Water-cooled wall combustion adjusting system based on model prediction

Abstract

The invention relates to the technical field of water-cooled wall parameter control, in particular to a water-cooled wall combustion adjusting system based on model prediction, which comprises a heat flow path dynamic analysis module, a heat flow path dynamic analysis module, a heat flow path dynamic analysis module and a heat flow path dynamic analysis module, the temperature and pressure data are compared with a preset temperature threshold value and a preset pressure threshold value, the high-heat-load area and the low-heat-load area are marked according to the preset temperature threshold value and the preset pressure threshold value, and the area with the temperature and the pressure continuously increased in the high-heat-load area and without the cooling trend between the adjacent measuring points is marked as the low-heat-load area in the low-heat-load area. And defining a local heat abnormal accumulation region, and generating an abnormal heat flow region analysis result. According to the method, the heat flow path in the water-cooled wall is dynamically analyzed, the surface temperature and pressure data of the water-cooled wall are innovatively acquired by adopting the sensor, and the high heat load area and the low heat load area are accurately marked in combination with threshold comparison and heat load analysis.

Description

Technical Field

[0001] The invention relates to the technical field of water-cooled wall parameter control, and in particular to a water-cooled wall combustion adjustment system based on model prediction. Background Art

[0002] The field of water-cooled wall parameter control technology mainly involves the operation control and optimization of boilers and their combustion systems, especially by adjusting the operating parameters of the water-cooled wall to achieve accurate management of the boiler's internal temperature field, heat load distribution, and combustion efficiency. This technology ensures the safe operation of the boiler and extends the life of the equipment by real-time monitoring and adjusting key parameters such as the temperature, pressure, and flow rate of the water-cooled wall, while improving energy efficiency.

[0003] Among them, the water-cooled wall combustion adjustment system is an optimization system based on water-cooled wall parameter control technology, which is mainly used to accurately adjust the combustion conditions inside the boiler to achieve rationalization of heat load distribution, improvement of combustion efficiency and control of pollutant emissions. By integrating multiple sensors, data processing and feedback control methods, the system can dynamically adjust relevant parameters during the combustion process to adapt to different working conditions, ensure the stability and economy of boiler operation, and protect the water-cooled wall structure.

[0004] Existing technologies rely on fixed water-cooled wall parameter control methods, which makes it difficult to perceive and accurately locate areas of abnormal heat accumulation in real time. The combustion conditions are adjusted only through overall parameters, resulting in the inability to effectively avoid the risk of local overheating. Due to the lack of dynamic analysis of the heat flow path, it is impossible to grasp the specific conduction direction of the heat flow in the high heat load area, making it difficult to predict and intervene in the diffusion trend of heat accumulation, and easily forming uncontrollable hot spots. In addition, the existing methods are unable to perform fine division and dynamic adjustment of abnormal heat load areas, and can only perform rough management of large-scale heat loads. The accuracy is insufficient and it is easy to cause problems of excessive or insufficient regulation. The lack of systematic evaluation of the heat transfer path and identification of key grids makes the ability to control the heat transfer state insufficient, and the response to heat flow disturbances is relatively passive, affecting the overall operating efficiency of the boiler and the life of the water-cooled wall. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose a water-cooled wall combustion adjustment system based on model prediction.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a water-cooled wall combustion adjustment system based on model prediction includes:

[0007] The heat flow path dynamic analysis module obtains the temperature and pressure data of each measuring point based on the dynamic conduction path of the heat flow inside the water-cooled wall. By comparing the temperature and pressure data with the preset temperature threshold and pressure threshold, the high heat load area and the low heat load area are marked according to the preset temperature threshold and pressure threshold. The area with continuous increase in temperature and pressure in the high heat load area and no cooling trend between adjacent measuring points is defined as the area of ​​abnormal local heat accumulation, and the abnormal heat flow area analysis results are generated;

[0008] The regional dynamic division and reconstruction module divides the area of ​​abnormal local heat accumulation in the identified abnormal heat flow area analysis results into multiple grids, while maintaining the original range of the low heat load area, obtains a grid division structure diagram, identifies the heat flow state in each grid area, and generates a dynamic division result of the abnormal area;

[0009] The heat transfer path assessment module collects all heat transfer information of the grids in the local heat abnormal accumulation area based on the abnormal area dynamic division result, calculates the probability of heat transfer from one grid to an adjacent grid, identifies the key grids where heat accumulation will diffuse, and generates grid heat transfer path assessment results;

[0010] The disturbance compensation and dynamic control module is based on the evaluation results of the grid heat transfer path. By simulating the future heat transfer and change trend inside the water-cooled wall, the future heat transfer state of the grid is predicted. If the temperature and pressure data in the current grid deviate from the predicted results, the disturbance compensation is triggered, the control parameters of the grid are dynamically adjusted, and the water-cooled wall combustion parameter adjustment results are generated.

[0011] As a further solution of the present invention, the abnormal heat flow area analysis results include the location of local abnormal heat accumulation, the heat flow transfer direction of the local abnormal heat accumulation area and its corresponding temperature and pressure range; the abnormal area dynamic division results include gridding processing of the abnormal heat flow area, heat load data of each grid, and dynamic boundary adjustment of the grid; the grid heat transfer path evaluation results include the heat transfer priority direction of the grid, identification information of the key grid, and the heat transfer probability between grids; the water-cooled wall combustion parameter adjustment results include fuel and air supply adjustments for the abnormal heat accumulation grid, the disturbance compensation type caused by the deviation between the predicted and current data, and dynamic control of diffusion direction abnormalities.

[0012] As a further solution of the present invention, the step of obtaining the mark of the high heat load area and the low heat load area is specifically:

[0013] Based on the dynamic conduction path of the heat flow inside the water-cooled wall, the temperature and pressure data of each measuring point where the water-cooled wall sensor is arranged are extracted, and the temperature and pressure data of the sensor are converted into numerical form according to the acquisition frequency. After eliminating the abnormal data, the data are sorted out to generate the temperature and pressure data set of each measuring point;

[0014] Based on the temperature and pressure data sets of each measuring point, a heat load analysis is performed on the water-cooled wall surface. Regional classification is performed by comparing the data with preset temperature and pressure thresholds. High heat load areas where both temperature and pressure are higher than the preset temperature and pressure thresholds, and low heat load areas where both temperature and pressure are lower than the preset temperature and pressure thresholds are marked, and a marked water-cooled wall heat load area distribution map is obtained.

