A soft measurement method and system for water-cooled wall temperature of boiler in thermal power generator set

Abstract

The present invention discloses a soft measurement method and system for the water-cooled wall temperature of a boiler in a thermal power generating set, belonging to the technical field of boiler safety. The method and system comprise the following steps: establishing a mathematical mechanism model of the boiler water-cooled wall temperature, and performing real-time calculation of the boiler water-cooled wall temperature based on the mathematical mechanism model to obtain a calculated value; performing real-time prediction of the boiler water-cooled wall temperature based on a machine learning algorithm according to measured data to obtain a predicted value; and constructing a comprehensive prediction function based on the calculated value and the predicted value, wherein the calculated value is updated in real time, and when the time is greater than a first threshold or when the variation of the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water-cooled wall temperature at this time. The method solves the problems of inaccurate and slow measurement of the original boiler water-cooled wall temperature, and makes the monitoring of the boiler water-cooled wall temperature more reasonable, stable and reliable.

Description

Technical Field

[0001] The invention belongs to the technical field of boiler safety, and in particular relates to a soft measurement method and system for the water-cooled wall temperature of a boiler of a thermal power generator set. Background Art

[0002] As the main steam pressure and temperature of the boiler continue to increase, the vaporization process of the working fluid becomes less obvious, and the role of the steam drum becomes increasingly smaller. Therefore, the steam drum is eliminated. Without the intermediate buffering function of the steam drum, the working fluid passes through the water-cooled wall, superheater, and reheater heating surfaces in one go. The fixed boundary of water vaporization during evaporation disappears, and the heat storage capacity of the entire evaporation process is relatively small. The wall temperature of the evaporation heating surface and the temperature of the working fluid decrease with the delay of feedwater and coal combustion. If the water-coal ratio is out of balance, the working fluid environment within the evaporation heating surface will rapidly deteriorate, and the temperature of the working fluid at the boiler outlet will rise rapidly. Excessively high steam temperature affects the wall temperature of the water-cooled wall, superheater, and reheater, and even the safety of the entire furnace, causing "four-pipe leakage" accidents. Therefore, for thermal power units, it is particularly important to accurately measure the temperature of each point in the furnace or predict the temperature changes in the furnace through modeling to prevent the occurrence of such accidents.

[0003] Research and prevention of water-cooled wall temperature of generator sets at home and abroad, among which the research on water-cooled wall temperature measurement is mainly divided into the following aspects: First, the temperature of the water-cooled wall is calculated by actual measurement technology, including the measurement of furnace heat and tube wall temperature. There is a large measurement error and it cannot accurately reflect the true temperature of the entire water-cooled wall; second, through numerical simulation or based on measurement data, the changes in the load and working fluid hydraulics in the furnace are calculated to build a furnace numerical model. The error is large and the algorithm accuracy is low; third, the boiler wall temperature curve is fitted by artificial intelligence, which is slow and cannot meet the needs of real-time monitoring of the boiler water-cooled wall temperature.

[0004] Based on the above technical problems, it is necessary to design a soft measurement method and system for the water-cooled wall temperature of the boiler of a thermal power generator set. Summary of the Invention

[0005] The purpose of the present invention is to provide a soft measurement method and system for the water-cooled wall temperature of a boiler of a thermal power generating unit.

[0006] In order to solve the above technical problems, the first aspect of the present invention provides a soft measurement method for the water-cooled wall temperature of a boiler of a thermal power generator unit, comprising:

[0007] Step S1: establishing a mathematical mechanism model of the boiler water wall temperature, and performing real-time calculation of the boiler water wall temperature based on the mathematical mechanism model to obtain a calculated value;

[0008] Step S2: Based on the measured data, the boiler water wall temperature is predicted in real time using a machine learning algorithm to obtain a predicted value;

[0009] Step S3: construct a comprehensive prediction function based on the calculated value and the predicted value, wherein the calculated value is updated in real time. When the time is greater than a first threshold or when the change in the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water-cooled wall temperature at this time.

