A method for visualizing the three-dimensional temperature field of a boiler furnace

Abstract

This invention provides a method for visualizing the three-dimensional temperature field of a boiler furnace, belonging to the field of boiler furnace technology. First, it completes the selection, one-to-one installation, and consistency calibration of temperature probes based on the burner array. Then, based on the calibrated furnace flame temperature acquisition hardware system, it synchronously acquires dual-band radiation signals from each burner and unit operating parameters, followed by alignment and noise reduction. Subsequently, based on a comprehensive raw monitoring database covering multiple operating conditions, it extracts the dual-band radiation characteristic values ​​of each burner, converts them using a colorimetric algorithm, and outputs the real-time reference temperature data of the outlet flame for each burner. Based on this, the invention has the ability to intelligently extrapolate sparse, near-wall discrete temperature measurement points into a high-resolution three-dimensional temperature field covering the high-temperature zone at the center of the furnace and the burnout air region, achieving a leap from finite "point" measurement to full-space "volume" reconstruction.

Description

Technical Field

[0001] This invention relates to the field of boiler furnace technology, specifically to a method for visualizing the three-dimensional temperature field of a boiler furnace. Background Technology

[0002] The three-dimensional temperature distribution inside the furnace of a coal-fired boiler is a core parameter for controlling the pulverized coal combustion process, ensuring safe unit operation, improving combustion efficiency, and controlling nitrogen oxide generation. Achieving high-resolution, near real-time visualization of the furnace's three-dimensional temperature field is a key underlying technology supporting intelligent operation and deep peak shaving in power plants.

[0003] Currently, the industry generally faces two major pain points: First, traditional temperature measurement methods, such as thermocouples and single-point infrared probes, are only deployed at limited points in the near-wall area of ​​the furnace, and can only collect local, discrete single-point temperatures, failing to provide sufficient data input for reconstructing the entire temperature field. Second, even with some measured data, existing methods remain at the stage of simple two-dimensional image inversion or purely relying on computational fluid dynamics (CFD) simulation, lacking a complete technical path that can organically integrate "limited measured data," "physical simulation laws," and "data-driven models," and systematically correct physical consistency deviations in the preliminary reconstruction results. Under complex operating conditions such as boiler low-load peak shaving and frequent fluctuations in coal quality, the accuracy and reliability of the reconstructed temperature field cannot meet the needs of refined combustion adjustment. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art, and provides a method for visualizing the three-dimensional temperature field of a boiler furnace.

[0005] This invention provides a method for visualizing the three-dimensional temperature field of a boiler furnace, comprising the following steps: Based on the burner array, the selection, one-to-one installation and consistency calibration of the temperature probe are completed, and the calibrated furnace flame temperature acquisition hardware system is output. Based on the calibrated furnace flame temperature acquisition hardware system, the dual-band radiation signals of each burner and the unit operating parameters are collected simultaneously. After alignment and noise reduction, a complete raw monitoring database of furnace combustion covering multiple operating conditions is output. Based on the original monitoring database of furnace combustion covering multiple operating conditions, dual-band radiation characteristic values ​​of each burner are extracted, converted by colorimetric algorithm, and the real-time reference temperature data of the outlet flame of each burner is output. A combustion CFD simulation model was built based on the boiler structure, and iterative calibration was performed using the real-time reference temperature data of the flame at the outlet of each burner as the boundary condition. The simulation result dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various operating conditions was output. Based on the simulation result dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various working conditions, simulation features are extracted, training samples are matched and constructed, a deep learning network model that integrates physical laws is trained, and an AI model for intelligent prediction of furnace temperature field with physical common sense is output. The real-time reference temperature data of the flame at the outlet of each burner is input into the intelligent prediction AI model of the furnace temperature field with common sense of physics. The model infers and generates an initial three-dimensional temperature field covering the entire furnace area and outputs structured initial temperature field voxel grid data with accurate spatial coordinates. Using the real-time reference temperature data of the flame at the outlet of each burner as a benchmark, the deviation of the structured initial temperature field voxel grid data with precise spatial coordinates at the corresponding points is extracted and the spatial correction coefficient field is reconstructed. The initial field is corrected point by point, and the standard effective data of the real-time three-dimensional temperature field of the furnace is output. Based on the standard valid data of the real-time three-dimensional temperature field of the furnace, pattern matching is performed by combining the abnormal feature rule base to identify abnormal areas and embed fault type alarm labels, and output an enhanced three-dimensional temperature field data volume with diagnostic alarm information. The system integrates standard valid data of real-time three-dimensional temperature field in the furnace with enhanced three-dimensional temperature field data volume containing diagnostic alarm information, generates a three-dimensional temperature cloud map through pseudo-color rendering, and pushes it to the terminal for dynamic visualization display.

[0006] Optionally, the hardware system for selecting, installing, and calibrating temperature probes based on the burner array, and outputting calibrated furnace flame temperature measurement and acquisition data, includes: Select dual-band narrowband temperature and flame detection instruments that can adapt to the harsh working conditions of high temperature and high dust in the furnace, and output the specific model and key performance parameters of the temperature and flame detection equipment that are compatible with this unit. Based on the specific model and key performance parameters of the temperature and flame detection equipment adapted to this unit, the monitoring points near the nozzle of each burner on each floor of the boiler are preset, and one-to-one precise alignment and installation are completed, outputting the deployment of the fixed flame detection probe array. The system is integrated and debugged for the installed and fixed flame detector array. At the same time, the initial temperature measurement accuracy and consistency of each probe are calibrated and debugged using a standard radiation source, and the calibrated furnace flame temperature measurement and acquisition hardware system is output.

[0007] Optionally, the calibrated furnace flame temperature acquisition hardware system simultaneously acquires dual-band radiation signals from each burner and unit operating parameters. After alignment and noise reduction, it outputs a complete raw monitoring database of furnace combustion covering multiple operating conditions, including: Based on a calibrated furnace flame temperature acquisition hardware system, the raw flame radiation signal intensity data of each burner in two narrow-band wavelength channels of 900nm and 1050nm are collected, and a dual-band raw flame radiation signal data stream with precise timestamps is output. Based on the dual-band flame raw radiation signal data stream with precise timestamps, real-time operating condition parameters are simultaneously acquired at the same time, and a full operating condition parameter data stream synchronized with the radiation signal is output. The dual-band flame raw radiation signal data stream with precise timestamps and the full range of operating condition parameter data stream synchronized with the radiation signal time are aligned on a unified time axis to output a full raw monitoring database of furnace combustion covering multiple operating conditions.