[0015] As a further solution of the present invention, the steps for obtaining the abnormal heat flow region analysis results are specifically as follows:

[0016] Based on the data in the marked water-cooled wall heat load area distribution map, extract the temperature and pressure data in the high heat load area, group and analyze the data according to the time series, monitor the temperature and pressure gradient changes between adjacent measuring points, and record the conduction direction of the heat flow to generate dynamic information of the heat flow in the high heat load area;

[0017] Based on the dynamic information of heat flow in the high heat load area, the formula is used:

[0018] ΔT=T i -T i-1 , ΔP=P i -P i-1

[0019] The temperature change rate ΔT between adjacent measuring points in the dynamic information of heat flow in the high heat load area and the pressure change rate ΔP between adjacent measuring points in the dynamic information of heat flow in the high heat load area are calculated respectively, and the abnormal heat flow area analysis results are generated by judging the changes between adjacent measuring points;

[0020] Among them, T i is the temperature value of the i-th measuring point, T i-1 is the temperature value of the i-1th measuring point, P i is the pressure value of the ith measuring point, P i-1 is the pressure value of the i-1th measuring point.

[0021] As a further solution of the present invention, the steps of obtaining the grid division structure diagram are specifically as follows:

[0022] Based on the abnormal heat flow area analysis results, the boundary coordinates of the local abnormal heat accumulation area are read, the area is divided into multiple grids according to the set grid size, and the temperature and pressure data in the grids are allocated to generate preliminary grid division results of the high heat load area;

[0023] Based on the characteristics of the low heat load area, the regional boundaries are extracted and the original range is maintained. The preliminary grid division results of the high heat load area and the boundary information of the low heat load area are integrated, the division results of all areas are uniformly encoded, and the regional grid division structure diagram is generated.

[0024] As a further solution of the present invention, the steps of obtaining the abnormal area dynamic division result are specifically:

[0025] Analyze the temperature and pressure data of each grid in the regional grid division structure diagram in time series, compare the temperature and pressure changes between adjacent grids, determine the conduction direction and intensity of the heat flow, and mark the heat flow state to generate heat flow state information for each grid;

[0026] Based on the heat flow state information of each grid, the heat accumulation phenomenon in the high heat load area is detected, and the dynamic division result of the abnormal area is generated by redefining the grid boundary, dynamically adjusting the boundary coordinates, and allocating the grid data.

[0027] As a further solution of the present invention, the step of obtaining the probability of calculating heat transfer from one grid to an adjacent grid is specifically:

[0028] Based on the abnormal area dynamic division result, the frequency, duration and transfer intensity information of heat transfer are extracted from the data recorded by the grid sensors, and the data are sorted and classified according to the grid numbers to generate a heat transfer information set;

[0029] Based on the heat transfer information set, the formula is used:

[0030]

[0031] Calculate the heat transfer probability P from grid i′ to grid j in the heat transfer information set i′j , indicating the possibility of heat transfer from grid i′ to grid j first, and the heat transfer analysis result is obtained;

[0032] Among them, f i′j is the frequency of heat transfer from grid i′ to a single adjacent grid j, is the total transmission frequency from grid i′ to all adjacent grids k, t i′j is the total duration of heat transfer from grid i′ to grid j, is the total duration from grid i′ to all adjacent grids k, and n is the number of all adjacent grids of grid i′.

[0033] As a further solution of the present invention, the step of obtaining the grid heat transfer path evaluation result is specifically as follows:

[0034] Based on the heat transfer analysis results, the probability between each grid and the adjacent grids is compared, and the transfer direction corresponding to the maximum probability is marked as the priority direction. By counting and marking the priority directions of the grids, a preliminary grid transfer path is generated;

[0035] Based on the preliminary grid transfer path, the transfer frequency and duration data between grids are counted, and the key grids where heat diffusion occurs are screened according to the frequency of transfer times and the cumulative transfer time. The key grids and their corresponding paths are integrated to generate grid heat transfer path evaluation results.

[0036] As a further solution of the present invention, the steps of simulating the future heat transfer and change trend inside the water-cooled wall are specifically as follows:

[0037] Based on the evaluation results of the grid heat transfer path, the grid data of the local abnormal heat accumulation area is integrated, and the priority direction, transfer frequency and duration in the transfer path are analyzed, and the dynamic heat flow relationship between the grids is combined to establish a prediction model for the water-cooled wall heat flow transfer process;

[0038] Based on the prediction model of the water-cooled wall heat flow transfer process, using the previously collected heat flow data and the real-time monitored temperature and pressure data, by simulating the heat transfer and state changes between grids within the future time range, analyzing the trend of heat accumulation and diffusion, and generating the simulation results of the heat transfer and change trend inside the water-cooled wall;

[0039] Based on the simulation results of the heat transfer and change trend inside the water-cooled wall, the formula is adopted:

[0040] T t+1 ′(j′)=T t ′(j′)+ΔT t ′(j′)

[0041] Calculate the thermal state T of grid j′ at the future time t+1 t+1 ′(j′), and obtain the prediction result of the future heat transfer state of the grid;

[0042] Among them, T t ′(j′) is the thermal state of grid j′ at the current time t, ΔT t ′(j′) is the heat change value of grid j′ at time t.

[0043] As a further solution of the present invention, the step of obtaining the water wall combustion parameter adjustment result is specifically as follows:

[0044] Based on the prediction result of the future heat transfer state of the grid, the real-time collected grid temperature and pressure data are compared and analyzed to verify the data deviation area and amplitude, and the abnormal situation in the transfer path is analyzed in combination with the priority direction of the grid transfer path and the distribution of key grids, and the disturbance type is identified, the disturbance compensation is triggered, and the disturbance type identification result is generated;

[0045] Based on the disturbance type identification result, the combustion state of the grid is optimized by adjusting the control parameters of fuel and air, the control parameter values ​​in the heat transfer path are dynamically adjusted to restore the transfer direction, the adjusted grid data is monitored in real time, it is verified that the deviation of heat transfer is restored to a normal range, and the water-cooled wall combustion parameter adjustment result is generated.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are:

[0047] In the present invention, by dynamically analyzing the heat flow path inside the water-cooled wall, sensors are innovatively used to collect the surface temperature and pressure data of the water-cooled wall, and combined with threshold comparison and heat load analysis, high heat load areas and low heat load areas are accurately marked. Based on the heat flow conduction trend in the high heat load area, the abnormal accumulation of local heat is dynamically identified and the heat flow distribution state result is generated. The key areas of heat transfer can be effectively distinguished and located in real time to ensure that the abnormally accumulated heat flow in the high heat load area is intervened in time. Through further grid division of the abnormal area and accurate analysis of the heat flow state, the boundary is dynamically corrected in units of grids, which significantly improves the refinement level of heat distribution control. For the evaluation of the heat transfer path, the heat diffusion direction is predicted by probability analysis, and the diffusion key grid can be accurately identified, providing a basis for subsequent regulation. The prediction model and disturbance compensation are comprehensively adopted to predict and control possible heat flow disturbances, effectively avoiding the risks caused by excessive accumulation of local heat or abnormal diffusion paths. Intelligent monitoring and dynamic regulation of the heat load and heat transfer inside the water-cooled wall are realized, the combustion efficiency is optimized, and the operation stability is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a system flow chart of the present invention;