[0010] By establishing a mathematical mechanism model of the boiler water-cooled wall temperature, the boiler water-cooled wall temperature is calculated in real time based on the above mathematical mechanism model to obtain a calculated value. According to the measured data, the boiler water-cooled wall temperature is predicted in real time based on a machine learning algorithm to obtain a predicted value. After obtaining the calculated value and the predicted value, by controlling the update speed of the calculated value and the predicted value, it is possible to achieve accurate prediction of the boiler water-cooled wall temperature while greatly accelerating the prediction speed at this time, solving the original problems of poor accuracy and slow speed.

[0011] By establishing a mathematical mechanism model of the boiler water-cooled wall temperature, the boiler water-cooled wall temperature is calculated in real time based on the above mathematical mechanism model to obtain a calculated value, which can realize the rapid calculation of the boiler water-cooled wall temperature. At the same time, the boiler water-cooled wall temperature is predicted by a machine learning algorithm, which can realize the accurate prediction of the boiler water-cooled wall temperature. A comprehensive prediction function is constructed based on the calculated value and the predicted value, wherein the calculated value is updated in real time. When the time is greater than a first threshold or when the change in the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water-cooled wall temperature at this time. Since the speed of predicting the boiler water-cooled wall temperature by the machine learning algorithm is slow, by timing or over-updating the predicted value in the comprehensive prediction function, the boiler water-cooled wall temperature obtained by the comprehensive prediction function at this time not only has good accuracy, but also its dynamic speed is greatly improved, so that the algorithm at this time can better meet the prediction of the boiler water-cooled wall temperature and become more accurate.

[0012] A further technical solution is that the specific steps of performing real-time calculation of the boiler water wall temperature based on the mathematical mechanism model to obtain the calculated value are:

[0013] S1 Based on the structural parameters of the furnace, the principal component analysis method is used to perform principal component analysis on the structural parameters and extract the characteristic quantities that affect the boiler water-cooled wall temperature;

[0014] S2 builds a mathematical mechanism model of the boiler water wall temperature based on the characteristic quantity;

[0015] S3 sends the real-time measurement point temperature into the mathematical mechanism model of the boiler water-cooled wall temperature to calculate the calculated value of the boiler water-cooled wall temperature at this time.

[0016] By adopting the principal component analysis method, the number of characteristic quantities for building the mathematical mechanism model of the boiler water-cooled wall temperature is reduced, and the calculation speed of the boiler water-cooled wall temperature can be further accelerated when the prediction accuracy of the boiler water-cooled wall temperature does not decrease much or does not decrease.

[0017] A further technical solution is that the mathematical mechanism model of the boiler water-cooled wall temperature performs segmented calculations on the flue gas temperature, combustion temperature, working fluid pressure, enthalpy value, and working fluid temperature.

[0018] A further technical solution is that before real-time prediction of the boiler water-cooled wall temperature is performed based on the measured data and the predicted value is obtained based on the machine learning algorithm, the measured data needs to be subjected to principal component analysis based on the PCA principal component analysis method to extract the characteristic values.

[0019] By extracting eigenvalues, the dimension of the measured data at this time is reduced, thereby further improving the prediction speed based on the machine learning algorithm.

[0020] A further technical solution is to perform denoising processing on the eigenvalues ​​based on the Canopy clustering algorithm.

[0021] The eigenvalue denoising process is performed based on the Canopy clustering algorithm, which avoids the disadvantage of the traditional clustering denoising algorithm that it is weak in noise interference resistance. It has the advantages of accurate clustering points and reduced number of clustering calculations, making the overall processing speed more accurate and the efficiency further improved.

[0022] A further technical solution is that the machine learning algorithm adopts a machine learning algorithm based on Attention-GRU.

[0023] By adopting the attention mechanism to further optimize the GRU algorithm, compared with the traditional LSTM algorithm, the forget gate and input gate are combined into an update gate, and the cell state information flow and the hidden layer state information flow are merged into one information flow. Its structure is simpler than the standard LSTM, and the convergence speed is further improved, thereby further improving the sensitivity of the prediction algorithm.