[0008] Optionally, based on a comprehensive raw monitoring database of furnace combustion covering multiple operating conditions, the dual-band radiation characteristic values ​​of each burner are extracted, converted using a colorimetric algorithm, and the real-time reference temperature data of the outlet flame of each burner is output, including: Based on the original monitoring database of furnace combustion covering multiple operating conditions, the flame radiation intensity characteristic values ​​of each burner at each sampling time corresponding to the wavelengths of 900nm and 1050nm are extracted, and the dual-band radiation intensity characteristic value sequence of each burner is output. Based on the extracted feature value sequence, it is substituted into the preset colorimetric thermometry algorithm model derived based on Planck's radiation law, and the intermediate calculation result of the flame temperature after disturbance reduction is output. Based on the intermediate calculation results of the flame temperature after interference reduction, and after unit conversion and verification, the real-time reference temperature data of the flame outlet of each burner is output.

[0009] Optionally, the combustion CFD simulation model is built based on the boiler structure, and iteratively calibrated using the real-time reference temperature data of the flame at the outlet of each burner as boundary conditions. The resulting simulation dataset contains the three-dimensional combustion flow field and temperature field distribution inside the furnace under various operating conditions, including: Based on boiler parameters, a basic CFD simulation model of the in-furnace combustion process adapted to this unit is built, and the initial state CFD simulation basic model is output. Select one or more stable operating conditions, compare the real-time reference temperature data of the outlet flame of each burner with the initial state CFD simulation basic model, and perform simulation calculations under the same operating conditions and in the same spatial location. By iteratively adjusting the model parameters in the simulation model, output a calibrated and accurate CFD simulation model that can be used for multi-condition extrapolation. Based on a calibrated and accurate CFD simulation model that can be used for multi-condition simulation, matrix simulation calculations are carried out for boiler load, different coal quality characteristics, and different air-coal ratio combinations. The simulation result dataset includes the three-dimensional combustion flow field and temperature field distribution inside the furnace under each condition.

[0010] Optionally, the process of extracting simulation features from a dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various operating conditions, matching and constructing training samples, training a deep learning network model that integrates physical laws, and outputting an AI model for intelligent prediction of the furnace temperature field with physical common sense includes: Based on the simulation result dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various working conditions, the core feature data is extracted and the core feature dataset of the furnace temperature field simulation under various working conditions is output. The real-time reference temperature data of the outlet flame of each burner is matched with the core feature dataset of the furnace temperature field simulation under each working condition, and the data is matched one by one according to time and working condition to output a dedicated training sample library for the AI ​​model. Based on an AI model-specific training sample library, the samples are enhanced by multi-feature fusion. At the same time, the algorithm simulates typical interference scenarios on site, generates a robust sample subset containing perturbation features and merges it into the original sample set, outputting the final complete training sample set. Based on the final complete training sample set, an LSTM deep neural network that integrates attention mechanism and working condition feature adaptation module is designed and iteratively trained to output an AI model for intelligent prediction of furnace temperature field with physical common sense.

[0011] Optionally, the step of inputting the real-time reference temperature data of the flame at the outlet of each burner into a furnace temperature field intelligent prediction AI model with physical common sense, inferring and generating an initial three-dimensional temperature field covering the entire furnace region, and outputting structured initial temperature field voxel mesh data with precise spatial coordinates includes: The AI ​​model for intelligent prediction of furnace temperature field with common sense of physics is used to import the actual operating parameters of the boiler and the real-time reference temperature data of the flame at the outlet of each burner in real time, and output the pre-processed and aligned model input data vector. Based on the preprocessed and aligned model input data vector, the AI ​​model performs forward inference operations and outputs the original data of the initial global three-dimensional temperature field of the furnace. The original three-dimensional temperature field data of the furnace is horizontally partitioned and decoupled according to the main burner layer, the burnout air layer, and the furnace outlet flue gas temperature zone. At the same time, the temperature data of the interface between adjacent zones is corrected and transitioned using a smoothing algorithm. Then, high-density numerical interpolation fitting is performed on the key layers of each burner layer to output the interpolated temperature field data set. Based on the interpolated temperature field data set, the furnace space is divided into three-dimensional voxel grid cells, and each cell is assigned a derived temperature value, outputting structured initial temperature field voxel grid data with accurate spatial coordinates.

[0012] Optionally, the step of using the real-time reference temperature data of the flame at the outlet of each burner as a scale, extracting the deviation of the structured initial temperature field voxel grid data with precise spatial coordinates at corresponding points and reconstructing the spatial correction coefficient field, performing point-by-point correction on the initial field, and outputting standard valid data of the real-time three-dimensional temperature field of the furnace, includes: Using the real-time reference temperature data of the flame at the outlet of each burner as the calibration point, find the grid cell in the structured initial temperature field voxel grid data with precise spatial coordinates that completely coincides with the spatial coordinates of the calibration point, calculate the temperature deviation value between the two, execute a spatial deviation field reconstruction method with physical field constraints, and output the temperature deviation calibration correction coefficient field acting on the entire furnace space. Based on the temperature deviation calibration correction coefficient field acting on the entire furnace space, point-to-point correction is performed on each grid cell in the structured initial temperature field voxel grid data with precise spatial coordinates, and the furnace grid temperature data with systematic deviations eliminated is output. Based on the furnace grid temperature data with systematic biases eliminated, the 3σ criterion is used to identify and remove low-probability outlier noise data caused by instantaneous operating condition fluctuations or occasional equipment interference, and output standard valid data of the furnace real-time three-dimensional temperature field.