[0049] Figure 2 A flow chart for marking high heat load areas and low heat load areas for the present invention;

[0050] Figure 3 A flow chart for obtaining abnormal heat flow region analysis results for the present invention;

[0051] Figure 4 A flow chart for obtaining a grid division structure diagram for the present invention;

[0052] Figure 5 A flow chart of the present invention for obtaining the result of dynamic division of abnormal areas;

[0053] Figure 6 A flow chart for calculating the probability of heat transfer from one grid to an adjacent grid for the present invention;

[0054] Figure 7 A flow chart for obtaining grid heat transfer path evaluation results for the present invention;

[0055] Figure 8 A flow chart for simulating the future heat transfer and change trend inside the water-cooled wall according to the present invention;

[0056] Fig. 9 The present invention is a flow chart for obtaining the water-cooled wall combustion parameter adjustment results. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0058] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0059] See also Figure 1 The water-cooled wall combustion adjustment system based on model prediction includes:

[0060] The heat flow path dynamic analysis module is based on the dynamic conduction path of the heat flow inside the water-cooled wall. It obtains the temperature and pressure data of each measuring point from the sensors arranged on the water-cooled wall. According to the temperature and pressure data, the overall heat load analysis of the water-cooled wall surface is performed. By comparing the temperature and pressure data with the preset temperature threshold and pressure threshold, if the current temperature and pressure are higher than the temperature threshold and pressure threshold, it is defined as a high heat load area. If the current temperature and pressure are lower than the temperature threshold and pressure threshold, it is defined as a low heat load area. The high heat load area and the low heat load area are marked, and the conduction direction of the heat flow in the high heat load area is monitored in real time. The area where the temperature and pressure in the high heat load area continue to increase and there is no cooling trend between adjacent measuring points is defined as an area of ​​abnormal local heat accumulation, and the abnormal heat flow area analysis results are generated;

[0061] The regional dynamic division and reconstruction module conducts detailed analysis on the area of ​​abnormal local heat accumulation in the analysis results of the identified abnormal heat flow area, divides the area of ​​abnormal local heat accumulation into multiple grids, and keeps the original range of the low heat load area, obtains the grid division structure diagram, and identifies the heat flow state in each grid area according to the temperature and pressure data in the grid in the grid division structure diagram, as well as the conduction direction in the high heat load path. By correcting the boundary position of the grid area in real time, the area of ​​abnormal local heat accumulation can be accurately divided and controlled, and the dynamic division result of the abnormal area is generated;

[0062] The heat transfer path assessment module collects all heat transfer information of the grids in the local heat abnormal accumulation area based on the dynamic division results of the abnormal area, calculates the probability of heat transfer from one grid to the adjacent grid with reference to the heat transfer frequency and duration, defines the transfer direction with the highest probability as the priority direction of the current grid, identifies the key grids where heat accumulation will diffuse, and generates the grid heat transfer path assessment results;

[0063] The disturbance compensation and dynamic control module uses the model predictive control method to perform disturbance compensation and dynamic control on the heat transfer process in the water-cooled wall based on the grid heat transfer path evaluation results. It establishes a prediction model for the water-cooled wall heat flow transfer process by integrating the data of the grid in the local heat abnormal accumulation area. It uses the prediction model of the water-cooled wall heat flow transfer process to simulate the future heat transfer and change trend inside the water-cooled wall, and predict the future heat transfer state of the grid. If the temperature and pressure data in the current grid deviate from the prediction results, the disturbance compensation is triggered, and the disturbance type is identified for the deviation area. The disturbance type includes the accumulation and diffusion direction deviating from the key grid, which is judged as an abnormal disturbance in the diffusion direction. The control parameters of the grid are dynamically adjusted, including the regulation of fuel and air to restore the priority direction transfer path, and generate the water-cooled wall combustion parameter adjustment results.

[0064] The analysis results of the abnormal heat flow area include the location of local abnormal heat accumulation, the heat flow transfer direction of the local abnormal heat accumulation area and its corresponding temperature and pressure range. The dynamic division results of the abnormal area include the gridding processing of the abnormal heat flow area, the specific heat load data of each grid, and the dynamic boundary adjustment of the grid. The grid heat transfer path evaluation results include the priority direction of grid heat transfer, the identification information of key grids, and the probability of heat transfer between grids. The water-cooled wall combustion parameter adjustment results include the fuel and air supply adjustment of the abnormal heat accumulation grid, the disturbance compensation type caused by the deviation between the predicted and actual data, and the dynamic control of the diffusion direction abnormality.

[0065] See also Figure 2 , the specific steps for marking the high heat load area and the low heat load area are:

[0066] Based on the dynamic conduction path of the heat flow inside the water-cooled wall, the temperature and pressure data of each measuring point where the water-cooled wall sensor is arranged are extracted, and the temperature and pressure data of the sensor are converted into numerical form according to the acquisition frequency. After eliminating the abnormal data, the data are sorted out to generate the temperature and pressure data set of each measuring point;

[0067] The data acquisition and preprocessing operations of the dynamic conduction path of heat flow inside the water-cooled wall are detailed as follows: the raw data are collected from the temperature sensors and pressure sensors arranged on the surface of the water-cooled wall. The raw data may contain noise due to environmental interference or equipment errors, so data cleaning is required. The cleaning steps include removing abnormal values ​​that exceed the rated range of the equipment. For example, the working range of the temperature sensor is -20℃ to 700℃, and the working range of the pressure sensor is 0MPa to 20MPa. Data outside this range will be directly filtered; for data points that are in the normal range but show sudden changes or deviate from the historical mean by 3 times the standard deviation, the sliding average method is used to smooth data fluctuations; the cleaned data are reordered by timestamp to form a continuous time series to ensure the accuracy of subsequent processing. Through preprocessing, the obtained temperature and pressure data can truly reflect the heat flow state collected by the sensor.

[0068] Based on the temperature and pressure data sets of each measuring point, the heat load analysis of the water-cooled wall surface is carried out. By comparing the data with the preset temperature and pressure thresholds, the regions are classified, and the high heat load regions where the temperature and pressure are higher than the preset temperature and pressure thresholds, and the low heat load regions where the temperature and pressure are lower than the preset temperature and pressure thresholds are marked, and the marked water-cooled wall heat load region distribution map is obtained;

[0069] Based on the temperature and pressure data obtained from the sensor, the overall heat load analysis of the water-cooled wall surface is performed. First, the temperature and pressure data of all measuring points are extracted, for example, collected once a minute and stored in the data recording system. The data collection and storage operations are performed through the LabVIEW tool. By decoding and processing the output signal of the sensor, the data point of each measuring point is converted into a specific temperature and pressure value through signal decoding. The temperature threshold is set to 450°C (determined according to the performance of the water-cooled wall heat-resistant material and the operating history data) and the pressure threshold is set to 10MPa (evaluated by the equipment design pressure upper limit and operating test data). The collected measuring point data will be classified through comparison logic. , the measuring point area with temperature greater than 450℃ and pressure greater than 10MPa is marked as a high heat load area. For example, it is detected that the three groups of sensors in area A continuously record temperatures of 460℃, 470℃, and 465℃ and pressures of 10.5MPa, 11MPa, and 10.8MPa respectively, and this area is marked as a high heat load area; conversely, it is detected that the sensors in area B record temperatures of 420℃, 430℃, and 425℃ and pressures of 9.5MPa, 9.8MPa, and 9.6MPa, and this area is marked as a low heat load area. The marking results of high and low heat load areas will be recorded in real time and displayed in the LabVIEW interface for subsequent heat flow analysis and abnormal monitoring.

