A coal-fired unit boiler heating surface risk level evaluation method and system

Abstract

This application provides a method and system for assessing the risk level of the heating surface of a coal-fired power plant boiler. The method includes: acquiring the operating parameters input from the fluid-thermal-structure coupled finite element model of the tube bank of the primary heating surface of an ultra-supercritical coal-fired power plant boiler, as well as the simulated physical state parameters of the tube bank under target operating conditions, and processing them into sample data to train a neural network model to establish a temperature and stress prediction model for the dangerous parts of the tube bank under target operating conditions; acquiring the real-time operating parameters of the boiler and inputting them into the temperature and stress prediction model, outputting the real-time temperature and real-time stress set of the dangerous parts; acquiring the material parameters of the dangerous parts corresponding to the real-time temperature, inputting the material parameters, real-time stress set, number of operating cycles, and creep life increment into the creep-fatigue life assessment model constructed based on the continuous damage mechanics model, and outputting the real-time total creep-fatigue damage of the dangerous parts to assess the failure risk level of the heating surface.

Description

Technical Field

[0001] This application relates to the technical field of boilers and their auxiliary equipment, and in particular to a method for assessing the risk level of failure of the heating surface of a coal-fired power unit boiler. Background Technology

[0002] To reduce carbon emissions, the traditional energy structure is shifting towards a new energy structure integrating wind, solar, hydro, thermal, and energy storage. Against this backdrop, coal-fired power units are gradually transforming from primary power sources to auxiliary service power sources. The constraints faced by coal-fired power units are reflected in three aspects: economy, safety, and flexibility. Among them, safety requirements are reflected in deep peak shaving and start-up / shutdown peak shaving.

[0003] Under peak-shaving operation, the boiler tubes of superheaters and reheaters in coal-fired power units are prone to drastic fluctuations and changes in temperature and stress. During operation, they not only bear the static stress brought about by stable operation, but also the fatigue damage caused by temperature fluctuations and steam pressure fluctuations. How to do a good job in metal monitoring and life prediction of hot-end components under deep peak-shaving has become a real problem faced by coal-fired power units, and has also put forward new requirements for the periodic inspection of power plant boilers. Summary of the Invention

[0004] This application provides a method and system for assessing the risk level of the heating surface of a coal-fired power plant boiler. It can accurately calculate the real-time total creep-fatigue damage of dangerous parts and assess the risk level of life failure of the heating surface, thereby improving the accuracy and real-time performance of the assessment.

[0005] The first aspect of this application provides a method for assessing the risk level of the heating surface of a coal-fired power unit boiler, the method comprising: The operating parameters of the fluid-thermal-structure coupled finite element model of the tube bank of the primary heating surface of the ultra-supercritical coal-fired power unit boiler are obtained, as well as the physical state parameters of the tube bank of the boiler under the target operating conditions are simulated and calculated. The obtained parameters are processed into sample data to train the neural network model and establish a temperature and stress prediction model of the dangerous parts of the tube bank of the boiler under the target operating conditions. The model obtains the real-time operating parameters of the boiler, inputs temperature and stress, and outputs the real-time temperature and stress set of the critical parts. Obtain the material parameters of the critical parts corresponding to the real-time temperature, input the material parameters, real-time stress set, and the number of operating cycles and creep life increment of the real-time operating parameters into the creep-fatigue life assessment model constructed based on the continuous damage mechanics model, and output the real-time total creep-fatigue damage of the critical parts. The failure risk level of the heated surface is assessed based on the real-time total creep-fatigue damage at the critical location.

[0006] In some embodiments, the target operating condition refers to the operating condition during cold start-up and stable operation under different evaporation rates.

[0007] In some embodiments, the sample data includes feature data and label data, and the acquired parameters are processed into sample data to train a neural network model, including: The temperature and stress fields in the physical state parameters are analyzed to extract the stress set and temperature of the dangerous parts. The operating parameters are used as feature data, and the stress set and temperature of the dangerous parts are used as label data to train the neural network model.