[0013] Optionally, the enhanced three-dimensional temperature field data body, based on standard valid data of the real-time three-dimensional temperature field of the furnace, combined with an anomaly feature rule base for pattern matching, identifies abnormal areas and embeds fault type alarm identifiers, and outputs diagnostic alarm information, includes: Based on the standard effective data of the real-time three-dimensional temperature field in the furnace, the spatial temperature gradient and its changing trend in each region of the furnace are calculated and compared with the normal operating mode defined in the expert experience database to output the coordinate range and key temperature characteristic information of the combustion abnormal area. Based on the coordinate range and key temperature feature information of the combustion anomaly area, it will be correlated with the pre-built temperature feature and fault type matching rule base. At the same time, when the feature pattern of the anomaly area matches a certain fault type in the rule base, the area will be automatically labeled with fault, and the potential fault type and corresponding hazard level judgment information obtained from the matching will be output. Based on the potential fault types and corresponding hazard level determination information obtained from the matching, exclusive combustion abnormality alarm identification points and fault type text labels are generated and embedded, and an enhanced three-dimensional temperature field data volume with diagnostic alarm information is output.

[0014] Optionally, the integrated real-time three-dimensional temperature field standard valid data of the furnace and the enhanced three-dimensional temperature field data volume with diagnostic alarm information are used to generate a three-dimensional temperature cloud map through pseudo-color rendering and pushed to the terminal for dynamic visualization display, including: The standard valid data of the real-time three-dimensional temperature field of the furnace and the enhanced three-dimensional temperature field data volume with diagnostic alarm information are aligned and integrated on the same spatial coordinate grid, and the basic integrated data package for visualization rendering is output for the front end to call. Based on the visualization rendering infrastructure integrated data package for front-end calls, and according to the preset pseudo-color mapping rules, a three-dimensional temperature distribution color cloud map of any cross section of the furnace and the entire furnace is transparent or semi-transparent is generated in real time. The generated three-dimensional temperature distribution color cloud map is pushed in real time via the network, enabling dynamic and continuous refreshing display of the three-dimensional temperature field.

[0015] The boiler furnace three-dimensional temperature field visualization method of this invention constructs a hybrid modeling framework of "actual calibration-simulation transfer-AI reconstruction". It uses high-precision colorimetric temperature measurement of the entire furnace burner outlet as the boundary calibration benchmark, and uses a CFD simulation model verified by actual data to generate a physical law dataset covering multiple operating conditions. It then trains a deep learning network with an attention mechanism, which has the ability to intelligently extrapolate sparse near-wall discrete temperature measurement points into a full-domain high-resolution three-dimensional temperature field covering the high-temperature zone in the center of the furnace and the burnout air zone. This achieves a leap from finite "point" measurement to full-space "volume" reconstruction.

[0016] Furthermore, by internalizing the physical laws of combustion and heat transfer verified by measured boundary calibration into the parameter structure of the LSTM deep neural network, the AI ​​model becomes a high-precision physical proxy model. It has the effect of physically consistently deducing the temperature distribution in areas without measurement points when only the burner outlet temperature is used as input, fundamentally making up for the data blind spots in the key combustion areas of the furnace by traditional methods.

[0017] Furthermore, by introducing a deviation field reconstruction method that uses measured calibration point deviations as observations and the physical characteristics of the initial temperature field as spatial constraints, and employing interpolation algorithms such as co-Kriging that consider multivariable spatial structures, a three-dimensional deviation correction coefficient field that conforms to the combustion organization characteristics in the furnace is generated. This method has the effect of performing physical consistency correction on the AI ​​initial reconstruction field and eliminates systematic deviations caused by unmodeled physical processes such as local coking and changes in radiation characteristics.

[0018] Furthermore, by executing a secondary fine correction closed loop of "physical field constraint spatial deviation field reconstruction - point-by-point correction - noise removal", it can significantly improve the accuracy and reliability of temperature field data under complex working conditions such as coal quality fluctuations and low load peak shaving, thus ensuring the reliability of subsequent combustion anomaly diagnosis and visualization decision-making. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 A flowchart of a method for visualizing the three-dimensional temperature field of a boiler furnace, provided in an embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0022] Unless otherwise specifically stated, the technical or scientific terms used in the embodiments of this invention should be understood in their ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains. The terms "comprising" or "including," as used in the embodiments of this invention, do not limit the shapes, numbers, steps, actions, operations, components, elements, and / or groups thereof mentioned, nor do they exclude the appearance or addition of one or more other different shapes, numbers, steps, actions, operations, components, elements, and / or groups thereof, or the inclusion of these.

[0023] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention. It should also be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale, and techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail; however, where appropriate, the illustrated techniques, methods, and apparatus should be considered part of the specification. In all the examples shown and discussed herein, any other specific example may have different values. It should be noted that similar symbols and letters in the following figures denote similar items; therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0024] In the description of the embodiments of the present invention, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In the embodiments of the present invention, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in the embodiments of the present invention, as well as the features of different embodiments or examples.

[0025] Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0026] like Figure 1 As shown in the figure, this embodiment of the invention provides a method for visualizing the three-dimensional temperature field of a boiler furnace, including the following steps: S1, Selection, deployment and debugging of temperature measurement hardware system: Based on the burner array, complete the selection, one-to-one installation and consistency calibration of temperature measurement probes, and output the calibrated furnace flame temperature measurement and acquisition hardware system.