[0070] See also Figure 3 , the specific steps for obtaining the analysis results of the abnormal heat flow area are:

[0071] Based on the data in the marked water-cooled wall heat load area distribution map, the temperature and pressure data in the high heat load area are extracted, the data are grouped and analyzed according to the time series, the temperature and pressure gradient changes between adjacent measuring points are monitored, and the conduction direction of the heat flow is recorded to generate dynamic information of the heat flow in the high heat load area;

[0072] The conduction direction of heat flow in the high heat load area is monitored in real time through LabVIEW software. After completing the marking of the high heat load area and the low heat load area, the temperature and pressure data of the high heat load area are transmitted, and the collected data are grouped and analyzed according to the time series. The temperature gradient and pressure gradient of adjacent measuring points in each group of data are analyzed. If the measuring points in a certain direction in the high heat load area continuously show a trend of increasing temperature and pressure, the heat flow conduction path is recorded and the conduction direction is marked. At the same time, the temperature and pressure change relationship between the high heat load area and the surrounding low heat load area is monitored to confirm whether there is a trend of heat flow transfer outward.

[0073] Based on the dynamic information of heat flow in the high heat load area, the formula is used:

[0074] ΔT=T i -T i-1 , ΔP=P i-P i-1

[0075] The temperature change rate ΔT between adjacent measuring points in the dynamic information of heat flow in the high heat load area and the pressure change rate ΔP between adjacent measuring points in the dynamic information of heat flow in the high heat load area are calculated respectively, and the abnormal heat flow area analysis results are generated by judging the changes between adjacent measuring points;

[0076] Among them, T i is the temperature value of the i-th measuring point, which is directly measured and recorded by the temperature sensor arranged in the high heat load area, T i-1 is the temperature value of the i-1th measuring point, which is used to calculate the temperature change between the i-th measuring point and the temperature of the i-th measuring point. i is the pressure value at the i-th measuring point, which is directly measured and recorded by the pressure sensor arranged in the high heat load area, P i-1 is the pressure value of the i-1th measuring point, which is used to calculate the pressure change between the i-th measuring point and the pressure at the i-th measuring point.

[0077] Assume that the temperature values ​​recorded at measuring points 1 and 2 in the high heat load area are 460°C and 470°C respectively, and the pressure values ​​are 10.5MPa and 11MPa respectively:

[0078] Calculate the rate of temperature change:

[0079] ΔT=T 2 -T 1 =470-460=10℃

[0080] Calculate the rate of change of pressure:

[0081] ΔP=P 2 -P 1 =11-10.5=0.5MPa

[0082] If the difference between measuring point 1 and measuring point 2 continues to be positive, there is no cooling trend between adjacent measuring points, and there are significant temperature and pressure gradients. This change trend indicates that there may be local abnormal heat accumulation in the high heat load area.

[0083] See also Figure 4 , the specific steps for obtaining the grid division structure diagram are:

[0084] Based on the abnormal heat flow area analysis results, the boundary coordinates of the local abnormal heat accumulation area are read, the area is divided into multiple grids according to the set grid size, and the temperature and pressure data in the grids are allocated to generate preliminary grid division results for the high heat load area;

[0085] The area of ​​abnormal local heat accumulation in the analysis results of the identified abnormal heat flow area is refined and divided into multiple grids. Assume that the boundary of the abnormal heat flow area has been determined as a rectangular area through analysis, with a length of 10 meters and a width of 5 meters. When performing grid division, the division parameters are set using MATLAB software to divide the area into 10×5 regular grids with a length of 1 meter and a width of 1 meter for each grid. The division process is as follows: read the boundary coordinates of the abnormal heat flow area, and define the four vertices of the rectangular area as (0, 0), (10, 0), (10, 5), and (0, 5). Set the size of each grid to 1 meter × 1 meter, calculate the number of grids in each row to 10, the number of grids in each column to 5, and a total of 50 grids. Using the grid division function in MATLAB, the boundary of the rectangular area and the grid size are used as input to generate the boundary coordinates of all grids. For example, the boundaries of the first grid are (0, 0), (1, 0), (1, 1), (0, 1), the boundaries of the second grid are (1, 0), (2, 0), (2, 1), (1, 1), and so on. The temperature and pressure data in the grid are assigned according to the sensor location. For example, the sensor contained in grid (2, 2) recorded a temperature of 450°C and a pressure of 10.5MPa, and these data were assigned to the corresponding grid. Output the boundary range and data distribution of all grids to form a preliminary grid division structure diagram. Each grid contains its number, the coordinates of the four vertices, and the temperature and pressure value set of the internal data points.

[0086] Based on the characteristics of the low heat load area, the regional boundaries are extracted and the original range is maintained. The preliminary grid division results of the high heat load area and the boundary information of the low heat load area are integrated, and the division results of all areas are uniformly coded to generate the regional grid division structure diagram;

[0087] Based on the characteristics of the low heat load area, the boundary points of the low heat load area are extracted from the temperature and pressure data recorded by the sensor to confirm the boundary stability and range continuity of these areas. There is no need to re-divide the grid. Instead, the existing regional boundaries are directly used to retain the low heat load area as a unit area. Subsequently, MATLAB software is used to integrate the fine grid division results of the high heat load area and the boundary information of the low heat load area. The division results of all areas are uniformly encoded to generate a complete regional grid division structure diagram, which contains the number of each grid or unit area, boundary coordinates, and temperature and pressure value distribution of data points in the area.