[0008] In some embodiments, the real-time stress set includes the first, second, and third principal stresses and the von Mises stress, and the expression for the creep-fatigue life assessment model is as follows: , in, This indicates creep-fatigue damage in the dangerous area that has not yet begun to function. This represents the total increase in damage. This indicates the number of times the hazardous area has experienced since it was put into operation. Each cycle consists of a cooling start-up, a stable run, and a shutdown. and These represent the total creep-fatigue damage at the critical site during the current and previous operations, respectively. , As a triaxiality factor, corrected respectively and , , For cold start ( = or stable operation = The hydrostatic pressure at that time , , These are the first, second, and third principal stresses, respectively. Von Mises stress at critical locations during cold start-up or stable operation. Poisson's ratio, This represents the fatigue life increment, which is equivalent to an increment of 1 in the number of starts. This represents the creep life increment, i.e., the stable operating time of critical parts. λ is the energy density coefficient, γ is the fatigue damage evolution index, α1 is the fatigue damage path correction index, α2 is the creep damage path correction index, λ is the characteristic length or time scale parameter of creep damage evolution, and r is the creep damage evolution index. The range of plastic strain at the critical location throughout the entire operating cycle is determined by utilizing the von Mises stress range at the critical location during start-up and shutdown. Calculations show that This represents the hardening coefficient in the elastoplastic constitutive relation of the critical region. The hardening index in the elastoplastic constitutive relation of the critical part; To stabilize the von Mises stress in critical areas during operation; , , γ, α1, α2, λ, r, and These are material parameters.

[0009] In some embodiments, assessing the failure risk level of a heated surface based on real-time total creep-fatigue damage at the critical location includes: The real-time total creep-fatigue damage of the hazardous parts is matched with the preset multi-level early warning thresholds to determine the failure risk level of the hazardous parts in the heated surface.

[0010] In some embodiments, the real-time total creep-fatigue damage of the hazardous area is matched with a preset multi-level early warning threshold to determine the failure risk level of the hazardous area in the heated surface, including: If the real-time total creep-fatigue damage is less than the first warning threshold, then the dangerous part is confirmed to be in a safe state. If the real-time total creep-fatigue damage is greater than or equal to the first warning threshold and less than the second warning threshold, then the dangerous part is confirmed to be in good condition. If the real-time total creep-fatigue damage is greater than or equal to the second warning threshold and less than the third warning threshold, then the dangerous part is confirmed to be in an alarm state. If the real-time total creep-fatigue damage is greater than or equal to the third warning threshold, then the dangerous part is confirmed to be in a dangerous state.

[0011] In some embodiments, the method further includes: Using the number of operating cycles as the x-axis and real-time total creep minus fatigue damage as the y-axis, the failure risk level is marked to plot the life curve of the critical parts of the heated surface.

[0012] In some embodiments, the fluid-thermal-structure coupling finite element model is constructed using the three-dimensional finite element method, and the neural network model is a long short-term memory network model.

[0013] In some embodiments, the different evaporation rates are B-MCR, 75%BMCR, 50%BMCR and 30%BMCR.

[0014] The second aspect of this application provides a risk level assessment system for the heating surface of a coal-fired power unit boiler, the assessment system including a control unit and a coal-fired power unit; The control unit is used to acquire the operating parameters input from the fluid-thermal-structure coupled finite element model of the tube bank of the primary heating surface of the ultra-supercritical coal-fired power unit boiler, as well as the physical state parameters of the tube bank of the boiler under the target operating conditions calculated by simulation. The acquired parameters are processed into sample data to train the neural network model and establish a temperature and stress prediction model for the dangerous parts of the tube bank of the boiler under the target operating conditions. The model obtains the real-time operating parameters of the boiler, inputs temperature and stress, and outputs the real-time temperature and stress set of the critical parts. Obtain the material parameters of the critical parts corresponding to the real-time temperature, input the material parameters, real-time stress set, and the number of operating cycles and creep life increment of the real-time operating parameters into the creep-fatigue life assessment model constructed based on the continuous damage mechanics model, and output the real-time total creep-fatigue damage of the critical parts. The failure risk level of the heated surface is assessed based on the real-time total creep-fatigue damage at the critical location.