[0027] The specific steps of S1 are as follows: S11, Instrument Selection: Select a dual-band narrow-band temperature measurement and flame detection instrument suitable for the harsh conditions of high temperature and high dust levels inside the furnace. Preferably, select 900nm and 1050nm, with center wavelengths in the near-infrared region, as the temperature measurement working band pair. The mechanism for this selection is that, for typical dust particle size distributions inside the furnace, the scattering, absorption, and other extinction characteristics of particles at these two adjacent wavelengths are significantly similar. Therefore, when calculating the ratio of the two wavelengths' radiation intensity using a subsequent colorimetric algorithm, the common attenuation interference caused by dust can be canceled to the greatest extent, thereby effectively improving the robustness and accuracy of temperature measurement under harsh dust conditions. Output the specific model and key performance parameters of the temperature measurement and flame detection equipment suitable for this unit. S12, Point Installation: Based on the specific model and key performance parameters of the temperature and flame detection equipment adapted to this unit output by S11, the monitoring points are preset near the nozzle of each burner on each floor of the boiler to complete one-to-one precise alignment and installation. During installation, the field of view of the probe needs to be finely adjusted to ensure that its field of view can completely and unobstructedly cover the core combustion area of ​​the flame at the outlet of the corresponding burner, forming a temperature probe array facing the entire furnace burner, and outputting the deployment of the fixed flame detection probe array; S13, Debugging and Calibration: Deployment of the installed and fixed fire detector array output from S12. The system was integrated and debugged to complete the power supply and redundant communication connection of the equipment. At the same time, the initial temperature measurement accuracy of each probe was calibrated and the consistency was adjusted using a standard radiation source to ensure that all probes in the array were in a ready state with usable and compliant communication. The calibrated furnace flame temperature measurement and acquisition hardware system was then output.

[0028] S2, Synchronous Acquisition of Multi-Source Combustion Related Data: Based on the calibrated furnace flame temperature acquisition hardware system output by S1, it synchronously acquires dual-band radiation signals from each burner and unit operating parameters. After alignment and noise reduction, it outputs a complete raw monitoring database of furnace combustion covering multiple operating conditions.

[0029] The specific steps of S2 are as follows: S21, Radiation Signal Acquisition: Based on the calibrated furnace flame temperature acquisition hardware system output by S1, the system continuously acquires the raw flame radiation signal intensity data of each burner in two narrow-band wavelength channels of 900nm and 1050nm in real time with a sampling period of 100 milliseconds, and outputs a dual-band raw flame radiation signal data stream with precise timestamps. S22, Operating Data Synchronization: Based on the dual-band flame raw radiation signal data stream with precise timestamps output by S21, a communication interface is established with the boiler distributed control system (DCS) to synchronously collect real-time operating parameters such as unit load, coal feed, air volume, and furnace pressure at the same time, and output a full range of operating parameter data stream synchronized with the radiation signal time. S23, Data Organization and Storage: Align the dual-band flame raw radiation signal data stream with precise timestamps output from S21 with the full range of operating condition parameter data stream synchronized with the radiation signal time output from S22 on a unified time axis, identify and remove abnormal jump values ​​caused by communication interference, and deduplicate duplicate data. After being organized and classified, the data is stored and output as a full raw monitoring database of furnace combustion covering multiple operating conditions.

[0030] S3, accurate conversion using colorimetric thermometry: Based on the full original monitoring database of furnace combustion covering multiple operating conditions output by S2, the dual-band radiation characteristic values ​​of each burner are extracted, converted by colorimetric algorithm, and the real-time reference temperature data of the outlet flame of each burner is output.

[0031] The specific steps of S3 are as follows: S31, Feature extraction: Based on the original monitoring database of furnace combustion covering multiple operating conditions output by S2, accurately extract the flame radiation intensity feature values ​​of each burner at each sampling time corresponding to the wavelengths of 900nm and 1050nm, and output the dual-band radiation intensity feature value sequence of each burner. S32, Algorithm Substitution Calculation: Based on the feature value sequence extracted in S31, it is substituted into the preset colorimetric thermometry algorithm model derived based on Planck's radiation law. The core of this algorithm is: by utilizing the significant similarity of the extinction characteristics of smoke and dust at two similar wavelengths as mentioned above, the common attenuation interference factor can be canceled to the greatest extent by calculating the ratio of their radiation intensities, thereby obtaining a calculated value that is closer to the actual flame temperature and outputting the intermediate calculated result of the flame temperature after interference cancellation. S33, Temperature Data Output: Based on the intermediate calculation results of the flame temperature after disturbance reduction output by S32, after unit conversion and verification, the real-time reference temperature data of the flame at the outlet of each burner is output, forming a measured calibration reference dataset for subsequent modeling and correction. Compared with traditional methods, the temperature measurement accuracy of this dataset is significantly improved under typical operating conditions.

[0032] S4, Boiler Combustion CFD Simulation Modeling and Calibration: Based on the boiler structure, a combustion CFD simulation model is built, and iterative calibration is performed using the real-time reference temperature data of the flame at the outlet of each burner output in S3 as boundary conditions. The output is a simulation result dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various operating conditions.

[0033] The specific steps of S4 are as follows: S41, Basic Model Building: Based on parameters such as boiler structural drawings, burner layout coordinates, and layout of heating surfaces at each stage, a basic CFD simulation model of the combustion process in the furnace adapted to this unit is built. This model couples core sub-models such as gas-solid two-phase flow, gas phase turbulence, pulverized coal combustion, and in-furnace radiative heat transfer, and outputs the initial state CFD simulation basic model. S42, Model Boundary Calibration: Select one or more stable operating conditions, and compare the real-time reference temperature data of the outlet flame of each burner output by S3 (as the measured calibration point) with the initial state CFD simulation base model output by S41 under the same operating conditions and in the same spatial location. By iteratively adjusting the model parameters related to wall heat transfer, combustion reaction rate, etc. in the simulation model, the simulation results are made to match the measured data. This step essentially uses limited high-precision measured data to provide reliable boundary conditions for the simulation model and outputs a calibrated and accurate CFD simulation model that can be used for multi-condition extrapolation. S43, Multi-condition Simulation Calculation: Based on the calibrated and accurate CFD simulation model output from S42, which can be used for multi-condition extrapolation, matrix simulation calculations are carried out covering boiler loads from 20% to 100%, different coal quality characteristics, and different air-coal ratio combinations. The output is a dataset of simulation results containing detailed three-dimensional combustion flow field and temperature field distribution inside the furnace under each condition, which contains physical laws.

[0034] S5, AI Temperature Prediction Model Sample Construction and Training: Based on the simulation result dataset of S4, which includes the three-dimensional combustion flow field and temperature field distribution inside the furnace under various working conditions, simulation features are extracted, matching and constructing training samples, training a deep learning network model that integrates physical laws, and outputting an intelligent AI model for predicting furnace temperature field with physical common sense.