[0024] A further technical solution is that the prediction steps of the Attention-GRU based machine learning algorithm are:

[0025] S1 extracts the actual operating data and further reduces the dimension of the input data through a dimensionality reduction algorithm to obtain the input vector at this time, including multiple groups of input quantities;

[0026] S2 normalizes the data based on the input vector, and assigns different weights to different input quantities in the normalized input vector according to their characteristics based on Attention, to obtain the corrected input vector;

[0027] S3 inputs the corrected input vector into the GRU-based prediction algorithm to obtain the predicted value of the boiler water-cooled wall temperature at this time.

[0028] A further technical solution is that the first threshold and the second threshold are determined according to the capacity and type of the boiler.

[0029] A further technical solution is to further include a third threshold value. When the variation of the measured data is greater than the third threshold value, the proportion of the calculated value in the comprehensive prediction function is increased, and the third threshold value is greater than the second threshold value.

[0030] When the boiler is in the startup and shutdown states, the predicted value is used to predict the boiler water-cooled wall temperature. Since the boiler state is not in normal operation, its prediction accuracy is low. Therefore, the proportion of calculated values ​​is increased to further improve the prediction accuracy without increasing the amount of calculation.

[0031] Another aspect of the present invention provides a soft measurement system for the water-cooled wall temperature of a boiler of a thermal power generator set, which adopts the above-mentioned soft measurement method for the water-cooled wall temperature of a boiler of a thermal power generator set, comprising:

[0032] Mathematical mechanism model calculation module, machine learning algorithm prediction module, result output module, wherein the mathematical mechanism model calculation module is responsible for establishing a mathematical mechanism model of the boiler water-cooled wall temperature, and performing real-time calculation of the boiler water-cooled wall temperature based on the above mathematical mechanism model to obtain a calculated value; the machine learning algorithm prediction module is responsible for performing real-time prediction of the boiler water-cooled wall temperature based on the machine learning algorithm according to the measured data to obtain a predicted value; the result output module is responsible for constructing a comprehensive prediction function based on the calculated value and the predicted value, wherein the calculated value is updated in real time, and when the time is greater than a first threshold or when the change in the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water-cooled wall temperature at this time.

[0033] Other features and advantages will be described in the following description, and in part will become apparent from the description, or understood by practicing the invention. The purpose and other advantages of the invention are realized and obtained by the structures particularly pointed out in the description and the drawings.

[0034] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 This is a flow chart of a soft measurement method for the water wall temperature of a thermal power generator boiler in Example 1;

[0037] Figure 2 This is a flow chart for real-time calculation of boiler water wall temperature based on the mathematical mechanism model in Example 1;

[0038] Figure 3 This is a flowchart of the prediction steps of the Attention-GRU-based machine learning algorithm in Example 1;

[0039] Figure 4 This is a diagram showing the structure of a soft measurement system for the water-cooled wall temperature of a boiler in a thermal power generator set in Example 2. DETAILED DESCRIPTION

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0041] As the main steam pressure and temperature of the boiler continue to increase, the vaporization process of the working fluid becomes less obvious, and the role of the steam drum becomes increasingly smaller. Therefore, the steam drum is eliminated. Without the intermediate buffering function of the steam drum, the working fluid passes through the water-cooled wall, superheater, and reheater heating surfaces in one go. The fixed boundary of water vaporization during evaporation disappears, and the heat storage capacity of the entire evaporation process is relatively small. The wall temperature of the evaporation heating surface and the temperature of the working fluid decrease with the delay of feedwater and coal combustion. If the water-coal ratio is out of balance, the working fluid environment within the evaporation heating surface will rapidly deteriorate, and the temperature of the working fluid at the boiler outlet will rise rapidly. Excessively high steam temperature affects the wall temperature of the water-cooled wall, superheater, and reheater, and even the safety of the entire furnace, causing "four-pipe leakage" accidents. Therefore, for thermal power units, it is particularly important to accurately measure the temperature of each point in the furnace or predict the temperature changes in the furnace through modeling to prevent the occurrence of such accidents.