[0088] See also Figure 5 ,The specific steps for obtaining the results of dynamic division of abnormal areas are:

[0089] The temperature and pressure data of each grid in the regional grid division structure diagram are analyzed in time series, the temperature and pressure changes between adjacent grids are compared, the conduction direction and intensity of the heat flow are determined, and the heat flow state is marked to generate the heat flow state information of each grid;

[0090] Based on the temperature and pressure data within the grid in the grid division structure diagram and the conduction direction in the high heat load path, the heat flow state in each grid area is identified. Assume that the grid division structure diagram of a high heat load area shows a 5×5 grid layout, in which the temperature and pressure data of each grid have been collected and classified by sensors. Assume that the sensor of grid A (1, 1) records a temperature of 470℃ and a pressure of 11.0MPa; grid B (1, 2) records a temperature of 465℃ and a pressure of 10.8MPa; grid C (2, 1) records a temperature of 472℃ and a pressure of 11.1MPa. By comparing the temperature and pressure data between adjacent grids, the conduction direction of the heat flow is determined. For example: from grid A to grid B, the temperature changes from 470℃→465℃ and the pressure changes from 11.0MPa→10.8MPa, indicating that heat is transferred from A to B. From grid A to grid C, the temperature changes from 470℃→472℃, and the pressure changes from 11.0MPa→11.1MPa, indicating that heat is transferred from C to A. The heat flow state is labeled according to the direction and intensity of the heat flow. For example, the heat flow state of grid A is marked as "output to B, input from C"; the state of grid B is "received from A"; and the state of grid C is "output to A". The heat flow state of all grids is recorded in the heat flow state table, which includes the grid number, heat flow conduction direction, temperature and pressure change information.

[0091] Based on the heat flow state information of each grid, the heat accumulation phenomenon in the high heat load area is detected, and the dynamic division results of the abnormal area are generated by redefining the grid boundary, dynamically adjusting the boundary coordinates, and allocating the grid data;

[0092] The boundary position of the grid area is corrected in real time, so that the area of ​​abnormal local heat accumulation can be accurately divided and controlled, and the dynamic division results of the abnormal area are generated. For example, in the grid division of the high heat load area, it is found that the temperature of grid A (2, 2) continues to rise to 480 ° C, while the temperatures of the adjacent grids B (2, 3) and C (3, 2) are 460 ° C and 465 ° C respectively, and the pressure of grid A also rises from 11.2 MPa to 11.5 MPa, indicating abnormal heat accumulation in the local area. When the system detects that the temperature and pressure data inside grid A exceed the average value of the adjacent grids and the gradient between the surrounding grids is large, the dynamic boundary adjustment is triggered. Based on the distribution of sensor data inside grid A, the area where high-temperature data points are concentrated is used as the basis for division, and the boundary of grid A is redefined. For example, grid A is adjusted to cover the range of high-temperature points, and the remaining part is reclassified as grid B and grid C. By correcting the boundary vertex coordinates of grid A, for example, the original boundaries are (2, 2), (3, 2), (3, 3), (2, 3), and the corrected boundaries are (2.5, 2.5), (3, 2.5), (3, 3), (2.5, 3), and the adjusted grid data is redistributed. The corrected grid division results are output, including the new number, boundary coordinates and internal data point distribution of each grid, to ensure that the local abnormal heat accumulation area is accurately defined, while maintaining a smooth transition of temperature and pressure changes between grids.

[0093] See also Figure 6 , the steps to calculate the probability of heat transfer from one grid to an adjacent grid are as follows:

[0094] Based on the results of the dynamic division of abnormal areas, the frequency, duration and intensity of heat transfer are extracted from the data recorded by the grid sensors, sorted and classified according to the grid numbers, and a heat transfer information set is generated;

[0095] Collect all heat transfer information of grids in the area of ​​abnormal local heat accumulation, extract the temperature, pressure and heat flow direction data recorded by the sensor from each grid, and directly read the time series information recorded by the sensor to count the frequency, transfer intensity and duration of heat transfer. Count the frequency of heat transfer per unit time: extract the number of times heat is transferred from one grid to the adjacent grid per unit time (for example, 1 minute) from the heat flow direction data recorded by the sensor. For example, the sensor record from grid A to grid B shows that there are 5 updates of heat flow direction data within 1 minute, so the heat transfer frequency is counted as 5 times / minute. Classify the transfer intensity between grids: According to the temperature difference recorded by the sensor, combined with the known thermal conductivity of the material, the transfer intensity level is classified by calculating the heat flux density. The basis for the division of transfer intensity is: low intensity: heat flux density is less than 20W / m 2 , usually corresponding to a temperature difference of less than 5°C. Medium intensity: heat flux density between 20-50W / m 2The temperature difference is usually between 510℃. High strength: heat flux is greater than 50W / m 2 , usually corresponding to a temperature difference of more than 10°C. For example, the temperature of grid A is 470°C, the temperature of grid B is 460°C, the thermal conductivity of the material is 5W / (m·K), and the corresponding heat flux density is 50W / m 2 , divided into high-intensity transfer. Mark the duration of heat transfer: By analyzing the time series data, identify the continuous heat flow transfer direction record segments. For example, the sensor data from grid A to grid B shows that the heat flow direction remains unchanged for 1 minute and is transferred steadily for 30 seconds, then the duration of the marked heat transfer is 30 seconds. Finally, this information is classified and sorted according to the grid number to generate a heat transfer information table containing the heat transfer path, frequency, intensity level and duration between grids.

[0096] Based on the heat transfer information collection, the formula is used:

[0097]

[0098] Calculate the heat transfer probability P from grid i′ to grid j i′j , indicating the possibility of heat transfer from grid i′ to grid j first, and the heat transfer analysis result is obtained;

[0099] Among them, f i′j is the frequency of heat transfer from grid i′ to a single adjacent grid j, which is obtained by counting the number of changes in the heat flow direction through the sensor. For example, if the transfer direction from grid i′ to grid j changes 4 times in 1 minute, the frequency is 4 times / minute. is the total transmission frequency from grid i′ to all adjacent grids k, t i′j is the total duration of heat transfer from grid i′ to grid j, indicating the cumulative time that the transfer direction remains unchanged. It is extracted from the time series data recorded by the sensor and determined by the cumulative value of the continuous time period of the transfer direction. The transfer direction from grid i′ to grid j remains stable for 20 seconds, 30 seconds and 10 seconds during the monitoring period, so the total duration is 20+30+10=60 seconds. is the total duration from grid i′ to all adjacent grids k, and n is the number of all adjacent grids of grid i′. For example, if grid i′ is adjacent to grid j 1 , j 2 , j 3 If the grid shares a boundary, the number of adjacent grids is 3.

[0100] For example, the adjacent grids of grid 1 are grid 2 and grid 3, and the data recorded by the sensor is as follows: The frequency of grid 1 → grid 2 is 4 times / minute (f 12 =4), the frequency of grid 1→grid 3 is 6 times / minute (f 13=6). The duration of grid 1→grid 2 is 60 seconds (t 12 =60), the duration of grid 1→grid 3 is 30 seconds (t 13 =30).

[0101] Normalized frequency calculation:

[0102]

[0103] Normalized duration calculation:

[0104]

[0105] Comprehensive probability calculation:

[0106] P 12 =0.4·0.6667=0.2667, P 13 =0.6·0.3333=0.2

[0107] The results show that the comprehensive transfer value from grid 1 to grid 2 is 0.2667, and the comprehensive transfer value from grid 1 to grid 3 is 0.2.