[0035] The specific steps of S5 are as follows: S51, Simulation Feature Extraction: Based on the simulation result dataset of the three-dimensional combustion flow field and temperature field distribution inside the furnace under various working conditions output by S4, core feature data such as three-dimensional temperature field distribution and spatial temperature gradient are extracted for different spatial regions inside the furnace, and the core feature dataset of furnace temperature field simulation under various working conditions is output. S52, Sample Dataset Construction and Enhancement: The real-time reference temperature data of the outlet flame of each burner output by S3 is matched with the core feature dataset of furnace temperature field simulation under various working conditions output by S51, and the data is matched one by one according to time and working condition to output a dedicated training sample library for the AI ​​model. S53, classify and label the sample library according to working condition characteristics: Based on the dedicated training sample library of the AI ​​model output by S52, the samples are enhanced by multi-feature fusion. The time-frequency domain features of the flame radiation signal (such as flicker frequency and fluctuation intensity) are fused as a new data dimension. At the same time, the algorithm simulates typical interference scenarios on site (such as high smoke and dust obstruction, signal attenuation, etc.), generates a robust sample subset with disturbance features and merges it into the original sample set, and outputs the final complete training sample set. S54, AI Model Training: Based on the final complete training sample set output by S53, an LSTM deep neural network integrating an attention mechanism and a working condition feature adaptation module is designed and iteratively trained. This training sample set not only includes temperature field distributions under multiple working conditions, but more importantly, its data inherently solidifies the combustion and heat transfer physical laws verified by S42 through actual measurement calibration. The AI ​​model trained with this sample set essentially learns and constructs a high-precision and high-efficiency physical proxy model through a data-driven approach, thereby achieving rapid inference of complex physical fields inside the furnace. The training terminates when the model's overall temperature distribution trend conformity on the validation set and the average relative error of key areas (such as the burner area and furnace outlet) reach the preset convergence criteria, outputting an intelligent prediction AI model of furnace temperature field with physical common sense.

[0036] S6, Initial Reconstruction of Furnace 3D Temperature Field: Input the real-time reference temperature data of the flame at the outlet of each burner output by S3 into the intelligent prediction AI model of furnace temperature field with physical common sense output by S5, infer and generate an initial 3D temperature field covering the entire furnace area, and output structured initial temperature field voxel mesh data with accurate spatial coordinates.

[0037] The specific steps of S6 are as follows: S61, Real-time Data Import: Deploy the AI ​​model for intelligent prediction of furnace temperature field with physical common sense output from S5, import the current actual operating parameters of the boiler and the real-time reference temperature data of the outlet flame of each burner output from S3, and output the pre-processed and aligned model input data vector. S62, Initial Field Calculation Generation: Based on the preprocessed and aligned model input data vector output by S61, the AI ​​model performs forward inference calculations. Through the coupled physical laws of combustion, flow and heat transfer in the furnace that it has learned and internalized during the training phase, the AI ​​model can intelligently deduce the flow field and temperature field distribution in the entire three-dimensional space of the furnace under the condition that only the burner outlet temperature is used as the boundary input. This enables the data reconstruction of the entire temperature field, including the high-temperature combustion zone in the center of the furnace, and outputs the original data of the initial three-dimensional temperature field of the furnace. S63, Combustion Zone Decoupling: The original three-dimensional temperature field data of the furnace output by S62 is horizontally decoupled into the main burner layer, the burnout air layer, and the furnace outlet flue gas temperature zone. At the same time, the temperature data of the interface between adjacent zones is corrected and transitioned using a smoothing algorithm to ensure the continuity of the data boundaries. Then, high-density numerical interpolation fitting is performed on key levels such as the horizontal cross section of each burner layer and the vertical longitudinal section of the furnace center to make the temperature values ​​within the section continuous and smooth, and output the interpolated temperature field data set. S64, Mesh Cell Division: Based on the interpolated temperature field data set output by S63, the furnace space is divided into three-dimensional voxel mesh cells according to the actual geometric dimensions of the furnace and the preset fine mesh accuracy. Each cell is assigned a derived temperature value, and structured initial temperature field voxel mesh data with accurate spatial coordinates is output.

[0038] S7, precise calibration and correction of temperature field data based on physical field constraints: using the real-time reference temperature data of the flame at the outlet of each burner output by S3 as a benchmark, extracting the deviation of the structured initial temperature field voxel grid data with precise spatial coordinates output by S6 at the corresponding points and reconstructing the spatial correction coefficient field, correcting the initial field point by point, and outputting the standard effective data of the real-time three-dimensional temperature field of the furnace.

[0039] The specific steps of S7 are as follows: S71, Physical constraint reconstruction of calibration coefficient field: Using the real-time reference temperature data of the outlet flame of each burner output by S3 as the calibration point, find the grid cell in the structured initial temperature field voxel grid data with precise spatial coordinates output by S6 that completely coincides with the spatial coordinates of the calibration point, calculate the temperature deviation value between the two, and on this basis, execute a spatial deviation field reconstruction method with physical field constraints to output the temperature deviation calibration correction coefficient field acting on the entire furnace space. The method specifically includes: taking the temperature deviation value of each calibration point as the observation value, and using the physical characteristics such as the temperature gradient field and burner jet trajectory contained in the initial temperature field output by S6 as the constraint conditions of spatial correlation. For example, the deviation correlation along the streamline direction is much greater than that perpendicular to the streamline direction. Using methods such as co-kriging interpolation to consider multivariable spatial structure, fit a three-dimensional deviation coefficient distribution field that conforms to the combustion organization characteristics in the furnace. S72, Grid Data Correction: Based on the temperature deviation calibration correction coefficient field output by S71 that acts on the entire furnace space, point-to-point fine correction is performed on each grid cell in the structured initial temperature field voxel grid data with precise spatial coordinates output by S6 to eliminate the systematic bias introduced by the AI ​​model during inference due to unmodeled physical processes (such as changes in radiation characteristics caused by local coking). Through this secondary calibration based on physical field constraints, the accuracy and reliability of the temperature field data under complex actual working conditions are significantly improved, and furnace grid temperature data with systematic bias eliminated is output. S73, Noise Data Removal: Based on the furnace grid temperature data output by S72 with systematic bias eliminated, the 3σ criterion is used to identify and remove low-probability outlier noise data caused by instantaneous operating condition fluctuations or occasional equipment interference. At the same time, the gaps left after removal are filled and smoothed using a neighboring grid interpolation algorithm, and finally the standard valid data of the real-time three-dimensional temperature field of the furnace is output.