[0042] Research and prevention of water-cooled wall temperature of generator sets at home and abroad, among which the research on water-cooled wall temperature measurement is mainly divided into the following aspects: First, the temperature of the water-cooled wall is calculated by actual measurement technology, including the measurement of furnace heat and tube wall temperature. There is a large measurement error and it cannot accurately reflect the true temperature of the entire water-cooled wall; second, through numerical simulation or based on measurement data, the changes in the load and working fluid hydraulics in the furnace are calculated to build a furnace numerical model. The error is large and the algorithm accuracy is low; third, the boiler wall temperature curve is fitted by artificial intelligence, which is slow and cannot meet the needs of real-time monitoring of the boiler water-cooled wall temperature.

[0043] Example 1

[0044] Figure 1 The present invention relates to a soft measurement method for the water-cooled wall temperature of a boiler of a thermal power generating unit, comprising:

[0045] Step S1: establishing a mathematical mechanism model of the boiler water wall temperature, and performing real-time calculation of the boiler water wall temperature based on the mathematical mechanism model to obtain a calculated value;

[0046] Step S2: Based on the measured data, the boiler water wall temperature is predicted in real time using a machine learning algorithm to obtain a predicted value;

[0047] Step S3: construct a comprehensive prediction function based on the calculated value and the predicted value, wherein the calculated value is updated in real time. When the time is greater than a first threshold or when the change in the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water-cooled wall temperature at this time.

[0048] By establishing a mathematical mechanism model of the boiler water-cooled wall temperature, the boiler water-cooled wall temperature is calculated in real time based on the above mathematical mechanism model to obtain a calculated value. According to the measured data, the boiler water-cooled wall temperature is predicted in real time based on a machine learning algorithm to obtain a predicted value. After obtaining the calculated value and the predicted value, by controlling the update speed of the calculated value and the predicted value, it is possible to achieve accurate prediction of the boiler water-cooled wall temperature while greatly accelerating the prediction speed at this time, solving the original problems of poor accuracy and slow speed.

[0049] By establishing a mathematical mechanism model of the boiler water-cooled wall temperature, the boiler water-cooled wall temperature is calculated in real time based on the above mathematical mechanism model to obtain a calculated value, which can realize the rapid calculation of the boiler water-cooled wall temperature. At the same time, the boiler water-cooled wall temperature is predicted by a machine learning algorithm, which can realize the accurate prediction of the boiler water-cooled wall temperature. A comprehensive prediction function is constructed based on the calculated value and the predicted value, wherein the calculated value is updated in real time. When the time is greater than a first threshold or when the change in the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water-cooled wall temperature at this time. Since the speed of predicting the boiler water-cooled wall temperature by the machine learning algorithm is slow, by timing or over-updating the predicted value in the comprehensive prediction function, the boiler water-cooled wall temperature obtained by the comprehensive prediction function at this time not only has good accuracy, but also its dynamic speed is greatly improved, so that the algorithm at this time can better meet the prediction of the boiler water-cooled wall temperature and become more accurate.

[0050] In another possible embodiment, Figure 2 As shown, the specific steps of performing real-time calculation of the boiler water wall temperature based on the mathematical mechanism model to obtain the calculated value are:

[0051] S1 Based on the structural parameters of the furnace, the principal component analysis method is used to perform principal component analysis on the structural parameters and extract the characteristic quantities that affect the boiler water-cooled wall temperature;

[0052] S2 builds a mathematical mechanism model of the boiler water wall temperature based on the characteristic quantity;

[0053] S3 sends the real-time measurement point temperature into the mathematical mechanism model of the boiler water-cooled wall temperature to calculate the calculated value of the boiler water-cooled wall temperature at this time.

[0054] By adopting the principal component analysis method, the number of characteristic quantities for building the mathematical mechanism model of the boiler water-cooled wall temperature is reduced, and the calculation speed of the boiler water-cooled wall temperature can be further accelerated when the prediction accuracy of the boiler water-cooled wall temperature does not decrease much or does not decrease.