[0015] Understandably, the risk level assessment method and system for the heating surface of a coal-fired power unit boiler provided in this application uses the input parameters of the fluid-thermal-structure coupled finite element model of the primary heating surface tube bank of an ultra-supercritical coal-fired power unit boiler and the physical state parameters under the target operating condition as sample data to train a neural network model to establish a temperature and stress prediction model for the dangerous parts of the tube bank. Combined with real-time operating parameters, it outputs a real-time temperature and stress set, links the material parameters corresponding to the real-time temperature with the creep-fatigue life assessment model constructed based on the continuous damage mechanics model, and can accurately calculate the real-time total creep-fatigue damage of the dangerous parts and assess the risk level of the heating surface life failure, thereby improving the accuracy and real-time performance of the assessment. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0017] Figure 1 A schematic diagram of a risk level assessment system for the heating surface of a coal-fired power unit boiler provided in an embodiment of this application; Figure 2 A flowchart illustrating a method for assessing the risk level of a coal-fired power unit boiler heating surface provided in an embodiment of this application; Figure 3 for Figure 2 A schematic diagram of the fluid-thermal-structure coupling finite element model of the tube bank of the primary heating surface of an ultra-supercritical coal-fired power unit boiler provided in the method shown. Figures 4 to 7 They are respectively Figure 3 The diagram shows the temperature field, steam flow field, total displacement, and von Mises stress of the lower left No. 2 furnace tube provided by the fluid-thermal-structure coupled finite element model. Figure 8 The heating surface life curve of the risk level assessment method for the heating surface of a coal-fired power unit boiler provided in the embodiments of this application.

[0018] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0019] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0020] The terms “first”, “second”, etc. used in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated.

[0021] Please see Figure 1 , Figure 1 This is a schematic diagram of a risk level assessment system for the heating surface of a coal-fired power unit boiler. The assessment system 100 includes a control unit 10 and a coal-fired power unit 20. The control unit 10 and the coal-fired power unit 20 cooperate to implement the risk level assessment method for the heating surface of a coal-fired power unit boiler provided in this application.

[0022] The technical solution of this application and how the technical solution of this application solves the technical problem are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0023] Please see Figure 2 , Figure 2 This is a flowchart illustrating a method for assessing the risk level of a coal-fired power unit boiler heating surface provided in this application. The main body executing this assessment method is the control unit, and the assessment method may include the following steps: Step S110: Obtain the operating parameters input to the fluid-thermal-structure coupled finite element model of the tube bank of the primary heating surface of the ultra-supercritical coal-fired power unit boiler, as well as the physical state parameters of the tube bank of the boiler under the target operating conditions calculated by simulation. Process the obtained parameters into sample data to train the neural network model and establish a temperature and stress prediction model for the dangerous parts of the tube bank of the boiler under the target operating conditions.

[0024] Specifically, ultra-supercritical coal-fired power units refer to coal-fired power units that employ ultra-supercritical parameters. A coal-fired power unit includes a boiler, which is equipped with a superheater and a reheater. The primary heating surface refers to the high-temperature section of the superheater or reheater. The heating surface is equipped with tube bundles, which are bundles of multiple furnace tubes. The parts of the furnace tubes that are susceptible to creep-fatigue damage due to temperature and stress are called hazardous areas.

[0025] The target operating condition refers to the operating conditions during cold start-up and stable operation under different evaporation rates. The different evaporation rates can be B-MCR, 75%BMCR, 50%BMCR and 30%BMCR.

[0026] A three-dimensional finite element method can be used to construct a fluid-thermal-structure coupled finite element model of the tube bank of the primary heating surface of an ultra-supercritical coal-fired power unit boiler. By inputting the boiler's medium flow rate, load, flue gas composition, working fluid flow rate in the boiler tubes, number of operating cycles, creep life increment, temperature, and pressure into this fluid-thermal-structure coupled finite element model, the physical state parameters of the tube bank, including temperature field, flow field in the tubes, total deformation, and stress set, can be calculated under B-MCR, 75%BMCR, 50%BMCR, and 30%BMCR, during cold start-up and stable operation of the boiler. Among them, the stress set includes the first, second, and third principal stresses and the von Mises equivalent stress.

[0027] For example, such as Figure 3 As shown, Figure 3 To construct a fluid-thermal-structure coupled finite element model of the tube bank of the primary heating surface of an ultra-supercritical coal-fired power unit boiler, this example uses T / P 92 (i.e., 9-12% Cr martensitic steel) superheater tube bank, with a cooling start-up every 10 days. The physical state parameters of the tube bank under cold start-up and stable operation at B-MCR, 75% BMCR, 50% BMCR, and 30% BMCR loads are calculated using this fluid-thermal-structure coupled finite element model. The following parameters can be analyzed: Figure 4 , 5 The temperature field, steam flow field, total displacement, and von Mises stress of the lower left furnace tube shown in Figures 6 and 7 are as follows:

[0028] One way to process the acquired parameters (operational parameters and physical state parameters) into sample data (including label data and feature data) is as follows: analyze the temperature field and stress field in the physical state parameters, extract the stress set and temperature of the dangerous parts, use the operation parameters as feature data, and use the stress set and temperature of the dangerous parts as label data to train a neural network model. The neural network model can be a Long Short-Term Memory Network (LSTM) model.

