quantitative value of the service life characteristics of the furnace tube, V represents the standard quantitative value of the static fluctuation characteristics of the furnace tube, I represents the standard quantitative value of the surface image characteristics of the furnace tube, and H represents the standard quantitative value of the thermal distribution characteristics of the surface of the furnace tube; ω M ,ω S ,ω V ,ω I and ω H They respectively represent the weight coefficients of the physicochemical characteristics of the furnace tube, the service life characteristics of the furnace tube, the static fluctuation characteristics of the furnace tube, the surface image characteristics of the furnace tube, and the thermal distribution characteristics of the furnace tube surface; ω MS ,ω SV ,ω SI and ω IH They respectively represent the weight coefficients of the interaction term between the physicochemical characteristics of the furnace tube and the service life characteristics of the furnace tube, the weight coefficients of the interaction term between the service life characteristics of the furnace tube and the static fluctuation characteristics of the furnace tube, the weight coefficients of the interaction term between the service life characteristics of the furnace tube and the image characteristics of the furnace tube surface, and the weight coefficients of the interaction term between the image characteristics of the furnace tube surface and the thermal distribution characteristics of the furnace tube surface.

5. The device according to claim 4, characterized in that The method for quantifying the heat distribution characteristics of the furnace tube surface includes: Use an infrared thermal imager to scan the furnace tube surface to obtain thermal distribution image data; Convert the thermal distribution image data into a temperature matrix, where each pixel corresponds to a temperature value; Calculate the heat distribution uniformity index of the furnace tube surface; the calculation formula is: Among them, SD represents the quantitative value of the heat distribution characteristics of the furnace tube surface, T i is the temperature of a single pixel, is the average temperature, and n is the total number of valid pixels in the thermal distribution image data.

6. The device according to claim 5, characterized in that The method for quantifying the furnace tube surface image features includes: Preprocess the furnace tube surface image taken by the high-definition camera, including grayscale, denoising and contrast enhancement; Automatically identify and segment defective areas in the furnace tube surface image using an image segmentation algorithm, wherein the defective areas include cracks and corrosion spots; Calculate the pixel area of ​​each defect area and add them up to get the total defect area; Calculate the effective visible area of ​​the surface of the furnace tube area with abnormal heat exchange; Divide the total defect area by the effective visible surface area to obtain the percentage of defect area to the abnormal furnace tube surface area.

Images