[0055] In another possible embodiment, the mathematical mechanism model of the boiler water-cooled wall temperature performs segmented calculations on the flue gas temperature, combustion temperature, working fluid pressure, enthalpy value, and working fluid temperature.

[0056] In another possible embodiment, before performing real-time prediction of the boiler water-cooled wall temperature based on the measured data based on a machine learning algorithm to obtain a predicted value, it is necessary to perform principal component analysis based on the PCA principal component analysis method on the measured data to extract characteristic values.

[0057] By extracting eigenvalues, the dimension of the measured data at this time is reduced, thereby further improving the prediction speed based on the machine learning algorithm.

[0058] In another possible embodiment, the eigenvalues ​​are subjected to denoising processing based on a Canopy clustering algorithm.

[0059] The eigenvalue denoising process is performed based on the Canopy clustering algorithm, which avoids the disadvantage of the traditional clustering denoising algorithm that it is weak in noise interference resistance. It has the advantages of accurate clustering points and reduced number of clustering calculations, making the overall processing speed more accurate and the efficiency further improved.

[0060] In another possible embodiment, the machine learning algorithm adopts an Attention-GRU-based machine learning algorithm.

[0061] By adopting the attention mechanism to further optimize the GRU algorithm, compared with the traditional LSTM algorithm, the forget gate and input gate are combined into an update gate, and the cell state information flow and the hidden layer state information flow are merged into one information flow. Its structure is simpler than the standard LSTM, and the convergence speed is further improved, thereby further improving the sensitivity of the prediction algorithm.

[0062] In another possible embodiment, Figure 3 As shown, the prediction steps of the Attention-GRU based machine learning algorithm are:

[0063] S1 extracts the actual operating data and further reduces the dimension of the input data through a dimensionality reduction algorithm to obtain the input vector at this time, including multiple groups of input quantities;

[0064] S2 normalizes the data based on the input vector, and assigns different weights to different input quantities in the normalized input vector according to their characteristics based on Attention, to obtain the corrected input vector;

[0065] S3 inputs the corrected input vector into the GRU-based prediction algorithm to obtain the predicted value of the boiler water-cooled wall temperature at this time.

[0066] In another possible embodiment, the first threshold and the second threshold are determined according to the capacity and type of the boiler.

[0067] In another possible embodiment, a third threshold is further included. When the variation of the measured data is greater than the third threshold, the proportion of the calculated value in the comprehensive prediction function is increased, and the third threshold is greater than the second threshold.

[0068] When the boiler is in the startup and shutdown states, the predicted value is used to predict the boiler water-cooled wall temperature. Since the boiler state is not in normal operation, its prediction accuracy is low. Therefore, the proportion of calculated values ​​is increased to further improve the prediction accuracy without increasing the amount of calculation.

[0069] like Figure 4 As shown, this embodiment 2 provides a soft measurement system for the water-cooled wall temperature of a boiler of a thermal power generator set, which adopts the above-mentioned soft measurement method for the water-cooled wall temperature of a boiler of a thermal power generator set, including:

[0070] Mathematical mechanism model calculation module, machine learning algorithm prediction module, result output module, wherein the mathematical mechanism model calculation module is responsible for establishing a mathematical mechanism model of the boiler water-cooled wall temperature, and performing real-time calculation of the boiler water-cooled wall temperature based on the above mathematical mechanism model to obtain a calculated value; the machine learning algorithm prediction module is responsible for performing real-time prediction of the boiler water-cooled wall temperature based on the machine learning algorithm according to the measured data to obtain a predicted value; the result output module is responsible for constructing a comprehensive prediction function based on the calculated value and the predicted value, wherein the calculated value is updated in real time, and when the time is greater than a first threshold or when the change in the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water-cooled wall temperature at this time.

[0071] In addition, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0072] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0073] With the above-described preferred embodiments of the present invention as a guide, and with reference to the above description, relevant personnel are fully capable of making various changes and modifications without departing from the technical scope of this invention. The technical scope of this invention is not limited to the contents of the specification and must be determined according to the scope of the claims.