[0029] Step S120: Obtain the real-time operating parameters of the boiler, input the temperature and stress prediction model, and output the real-time temperature and stress set of the dangerous parts.

[0030] Specifically, taking the superheater as an example, a 32-point monitoring network is formed by arranging K-type thermocouples (accuracy ±1.5℃), piezoelectric pressure sensors (range 0-20MPa), and fiber optic strain sensors (resolution 1με) in the middle section of the superheater's inlet header, outlet header, and pipe arrangement in a screen-type structure, as well as at typical bends. This network achieves second-level sampling and minute-level averaging through a data acquisition system. Simultaneously, it acquires data such as flow rate, load, and flue gas composition collected by the DCS system. After data preprocessing, it obtains real-time operating parameters such as the working fluid flow rate, temperature, number of operating cycles, creep life increment, pressure, and combustion status in the furnace of the boiler.

[0031] Step S130: Obtain the material parameters of the critical parts corresponding to the real-time temperature, input the material parameters, real-time stress set, and the number of running cycles and creep life increment of the real-time operating parameters into the creep-fatigue life assessment model constructed based on the continuous damage mechanics model, and output the real-time total creep-fatigue damage of the critical parts.

[0032] Specifically, by conducting creep-fatigue tests on the steel used in the heated surface, the material parameters corresponding to different temperatures in the critical parts of the heated surface are obtained.

[0033] The expression for the creep-fatigue life assessment model is shown below: , in, This indicates creep-fatigue damage in the dangerous area that has not yet begun to function. This represents the total increase in damage. This indicates the number of times the hazardous area has experienced since it was put into operation. Each cycle consists of a cooling start-up, a stable run, and a shutdown. and These represent the total creep-fatigue damage at the critical site during the current and previous operations, respectively. , As a triaxiality factor, corrected respectively and , , For cold start ( = or stable operation = The hydrostatic pressure at that time , , These are the first, second, and third principal stresses, respectively. Von Mises stress at critical locations during cold start-up or stable operation. Poisson's ratio, This represents the fatigue life increment, which is equivalent to an increment of 1 in the number of starts. This represents the creep life increment, i.e., the stable operating time of critical parts. λ is the energy density coefficient, γ is the fatigue damage evolution index, α1 is the fatigue damage path correction index, α2 is the creep damage path correction index, λ is the characteristic length or time scale parameter of creep damage evolution, and r is the creep damage evolution index. The range of plastic strain at the critical location throughout the entire operating cycle is determined by utilizing the von Mises stress range at the critical location during start-up and shutdown. Calculations show that This represents the hardening coefficient in the elastoplastic constitutive relation of the critical region. The hardening index in the elastoplastic constitutive relation of the critical part; To stabilize the von Mises stress in critical areas during operation; , , γ, α1, α2, λ, r, and These are material parameters.

[0034] Following the previous example using T / P 92 superheater tube bank, at a real-time temperature of 600℃, what are the corresponding material parameters for the critical parts of the tube bank? α, γ, and α1 take values ​​of 14, 0.6, and 2 respectively, while α2, r, and λ take values ​​of 2, 8, and 585 respectively. , 1100 , The value is 0.12. Therefore, the expression for the creep-fatigue life assessment model is as follows: , Because of errors in the finite element model simulation and stress concentration issues caused by the structure, when assessing the risk level of life failure, a stress correction factor is introduced to adjust the maximum von Mises stress output by the finite element model. The von Mises stress is obtained from the input creep-fatigue life assessment model. Then the triaxiality factor is calculated. , and plastic strain range The specific values ​​are shown in Table 1 below. Next, the values ​​in the table... , , and Substituting into the creep-fatigue life assessment model, the total creep-fatigue damage is calculated. .