[0108] See also Figure 7 , the specific steps for obtaining the grid heat transfer path evaluation results are:

[0109] Based on the heat transfer analysis results, the probability of each grid and the adjacent grids is compared, and the transfer direction corresponding to the maximum probability is marked as the priority direction. By counting and marking the priority directions of the grids, a preliminary grid transfer path is generated;

[0110] The probability of heat transfer from one grid to the adjacent grid is brought into the analysis, and the transfer direction with the maximum probability is defined as the priority direction of the current grid. Assuming that the adjacent grids of grid 1 are 2 and 3, the calculated transfer probabilities are: Grid 1 to grid 2: P 12 =0.2667 (approximately 26.67%), grid 1 to grid 3: P 13 =0.2 (about 20%). After converting these two probability values ​​into percentage form and comparing them, it is determined that the grid 2 with a larger probability is the priority direction of the current grid 1. Therefore, the priority transmission path of the marked grid 1′ is 1→2. Further assume that the adjacent grids of grid 2 are 1 and 4, and their transmission probabilities are: Grid 2 to grid 1: P 2,1 =0.15 (approximately 15%), grid 2 to grid 4: P 2,4=0.3 (about 30%). At this time, grid 4 is defined as the priority direction of grid 2, and the transfer path is marked as 2→4. By calculating the transfer probability grid by grid and marking the priority direction, the heat transfer path can be further drawn. For example: the starting point of the path is grid 1, and the priority path is 1→2→4; all priority paths are integrated to form the overall heat transfer network.

[0111] Based on the preliminary grid transfer path, the transfer frequency and duration data between grids are counted. According to the frequency of transfer times and the cumulative transfer time, the key grids where heat diffusion occurs are screened, and the key grids and their corresponding paths are integrated to generate the grid heat transfer path evaluation results.

[0112] Based on the determined priority transfer direction and path identification, the key grids where heat accumulation may diffuse are identified. Assuming that the priority transfer path of grid 1 is 1→2→4, the frequency and duration data of inter-grid transfer are further analyzed to identify grids that may become key diffusion points: Assuming that the transfer frequency of grid 2 in the transfer path 1→2→4 is the highest among all paths, with 5 transfers recorded per minute, while the transfer frequencies of adjacent grids 1 and 4 are 3 and 2 times, respectively. At the same time, the transfer duration of grid 2 is 50 seconds, which is significantly longer than other grids (such as 30 seconds for 1 and 20 seconds for 4). The high frequency and long duration of grid 2 indicate that grid 2 is a potential key point for heat accumulation diffusion. Mark grid 2 as a key grid and give high priority to its diffusion path 1→2→4. Assuming that grid 5 in another path 3→5 also meets the key grid conditions (high frequency: 6 times per minute; long duration: 60 seconds), grid 5 is also marked as a key grid. The final result marks grids 2 and 5 as key diffusion grids and generates grid heat transfer path assessment results, which include the following: the number of each key grid (such as 2, 5), the preferred transfer path between key grids (such as 1→2→4 and 3→5), and the transfer frequency and duration data of each key grid.

[0113] See also Figure 8 , the specific steps for simulating the future heat transfer and change trend inside the water-cooled wall are:

[0114] Based on the evaluation results of grid heat transfer paths, the grid data of the local abnormal heat accumulation area is integrated. By analyzing the priority direction, transfer frequency and duration in the transfer path and combining the dynamic heat flow relationship between grids, a prediction model for the water-cooled wall heat flow transfer process is established.

[0115] Firstly, the evaluation results of grid heat transfer paths are collected and sorted, including the transfer frequency, duration, priority direction and key grid identification of each grid. Simulink of MATLAB is selected for model construction. The core structural model of the heat transfer system is designed through Simulink, including the connection relationship between the transfer paths between grids, the input and output heat flux values ​​of each grid, the dynamic response of heat transfer delay and cumulative effect. Then, the real-time temperature and pressure data of the local grid are obtained through the sensor network, and a dynamic input interface is established. These real-time data are input into the model for simulation and calibration to ensure that the prediction model reflects the dynamic behavior of the actual heat transfer of the water-cooled wall. Finally, by verifying the consistency between historical data and model prediction results, the model parameters are optimized so that it can accurately predict the heat distribution trend in the water-cooled wall and the heat flux change process of the key grids, forming an operational water-cooled wall heat flux transfer prediction model.

[0116] Based on the prediction model of the heat flow transfer process of the water-cooled wall, the previously collected heat flow data and the real-time monitored temperature and pressure data are used to simulate the heat transfer and state changes between grids in the future time range, analyze the trend of heat accumulation and diffusion, and generate simulation results of heat transfer and change trends inside the water-cooled wall;

[0117] Through the established water-cooled wall heat flow transfer prediction model, the historical heat flow data and the real-time temperature and pressure data collected by the current sensor are input into the model, the simulation time range and step size are set in the MATLAB Simulink environment, and the modular simulation design is used to realize the dynamic transfer of heat flow between grids. The simulation process calculates the heat state of each grid in the future time step according to the priority direction and transfer frequency of the heat flow path, and records the cumulative effect and change trend of heat transfer of each grid during the simulation process. Then, by visualizing the simulation results, the heat distribution map of each grid at different time steps is generated. Combined with the transfer direction and key grid identification, the area where heat accumulation may spread and the dynamic change of the transfer state between grids are further analyzed. Finally, the simulation results are verified with the actual observations to ensure the accuracy of the model prediction and form a detailed description of the future heat change trend inside the water-cooled wall.

[0118] Based on the simulation results of heat transfer and change trend inside the water-cooled wall, the formula is adopted:

[0119] T t+1 ′(j′)=T t ′(j′)+ΔT t ′(j′)

[0120] Calculate the thermal state T of grid j′ at the future time t+1 t+1 ′(j′), and obtain the prediction result of the future heat transfer state of the grid;

[0121] Among them, T t ′(j′) is the thermal state of grid j′ at the current time t. The thermal state value at the current time is obtained directly through the grid monitoring device (such as temperature sensor or calorimeter), ΔT t ′(j′) is the heat change value of grid j′ at time t, indicating the heat transfer, accumulation or loss. t ′(j′) and time step Δt are calculated using the formula: ΔT t ′(j′)=R t ′(j′)·Δt, heat change rate R t ′(j′) is measured from the heat flux density transferred between the grids. The heat flux density can be experimentally recorded as the rate of change of heat flux energy per second. The time step Δt is based on the time interval value set in the simulation, such as 1 minute or 1 second.

[0122] For example, if the initial thermal state of grid j′ is T t ′(j′)=300K, the rate of heat change is R t ′(j′)=5K / min, time step Δt=1min.