[0040] S8, Combustion Abnormal Condition Adaptation and Alarm: Based on the standard valid data of the real-time three-dimensional temperature field of the furnace output by S7, it combines the abnormal feature rule base for pattern matching, identifies abnormal areas and embeds fault type alarm labels, and outputs an enhanced three-dimensional temperature field data volume with diagnostic alarm information.

[0041] The specific steps of S8 are as follows: S81, Abnormal Area Identification: Based on the standard effective data of the real-time three-dimensional temperature field of the furnace output by S7, by calculating the spatial temperature gradient and its changing trend of each region in the furnace, and comparing it with the normal operating mode defined in the expert experience database, it intelligently identifies the core areas of combustion abnormalities such as local overheating and severe temperature field deviation, and outputs the coordinate range and key temperature characteristic information of the combustion abnormal area. S82, Fault Threshold Matching and Labeling: Based on the coordinate range and key temperature feature information of the combustion anomaly area output by S81, it will be correlated with the pre-built temperature feature and fault type matching rule base. This rule base defines the local high temperature and gradient increase characteristics in the early stage of coking of water-cooled wall, the deflection heat load characteristics caused by flame scouring of water-cooled wall, and the smoke temperature distribution characteristics of the combustion center shifting upward or downward. At the same time, when the feature pattern of the anomaly area matches a certain fault type in the rule base, the area is automatically labeled with faults, and the potential fault type and corresponding hazard level judgment information obtained from the matching are output. S83, Enhanced Data Volume Output: Based on the potential fault types and corresponding hazard level judgment information obtained from the matching of S82, at the corresponding spatial coordinate positions in the standard valid data of the real-time three-dimensional temperature field of the furnace, a unique combustion abnormality alarm mark point and fault type text label are generated and embedded, and an enhanced three-dimensional temperature field data volume with diagnostic alarm information is output.

[0042] S9, Dynamic Visualization of 3D Temperature Field: Integrates the standard valid data of real-time 3D temperature field of furnace output by S7 with the enhanced 3D temperature field data volume with diagnostic alarm information output by S8, generates a 3D temperature cloud map through pseudo-color rendering, and pushes it to the terminal to achieve dynamic visualization display.

[0043] The specific steps of S9 are as follows: S91, Rendering Data Integration: Align and integrate the standard valid data of the real-time three-dimensional temperature field of the furnace output by S7 with the enhanced three-dimensional temperature field data volume with diagnostic alarm information output by S8 on the same spatial coordinate grid, and output a basic integrated data package for visualization rendering for front-end use. S92, Temperature Cloud Map Rendering: Based on the visualization rendering basic integrated data package output by S91 for front-end calls, according to the preset pseudo-color mapping rules from low temperature blue to high temperature red and bright white, the computer graphics rendering engine is called to generate a three-dimensional temperature distribution color cloud map of any cross section of the furnace and the entire furnace transparent or semi-transparent in real time. The temperature value is displayed in the cloud map in the form of color mark, and the alarm information is displayed with a bright flashing warning color mark. S93, Terminal Visualization Output: The three-dimensional temperature distribution color cloud map generated by S92 is pushed in real time to the operator station screen or human-machine interface in the power plant's central control room via the network, realizing dynamic and continuous refresh display of the three-dimensional temperature field; the system also supports backtracking query of the temperature field at any historical moment, and automatically pops up an alarm prompt when combustion abnormality is detected, providing intuitive and reliable decision-making basis for operators' fine-tuning and fault prediction.

[0044] The boiler furnace three-dimensional temperature field visualization method of this invention constructs a hybrid modeling framework of "actual calibration-simulation transfer-AI reconstruction". It uses high-precision colorimetric temperature measurement of the entire furnace burner outlet as the boundary calibration benchmark, and uses a CFD simulation model verified by actual data to generate a physical law dataset covering multiple operating conditions. It then trains a deep learning network with an attention mechanism, which has the ability to intelligently extrapolate sparse near-wall discrete temperature measurement points into a full-domain high-resolution three-dimensional temperature field covering the high-temperature zone in the center of the furnace and the burnout air zone. This achieves a leap from finite "point" measurement to full-space "volume" reconstruction.

[0045] Furthermore, by internalizing the physical laws of combustion and heat transfer verified by measured boundary calibration into the parameter structure of the LSTM deep neural network, the AI ​​model becomes a high-precision physical proxy model. It has the effect of physically consistently deducing the temperature distribution in areas without measurement points when only the burner outlet temperature is used as input, fundamentally making up for the data blind spots in the key combustion areas of the furnace by traditional methods.

[0046] Furthermore, by introducing a deviation field reconstruction method that uses measured calibration point deviations as observations and the physical characteristics of the initial temperature field as spatial constraints, and employing interpolation algorithms such as co-Kriging that consider multivariable spatial structures, a three-dimensional deviation correction coefficient field that conforms to the combustion organization characteristics in the furnace is generated. This method has the effect of performing physical consistency correction on the AI ​​initial reconstruction field and eliminates systematic deviations caused by unmodeled physical processes such as local coking and changes in radiation characteristics.

[0047] Furthermore, by executing a secondary fine correction closed loop of "physical field constraint spatial deviation field reconstruction - point-by-point correction - noise removal", it can significantly improve the accuracy and reliability of temperature field data under complex working conditions such as coal quality fluctuations and low load peak shaving, thus ensuring the reliability of subsequent combustion anomaly diagnosis and visualization decision-making.