[0035] Table 1: Stress-related parameters required for creep-fatigue life assessment model

[0036]

[0037] Step S140: Assess the failure risk level of the heated surface based on the real-time total creep-fatigue damage of the hazardous area.

[0038] Understandably, in the above technical solution, by using the input parameters of the fluid-thermal-structure coupled finite element model of the primary heating surface tube bank of an ultra-supercritical coal-fired power unit boiler and the physical state parameters under the target operating condition as sample data, a neural network model is trained to establish a temperature and stress prediction model for the dangerous parts of the tube bank. Combined with real-time operating parameters, the real-time temperature and stress set is output, linked to the material parameters corresponding to the real-time temperature and the creep-fatigue life assessment model constructed based on the continuous damage mechanics model. This allows for the accurate calculation of the real-time total creep-fatigue damage of the dangerous parts and the assessment of the life failure risk level of the heating surface, improving the accuracy and real-time performance of the assessment. In other words, based on the fluid-thermal-structure coupled simulation calculation and neural network model of the coal-fired power unit's heating surface, combined with real-time operating parameters such as the flow rate, temperature, and pressure of the working fluid inside the boiler tubes collected from on-site measuring points, as well as the combustion status in the furnace, the total creep-fatigue damage of the dangerous parts of the tube bank can be analyzed in real time, and based on this, the life failure risk level of the heating surface can be given, improving the accuracy and real-time performance of the assessment.

[0039] In some embodiments, step S140: assessing the failure risk level of the heated surface based on the real-time total creep-fatigue damage of the hazardous parts includes: matching the real-time total creep-fatigue damage of the hazardous parts with preset multi-level early warning thresholds to determine the failure risk level of the hazardous parts in the heated surface.

[0040] In one implementation, the failure risk level is divided into safe state, good state, alarm state, and dangerous state. That is, if the real-time total creep-fatigue damage is less than a first warning threshold, the dangerous part is confirmed to be in a safe state; if the real-time total creep-fatigue damage is greater than or equal to the first warning threshold and less than a second warning threshold, the dangerous part is confirmed to be in a good state; if the real-time total creep-fatigue damage is greater than or equal to the second warning threshold and less than a third warning threshold, the dangerous part is confirmed to be in an alarm state; if the real-time total creep-fatigue damage is greater than or equal to the third warning threshold, the dangerous part is confirmed to be in a dangerous state.

[0041] Following the previous example using T / P 92 superheater tube banks, the four-level warning thresholds are as follows: , In some embodiments, the evaluation method further includes: plotting the number of operating cycles as the x-axis and the real-time total creep-fatigue damage as the y-axis, and labeling the failure risk level to draw a heat-sustaining surface life curve for the critical parts of the heat-sustaining surface, as shown in the heat-sustaining surface life curve. Figure 8 As shown.

[0042] Understandably, the risk level of dangerous parts of the heated surface is visualized through the life curve of the heated surface.

[0043] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for assessing the risk level of a coal-fired power unit boiler heating surface, characterized in that the method... include: The operating parameters of the fluid-thermal-structure coupled finite element model of the tube bank of the primary heating surface of the ultra-supercritical coal-fired power unit boiler are obtained, as well as the physical state parameters of the tube bank of the boiler under the target operating conditions are simulated and calculated. The obtained parameters are processed into sample data to train the neural network model and establish a temperature and stress prediction model of the dangerous parts of the tube bank of the boiler under the target operating conditions. The real-time operating parameters of the boiler are obtained and input into the temperature and stress prediction model, which outputs the real-time temperature and stress set of the critical parts. Obtain the material parameters of the critical parts corresponding to the real-time temperature, input the material parameters, real-time stress set, and the number of operating cycles and creep life increment of the real-time operating parameters into the creep-fatigue life assessment model constructed based on the continuous damage mechanics model, and output the real-time total creep-fatigue damage of the critical parts. The failure risk level of the heated surface is assessed based on the real-time total creep-fatigue damage at the critical location.

2. The method according to claim 1, characterized in that, The target operating conditions refer to the operating conditions during cold start-up and stable operation under different evaporation rates.

3. The method according to claim 1, characterized in that, The sample data includes feature data and label data. The process of processing the acquired parameters into sample data to train the neural network model includes: The temperature and stress fields in the physical state parameters are analyzed to extract the stress set and temperature of the dangerous parts. The operating parameters are used as feature data, and the stress set and temperature of the dangerous parts are used as label data to train the neural network model.