[0123] According to the heat change formula:

[0124] ΔT t ′(j′)=R t ′(j′)·Δt=5K / min×1min=5K

[0125] Substituting the heat change value into the main formula:

[0126] T t+1 ′(j′)=T t ′(j′)+ΔT t ′(j′)=300K+5K=305K

[0127] The results show that the thermal state of grid j' in the future can be dynamically calculated through the prediction formula. In this example, the future thermal state of grid j' increases from the current 300K to 305K. This shows that the heat transfer trend causes the grid temperature to rise. Combined with this result, it can be used to further control the combustion process or identify potential overheating risks, providing a quantitative basis for optimizing the operation of the water-cooled wall.

[0128] See also Fig. 9 , the specific steps for obtaining the water-cooled wall combustion parameter adjustment results are:

[0129] Based on the prediction results of the future heat transfer state of the grid, the real-time collected grid temperature and pressure data are compared and analyzed to confirm the data deviation area and amplitude. Combined with the priority direction of the grid transfer path and the distribution of key grids, the abnormal situation in the transfer path is analyzed, and the disturbance type is identified, which triggers disturbance compensation and generates disturbance type identification results.

[0130] If the temperature and pressure data in the current grid deviate from the predicted results, the following process triggers disturbance compensation and identifies the disturbance type: First, collect the real-time temperature and pressure data of the current grid and compare them one by one with the future heat state calculated in the prediction model. For example, the current temperature of grid j' detected by the sensor is 305K, while the predicted value is 300K, and the temperature deviation is +5K. Similarly, the actual value of the pressure data is 101kPa, the predicted value is 100kPa, and the deviation is +1kPa. Combined with the deviation data, the disturbance compensation is triggered and the analysis process of the deviation area is entered. For the deviation area, first confirm whether it is a local heat accumulation problem through the transfer path data between grids. For example, confirm that the heat transfer path of grid j' is compared with the direction of the key grid k', and find that the transfer frequency of grid j' is significantly lower than the historical average, and the transfer direction deviates from the key grid k'. Subsequently, the abnormal situation in the diffusion direction is further confirmed by time series analysis. For example, the heat flux between grids is reduced by 15%, and the transfer delay time exceeds 2 seconds. It is judged that such deviations belong to abnormal disturbances in the diffusion direction. After the compensation mechanism is activated, dynamic recovery is performed by adjusting relevant control parameters, such as increasing the fuel supply of grid j' to increase the heat flux density, optimizing the air supply ratio to improve combustion efficiency, and ensuring that the transfer path gradually recovers to the priority direction. Finally, the heat state data of the adjusted grid is monitored in real time to confirm that the temperature deviation is reduced to within 0.5K, the pressure deviation disappears, and the disturbance type identification and compensation adjustment are completed.

[0131] Based on the disturbance type identification results, the combustion state of the grid is optimized by adjusting the control parameters of fuel and air, and the control parameter values ​​in the heat transfer path are dynamically adjusted to restore the transfer direction. The adjusted grid data is monitored in real time to confirm that the deviation of heat transfer has returned to the normal range, and the water-cooled wall combustion parameter adjustment results are generated;

[0132] After identifying the deviation and triggering disturbance compensation, the priority direction transfer path is restored by dynamically adjusting the control parameters of the grid. The specific process is as follows: Assume that the temperature deviation of grid j′ is +5K and the pressure deviation is

[0133] +1kPa, and it is determined to be an abnormal disturbance in the diffusion direction. In response to this situation, the basic fuel and air supply parameters of grid j' (for example, the fuel flow rate is 100mL / min and the air flow rate is 200mL / min) are first extracted and used as the basis for adjustment. Subsequently, the adjustment amount is calculated according to the deviation amplitude. For example, the temperature deviation +5K corresponds to an increase of 10mL / min in the fuel supply amount and a synchronous increase of 15mL / min in the air supply ratio. These adjustment values ​​are transmitted to the supply device of grid j' in real time through the control module to update the combustion parameters of the grid. After the adjustment is completed, the temperature and pressure changes after adjustment are recorded in real time by the monitoring equipment to confirm that the temperature of grid j' has returned to 300.5K and the pressure has returned to 100 kPa. At the same time, the direction of the transfer path is realigned to the key grid k'. The adjusted fuel and air parameters are stored as new reference values ​​(fuel flow rate 110mL / min, air flow rate 215mL / min) to generate the water-cooled wall combustion parameter adjustment results.

[0134] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A water-cooled wall combustion adjustment system based on model prediction, characterized in that: The system comprises: The heat flow path dynamic analysis module obtains the temperature and pressure data of each measuring point based on the dynamic conduction path of the heat flow inside the water-cooled wall. By comparing the temperature and pressure data with the preset temperature threshold and pressure threshold, the high heat load area and the low heat load area are marked according to the preset temperature threshold and pressure threshold. The area with continuous increase in temperature and pressure in the high heat load area and no cooling trend between adjacent measuring points is defined as the area of ​​abnormal local heat accumulation, and the abnormal heat flow area analysis results are generated; The regional dynamic division and reconstruction module divides the area of ​​abnormal local heat accumulation in the identified abnormal heat flow area analysis results into multiple grids, while maintaining the original range of the low heat load area, obtains a grid division structure diagram, identifies the heat flow state in each grid area, and generates a dynamic division result of the abnormal area; The heat transfer path assessment module collects all heat transfer information of the grids in the local heat abnormal accumulation area based on the abnormal area dynamic division result, calculates the probability of heat transfer from one grid to an adjacent grid, identifies the key grids where heat accumulation will diffuse, and generates grid heat transfer path assessment results; The disturbance compensation and dynamic control module is based on the evaluation results of the grid heat transfer path. By simulating the future heat transfer and change trend inside the water-cooled wall, the future heat transfer state of the grid is predicted. If the temperature and pressure data in the current grid deviate from the predicted results, the disturbance compensation is triggered, the control parameters of the grid are dynamically adjusted, and the water-cooled wall combustion parameter adjustment results are generated.

2. The water-cooled wall combustion adjustment system based on model prediction according to claim 1 is characterized in that: The abnormal heat flow area analysis results include the location of local abnormal heat accumulation, the heat flow transfer direction of the local abnormal heat accumulation area and its corresponding temperature and pressure range; the abnormal area dynamic division results include the gridding processing of the abnormal heat flow area, the heat load data of each grid, and the dynamic boundary adjustment of the grid; the grid heat transfer path evaluation results include the heat transfer priority direction of the grid, the identification information of the key grid, and the heat transfer probability between grids; the water-cooled wall combustion parameter adjustment results include the fuel and air supply adjustment of the abnormal heat accumulation grid, the disturbance compensation type caused by the deviation between the prediction and current data, and the dynamic control of the diffusion direction abnormality.