[0048] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for visualizing a three-dimensional temperature field in a boiler furnace, characterized by Includes the following steps: Based on the burner array, the selection, one-to-one installation and consistency calibration of the temperature probe are completed, and the calibrated furnace flame temperature acquisition hardware system is output. Based on the calibrated furnace flame temperature acquisition hardware system, the dual-band radiation signals of each burner and the unit operating parameters are collected simultaneously. After alignment and noise reduction, a complete raw monitoring database of furnace combustion covering multiple operating conditions is output. Based on the original monitoring database of furnace combustion covering multiple operating conditions, dual-band radiation characteristic values ​​of each burner are extracted, converted by colorimetric algorithm, and the real-time reference temperature data of the outlet flame of each burner is output. A combustion CFD simulation model was built based on the boiler structure, and iterative calibration was performed using the real-time reference temperature data of the flame at the outlet of each burner as the boundary condition. The simulation result dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various operating conditions was output. Based on the simulation result dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various working conditions, simulation features are extracted, training samples are matched and constructed, a deep learning network model that integrates physical laws is trained, and an AI model for intelligent prediction of furnace temperature field with physical common sense is output. The real-time reference temperature data of the flame at the outlet of each burner is input into the intelligent prediction AI model of the furnace temperature field with common sense of physics. The model infers and generates an initial three-dimensional temperature field covering the entire furnace area and outputs structured initial temperature field voxel grid data with accurate spatial coordinates. Using the real-time reference temperature data of the flame at the outlet of each burner as a benchmark, the deviation of the structured initial temperature field voxel grid data with precise spatial coordinates at the corresponding points is extracted and the spatial correction coefficient field is reconstructed. The initial field is corrected point by point, and the standard effective data of the real-time three-dimensional temperature field of the furnace is output. Based on the standard valid data of the real-time three-dimensional temperature field of the furnace, pattern matching is performed by combining the abnormal feature rule base to identify abnormal areas and embed fault type alarm labels, and output an enhanced three-dimensional temperature field data volume with diagnostic alarm information. The system integrates standard valid data of real-time three-dimensional temperature field in the furnace with enhanced three-dimensional temperature field data volume containing diagnostic alarm information, generates a three-dimensional temperature cloud map through pseudo-color rendering, and pushes it to the terminal for dynamic visualization display.

2. The method of visualizing a three-dimensional temperature field in a boiler furnace according to claim 1, characterized in that, The hardware system for selecting, installing, and calibrating temperature probes based on burner arrays, and outputting calibrated furnace flame temperature acquisition data, includes: Select dual-band narrowband temperature and flame detection instruments that can adapt to the harsh working conditions of high temperature and high dust in the furnace, and output the specific model and key performance parameters of the temperature and flame detection equipment that are compatible with this unit. Based on the specific model and key performance parameters of the temperature and flame detection equipment adapted to this unit, the monitoring points near the nozzle of each burner on each floor of the boiler are preset, and one-to-one precise alignment and installation are completed, outputting the deployment of the fixed flame detection probe array. The system is integrated and debugged for the installed and fixed flame detector array. At the same time, the initial temperature measurement accuracy and consistency of each probe are calibrated and debugged using a standard radiation source, and the calibrated furnace flame temperature measurement and acquisition hardware system is output.

3. The method of visualizing a three-dimensional temperature field in a boiler furnace according to claim 1, characterized in that, The calibrated furnace flame temperature measurement and acquisition hardware system simultaneously acquires dual-band radiation signals from each burner and unit operating parameters. After alignment and noise reduction, it outputs a complete raw monitoring database of furnace combustion covering multiple operating conditions, including: Based on a calibrated furnace flame temperature acquisition hardware system, the raw flame radiation signal intensity data of each burner in two narrow-band wavelength channels of 900nm and 1050nm are collected, and a dual-band raw flame radiation signal data stream with precise timestamps is output. Based on the dual-band flame raw radiation signal data stream with precise timestamps, real-time operating condition parameters are simultaneously acquired at the same time, and a full operating condition parameter data stream synchronized with the radiation signal is output. The dual-band flame raw radiation signal data stream with precise timestamps and the full range of operating condition parameter data stream synchronized with the radiation signal time are aligned on a unified time axis to output a full raw monitoring database of furnace combustion covering multiple operating conditions.

4. The method of visualizing a three-dimensional temperature field in a boiler furnace according to claim 1, characterized in that, The database, based on a comprehensive raw monitoring database of furnace combustion covering multiple operating conditions, extracts dual-band radiation characteristic values ​​of each burner. These values ​​are then converted using a colorimetric algorithm to output real-time reference temperature data for the outlet flame of each burner, including: Based on the original monitoring database of furnace combustion covering multiple operating conditions, the flame radiation intensity characteristic values ​​of each burner at each sampling time corresponding to the wavelengths of 900nm and 1050nm are extracted, and the dual-band radiation intensity characteristic value sequence of each burner is output. Based on the extracted feature value sequence, it is substituted into the preset colorimetric thermometry algorithm model derived based on Planck's radiation law, and the intermediate calculation result of the flame temperature after disturbance reduction is output. Based on the intermediate calculation results of the flame temperature after interference reduction, and after unit conversion and verification, the real-time reference temperature data of the flame outlet of each burner is output.

5. The method of visualizing a three-dimensional temperature field in a boiler furnace according to claim 1, characterized in that, The combustion CFD simulation model is built based on the boiler structure, and iteratively calibrated using the real-time reference temperature data of the flame at the outlet of each burner as boundary conditions. The output is a simulation result dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various operating conditions, including: Based on boiler parameters, a basic CFD simulation model of the in-furnace combustion process adapted to this unit is built, and the initial state CFD simulation basic model is output. Select one or more stable operating conditions, compare the real-time reference temperature data of the outlet flame of each burner with the initial state CFD simulation basic model, and perform simulation calculations under the same operating conditions and in the same spatial location. By iteratively adjusting the model parameters in the simulation model, output a calibrated and accurate CFD simulation model that can be used for multi-condition extrapolation. Based on a calibrated and accurate CFD simulation model that can be used for multi-condition simulation, matrix simulation calculations are carried out for boiler load, different coal quality characteristics, and different air-coal ratio combinations. The simulation result dataset includes the three-dimensional combustion flow field and temperature field distribution inside the furnace under each condition.