Claims

1. A soft measurement method for the water-cooled wall temperature of a thermal power generator boiler, characterized in that: include: Step S1: establishing a mathematical mechanism model of the boiler water wall temperature, and performing real-time calculation of the boiler water wall temperature based on the mathematical mechanism model to obtain a calculated value; Step S2: Based on the measured data, the boiler water wall temperature is predicted in real time using a machine learning algorithm to obtain a predicted value; Step S3, constructing a comprehensive prediction function based on the calculated value and the predicted value, wherein the calculated value is updated in real time, and when the time is greater than a first threshold or when the change in the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water wall temperature at that time; It also includes a third threshold. When the change in the measured data is greater than the third threshold, the proportion of the calculated value in the comprehensive prediction function is increased, and the third threshold is greater than the second threshold.

2. The soft sensing method according to claim 1, wherein: The specific steps of performing real-time calculation of the boiler water wall temperature based on the mathematical mechanism model to obtain the calculated value are as follows: S1 Based on the structural parameters of the furnace, the principal component analysis method is used to perform principal component analysis on the structural parameters and extract the characteristic quantities that affect the boiler water-cooled wall temperature; S2 builds a mathematical mechanism model of the boiler water wall temperature based on the characteristic quantity; S3 sends the real-time measurement point temperature into the mathematical mechanism model of the boiler water-cooled wall temperature to calculate the calculated value of the boiler water-cooled wall temperature at this time.

3. The soft measurement method according to claim 2, characterized in that The mathematical mechanism model of the boiler water-cooled wall temperature performs segmented calculations on the flue gas temperature, combustion temperature, working fluid pressure, enthalpy value, and working fluid temperature.

4. The soft sensing method according to claim 2, wherein: Before performing real-time prediction of the boiler water-cooled wall temperature based on the machine learning algorithm according to the measured data to obtain the predicted value, it is necessary to perform principal component analysis based on the PCA principal component analysis method on the measured data to extract the eigenvalues.

5. The soft measurement method according to claim 4, characterized in that The eigenvalues ​​are subjected to denoising processing based on the Canopy clustering algorithm.

6. The soft sensing method according to claim 1, wherein: The machine learning algorithm adopts an Attention-GRU-based machine learning algorithm.

7. The soft sensing method according to claim 6, characterized in that: The prediction steps of the Attention-GRU based machine learning algorithm are: S1 extracts the actual operating data and further reduces the dimension of the input data through a dimensionality reduction algorithm to obtain the input vector at this time, including multiple groups of input quantities; S2 normalizes the data based on the input vector, and assigns different weights to different input quantities in the normalized input vector according to their characteristics based on Attention, to obtain the corrected input vector; S3 inputs the corrected input vector into the GRU-based prediction algorithm to obtain the predicted value of the boiler water-cooled wall temperature at this time.

8. The soft sensing method according to claim 1, wherein: The first threshold and the second threshold are determined according to the capacity and type of the boiler.

9. A soft measurement system for the water wall temperature of a thermal power generator boiler, using the soft measurement method for the water wall temperature of a thermal power generator boiler according to any one of claims 1 to 8, comprising: Mathematical mechanism model calculation module, machine learning algorithm prediction module, and result output module. The mathematical mechanism model calculation module is responsible for establishing a mathematical mechanism model of the boiler water-cooled wall temperature and performing real-time calculation of the boiler water-cooled wall temperature based on the mathematical mechanism model to obtain a calculated value. The machine learning algorithm prediction module is responsible for real-time prediction of the boiler water wall temperature based on the measured data and the machine learning algorithm to obtain the predicted value; The result output module is responsible for constructing a comprehensive prediction function based on the calculated value and the predicted value, wherein the calculated value is updated in real time. When the time is greater than a first threshold or when the change in the measured data is greater than a second threshold, the predicted value is updated to obtain the boiler water-cooled wall temperature at this time.

Images