3. The water-cooled wall combustion adjustment system based on model prediction according to claim 2 is characterized in that: The steps of obtaining the high heat load area and the low heat load area by marking are specifically as follows: Based on the dynamic conduction path of the heat flow inside the water-cooled wall, the temperature and pressure data of each measuring point where the water-cooled wall sensor is arranged are extracted, and the temperature and pressure data of the sensor are converted into numerical form according to the acquisition frequency. After eliminating the abnormal data, the data are sorted out to generate the temperature and pressure data set of each measuring point; Based on the temperature and pressure data sets of each measuring point, a heat load analysis is performed on the water-cooled wall surface. Regional classification is performed by comparing the data with preset temperature and pressure thresholds. High heat load areas where both temperature and pressure are higher than the preset temperature and pressure thresholds, and low heat load areas where both temperature and pressure are lower than the preset temperature and pressure thresholds are marked, and a marked water-cooled wall heat load area distribution map is obtained.

4. The water-cooled wall combustion adjustment system based on model prediction according to claim 3 is characterized in that: The steps for obtaining the abnormal heat flow area analysis results are specifically as follows: Based on the data in the marked water-cooled wall heat load area distribution map, extract the temperature and pressure data in the high heat load area, group and analyze the data according to the time series, monitor the temperature and pressure gradient changes between adjacent measuring points, and record the conduction direction of the heat flow to generate dynamic information of the heat flow in the high heat load area; Based on the dynamic information of heat flow in the high heat load area, the formula is used: ΔT=T i -T i-1 ,ΔP=P i -P i-1 The temperature change rate ΔT between adjacent measuring points in the dynamic information of heat flow in the high heat load area and the pressure change rate ΔP between adjacent measuring points in the dynamic information of heat flow in the high heat load area are calculated respectively, and the abnormal heat flow area analysis results are generated by judging the changes between adjacent measuring points; Among them, T i is the temperature value of the i-th measuring point, T i-1 is the temperature value of the i-1th measuring point, P i is the pressure value of the ith measuring point, P i-1 is the pressure value of the i-1th measuring point.

5. The water-cooled wall combustion adjustment system based on model prediction according to claim 4 is characterized in that: The steps for obtaining the grid division structure diagram are specifically as follows: Based on the abnormal heat flow area analysis results, the boundary coordinates of the local abnormal heat accumulation area are read, the area is divided into multiple grids according to the set grid size, and the temperature and pressure data in the grids are allocated to generate preliminary grid division results of the high heat load area; Based on the characteristics of the low heat load area, the regional boundaries are extracted and the original range is maintained. The preliminary grid division results of the high heat load area and the boundary information of the low heat load area are integrated, the division results of all areas are uniformly encoded, and the regional grid division structure diagram is generated.

6. The water-cooled wall combustion adjustment system based on model prediction according to claim 5 is characterized in that: The specific steps for obtaining the abnormal area dynamic division result are: Analyze the temperature and pressure data of each grid in the regional grid division structure diagram in time series, compare the temperature and pressure changes between adjacent grids, determine the conduction direction and intensity of the heat flow, and mark the heat flow state to generate heat flow state information for each grid; Based on the heat flow state information of each grid, the heat accumulation phenomenon in the high heat load area is detected, and the dynamic division result of the abnormal area is generated by redefining the grid boundary, dynamically adjusting the boundary coordinates, and allocating the grid data.

7. The water-cooled wall combustion adjustment system based on model prediction according to claim 6 is characterized in that: The steps for obtaining the probability of heat transfer from one grid to an adjacent grid are specifically as follows: Based on the abnormal area dynamic division result, the frequency, duration and transfer intensity information of heat transfer are extracted from the data recorded by the grid sensors, and the data are sorted and classified according to the grid numbers to generate a heat transfer information set; Based on the heat transfer information set, the formula is used: Calculate the heat transfer probability P from grid i′ to grid j in the heat transfer information set i′j , indicating the possibility of heat transfer from grid i′ to grid j first, and the heat transfer analysis result is obtained; Among them, f i′j is the frequency of heat transfer from grid i′ to a single adjacent grid j, is the total transmission frequency from grid i′ to all adjacent grids k, t i′j is the total duration of heat transfer from grid i′ to grid j, is the total duration from grid i′ to all adjacent grids k, and n is the number of all adjacent grids of grid i′.

8. The water-cooled wall combustion adjustment system based on model prediction according to claim 7 is characterized in that: The steps for obtaining the grid heat transfer path evaluation result are specifically as follows: Based on the heat transfer analysis results, the probability between each grid and the adjacent grids is compared, and the transfer direction corresponding to the maximum probability is marked as the priority direction. By counting and marking the priority directions of the grids, a preliminary grid transfer path is generated; Based on the preliminary grid transfer path, the transfer frequency and duration data between grids are counted, and the key grids where heat diffusion occurs are screened according to the frequency of transfer times and the cumulative transfer time. The key grids and their corresponding paths are integrated to generate grid heat transfer path evaluation results.

9. The water-cooled wall combustion adjustment system based on model prediction according to claim 8, characterized in that: The steps for simulating the future heat transfer and change trend inside the water-cooled wall are specifically as follows: Based on the evaluation results of the grid heat transfer path, the grid data of the local abnormal heat accumulation area is integrated, and the priority direction, transfer frequency and duration in the transfer path are analyzed, and the dynamic heat flow relationship between the grids is combined to establish a prediction model for the water-cooled wall heat flow transfer process; Based on the prediction model of the water-cooled wall heat flow transfer process, using the previously collected heat flow data and the real-time monitored temperature and pressure data, by simulating the heat transfer and state changes between grids within the future time range, analyzing the trend of heat accumulation and diffusion, and generating the simulation results of the heat transfer and change trend inside the water-cooled wall; Based on the simulation results of the heat transfer and change trend inside the water-cooled wall, the formula is adopted: T t+1 ′(j′)=T t ′(j′)+ΔT t ′(j′) Calculate the thermal state T of grid j′ at the future time t+1 t+1 ′(j′), and obtain the prediction result of the future heat transfer state of the grid; Among them, T t ′(j′) is the thermal state of grid j′ at the current time t, ΔT t ′(j′) is the heat change value of grid j′ at time t.

10. The water-cooled wall combustion adjustment system based on model prediction according to claim 9, characterized in that: The steps for obtaining the water wall combustion parameter adjustment result are specifically as follows: Based on the prediction result of the future heat transfer state of the grid, the real-time collected grid temperature and pressure data are compared and analyzed to verify the data deviation area and amplitude, and the abnormal situation in the transfer path is analyzed in combination with the priority direction of the grid transfer path and the distribution of key grids, and the disturbance type is identified, the disturbance compensation is triggered, and the disturbance type identification result is generated; Based on the disturbance type identification result, the combustion state of the grid is optimized by adjusting the control parameters of fuel and air, the control parameter values ​​in the heat transfer path are dynamically adjusted to restore the transfer direction, the adjusted grid data is monitored in real time, it is verified that the deviation of heat transfer is restored to a normal range, and the water-cooled wall combustion parameter adjustment result is generated.

Images