6. The method of visualizing a three-dimensional temperature field in a boiler furnace according to claim 1, characterized in that, The simulation result dataset, which includes the three-dimensional combustion flow field and temperature field distribution inside the furnace under various operating conditions, extracts simulation features, matches and constructs training samples, trains a deep learning network model that integrates physical laws, and outputs an intelligent AI model for predicting the furnace temperature field with common sense in physics, including: Based on the simulation result dataset containing the three-dimensional combustion flow field and temperature field distribution inside the furnace under various working conditions, the core feature data is extracted and the core feature dataset of the furnace temperature field simulation under various working conditions is output. The real-time reference temperature data of the outlet flame of each burner is matched with the core feature dataset of the furnace temperature field simulation under each working condition, and the data is matched one by one according to time and working condition to output a dedicated training sample library for the AI ​​model. Based on an AI model-specific training sample library, the samples are enhanced by multi-feature fusion. At the same time, the algorithm simulates typical interference scenarios on site, generates a robust sample subset containing perturbation features and merges it into the original sample set, outputting the final complete training sample set. Based on the final complete training sample set, an LSTM deep neural network that integrates attention mechanism and working condition feature adaptation module is designed and iteratively trained to output an AI model for intelligent prediction of furnace temperature field with physical common sense.

7. The method of visualizing a three-dimensional temperature field in a boiler furnace according to claim 1, characterized in that, The process involves inputting real-time reference temperature data of the flame exiting each burner into a furnace temperature field intelligent prediction AI model with physical common sense, inferring and generating an initial three-dimensional temperature field covering the entire furnace region, and outputting structured initial temperature field voxel mesh data with precise spatial coordinates, including: The AI ​​model for intelligent prediction of furnace temperature field with common sense of physics is used to import the actual operating parameters of the boiler and the real-time reference temperature data of the flame at the outlet of each burner in real time, and output the pre-processed and aligned model input data vector. Based on the preprocessed and aligned model input data vector, the AI ​​model performs forward inference operations and outputs the original data of the initial global three-dimensional temperature field of the furnace. The original three-dimensional temperature field data of the furnace is horizontally partitioned and decoupled according to the main burner layer, the burnout air layer, and the furnace outlet flue gas temperature zone. At the same time, the temperature data of the interface between adjacent zones is corrected and transitioned using a smoothing algorithm. Then, high-density numerical interpolation fitting is performed on the key layers of each burner layer to output the interpolated temperature field data set. Based on the interpolated temperature field data set, the furnace space is divided into three-dimensional voxel grid cells, and each cell is assigned a derived temperature value, outputting structured initial temperature field voxel grid data with accurate spatial coordinates.

8. The method for visualizing the three-dimensional temperature field of a boiler furnace according to claim 1, characterized in that, The process uses the real-time reference temperature data of the flame at the outlet of each burner as a benchmark, extracts the deviation of the structured initial temperature field voxel grid data with precise spatial coordinates at corresponding points, reconstructs the spatial correction coefficient field, performs point-by-point correction on the initial field, and outputs standard valid data of the real-time three-dimensional temperature field of the furnace, including: Using the real-time reference temperature data of the flame at the outlet of each burner as the calibration point, find the grid cell in the structured initial temperature field voxel grid data with precise spatial coordinates that completely coincides with the spatial coordinates of the calibration point, calculate the temperature deviation value between the two, execute a spatial deviation field reconstruction method with physical field constraints, and output the temperature deviation calibration correction coefficient field acting on the entire furnace space. Based on the temperature deviation calibration correction coefficient field acting on the entire furnace space, point-to-point correction is performed on each grid cell in the structured initial temperature field voxel grid data with precise spatial coordinates, and the furnace grid temperature data with systematic deviations eliminated is output. Based on the furnace grid temperature data with systematic biases eliminated, the 3σ criterion is used to identify and remove low-probability outlier noise data caused by instantaneous operating condition fluctuations or occasional equipment interference, and output standard valid data of the furnace real-time three-dimensional temperature field.

9. The method for visualizing the three-dimensional temperature field of a boiler furnace according to claim 1, characterized in that, The enhanced three-dimensional temperature field data volume, based on the standard valid data of the real-time three-dimensional temperature field of the furnace, is generated by pattern matching combined with an anomaly feature rule base to identify abnormal areas and embed fault type alarm identifiers, and outputs diagnostic alarm information. This includes: Based on the standard effective data of the real-time three-dimensional temperature field in the furnace, the spatial temperature gradient and its changing trend in each region of the furnace are calculated and compared with the normal operating mode defined in the expert experience database to output the coordinate range and key temperature characteristic information of the combustion abnormal area. Based on the coordinate range and key temperature feature information of the combustion anomaly area, it will be correlated with the pre-built temperature feature and fault type matching rule base. At the same time, when the feature pattern of the anomaly area matches a certain fault type in the rule base, the area will be automatically labeled with fault, and the potential fault type and corresponding hazard level judgment information obtained from the matching will be output. Based on the potential fault types and corresponding hazard level determination information obtained from the matching, exclusive combustion abnormality alarm identification points and fault type text labels are generated and embedded, and an enhanced three-dimensional temperature field data volume with diagnostic alarm information is output.

10. The method for visualizing the three-dimensional temperature field of a boiler furnace according to claim 1, characterized in that, The integrated real-time three-dimensional temperature field standard valid data of the furnace and the enhanced three-dimensional temperature field data volume with diagnostic alarm information are used to generate a three-dimensional temperature cloud map through pseudo-color rendering, which is then pushed to the terminal for dynamic visualization display, including: The standard valid data of the real-time three-dimensional temperature field of the furnace and the enhanced three-dimensional temperature field data volume with diagnostic alarm information are aligned and integrated on the same spatial coordinate grid, and the basic integrated data package for visualization rendering is output for the front end to call. Based on the visualization rendering infrastructure integrated data package for front-end calls, and according to the preset pseudo-color mapping rules, a three-dimensional temperature distribution color cloud map of any cross section of the furnace and the entire furnace is transparent or semi-transparent is generated in real time. The generated three-dimensional temperature distribution color cloud map is pushed in real time via the network, enabling dynamic and continuous refreshing display of the three-dimensional temperature field.

Images