4. The method according to claim 1, characterized in that, The real-time stress set includes the first, second, and third principal stresses and the von Mises stress, and the expression of the creep-fatigue life assessment model is as follows: ， in, This indicates creep-fatigue damage in the dangerous area that has not yet begun to function. This represents the total increase in damage. This indicates the number of times the hazardous area has experienced since it was put into operation. Each cycle consists of a cooling start-up, a stable run, and a shutdown. and These represent the total creep-fatigue damage at the critical site during the current and previous operations, respectively. , As a triaxiality factor, corrected respectively and , , For cold start ( = or stable operation = The hydrostatic pressure at that time , , These are the first, second, and third principal stresses, respectively. Von Mises stress at critical locations during cold start-up or stable operation. Poisson's ratio, This represents the fatigue life increment, which is equivalent to an increment of 1 in the number of starts. This represents the creep life increment, i.e., the stable operating time of critical parts. λ is the energy density coefficient, γ is the fatigue damage evolution index, α1 is the fatigue damage path correction index, α2 is the creep damage path correction index, λ is the characteristic length or time scale parameter of creep damage evolution, and r is the creep damage evolution index. The range of plastic strain at the critical location throughout the entire operating cycle is determined by utilizing the von Mises stress range at the critical location during start-up and shutdown. Calculations show that This represents the hardening coefficient in the elastoplastic constitutive relation of the critical region. The hardening index in the elastoplastic constitutive relation of the critical part; To stabilize the von Mises stress in critical areas during operation; , , γ, α1, α2, λ, r, and These are material parameters.

5. The method according to claim 1, characterized in that, The assessment of the failure risk level of the heated surface based on the real-time total creep-fatigue damage of the hazardous area includes: The real-time total creep-fatigue damage of the hazardous parts is matched with the preset multi-level early warning thresholds to determine the failure risk level of the hazardous parts in the heated surface.

6. The method according to claim 5, characterized in that, The method of matching the real-time total creep-fatigue damage of the hazardous area with preset multi-level early warning thresholds to determine the failure risk level of the hazardous area in the heated surface includes: If the real-time total creep-fatigue damage is less than the first warning threshold, then the dangerous part is confirmed to be in a safe state. If the real-time total creep-fatigue damage is greater than or equal to the first warning threshold and less than the second warning threshold, then the dangerous part is confirmed to be in good condition. If the real-time total creep-fatigue damage is greater than or equal to the second warning threshold and less than the third warning threshold, then the dangerous part is confirmed to be in an alarm state. If the real-time total creep-fatigue damage is greater than or equal to the third warning threshold, then the dangerous part is confirmed to be in a dangerous state.

7. The method according to claim 6, characterized in that, The method further includes: Using the number of operating cycles as the x-axis and real-time total creep minus fatigue damage as the y-axis, the failure risk level is marked to plot the life curve of the critical parts of the heated surface.

8. The method according to claim 1, characterized in that, The fluid-thermal-structure coupled finite element model is constructed using the three-dimensional finite element method, and the neural network model is a long short-term memory network model.

9. The method according to claim 2, characterized in that, The different evaporation rates are B-MCR, 75%BMCR, 50%BMCR and 30%BMCR.

10. A risk level assessment system for the heating surface of a coal-fired power unit boiler, characterized in that, The evaluation system includes a control unit and a coal-fired power unit; The control unit is used to acquire the operating parameters input from the fluid-thermal-structure coupled finite element model of the tube bank of the primary heating surface of the ultra-supercritical coal-fired power unit boiler, as well as the physical state parameters of the tube bank of the boiler under the target operating conditions calculated by simulation, and to process the acquired parameters into sample data to train the neural network model and establish a temperature and stress prediction model for the dangerous parts of the tube bank of the boiler under the target operating conditions. The real-time operating parameters of the boiler are obtained and input into the temperature and stress prediction model, which outputs the real-time temperature and stress set of the critical parts. Obtain the material parameters of the critical parts corresponding to the real-time temperature, input the material parameters, real-time stress set, and the number of operating cycles and creep life increment of the real-time operating parameters into the creep-fatigue life assessment model constructed based on the continuous damage mechanics model, and output the real-time total creep-fatigue damage of the critical parts. The failure risk level of the heated surface is assessed based on the real-time total creep-fatigue damage at the critical location.

Images