A multi-sensor fusion-based boiler furnace temperature field measurement system

Abstract

The present application relates to the technical field of boiler temperature measurement, and discloses a boiler furnace temperature field measurement system based on multi-sensor fusion. The temperature acquisition module of the system is arranged with multiple groups of infrared temperature measurement sensors and thermocouple sensors in the preset spatial grid of the boiler furnace to synchronously acquire real-time temperature data of each spatial grid; the data fusion module performs time alignment and spatial registration on the collected data to generate a time-space synchronous original temperature data set; the noise suppression module performs multi-scale decomposition on the original data set based on wavelet transform, extracts high-frequency noise components and adaptively filters them to output a temperature data sequence after noise reduction; the temperature field reconstruction module reconstructs a three-dimensional temperature field distribution model of the furnace according to the temperature data sequence after noise reduction through a radial basis function interpolation algorithm; and the dynamic compensation module calculates a heat conduction lag compensation amount of the temperature field distribution model according to the flue gas flow speed in the furnace and the working condition parameters of the burner to generate a compensated dynamic temperature field model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of boiler temperature measurement, in particular to a boiler furnace temperature field measurement system based on multi-sensor fusion. BACKGROUND

[0002] In the modern industrial production system, the power generation boiler is the core thermal equipment of the thermal power plant, and its core function is to convert fuel chemical energy into steam heat energy, and then drive the steam turbine to complete the power generation process, which directly determines the continuity and economy of power production. For power generation boilers that burn high-alkali coal, because high-alkali coal contains more alkali metal compounds, special problems such as slagging and fouling are easily caused during the combustion process, making the stability and efficiency of the boiler operation face more severe challenges.

[0003] The distribution state of the furnace temperature field is a key factor affecting the operation quality of the power generation boiler that burns high-alkali coal, and is directly related to the stability of the boiler operation, energy utilization efficiency and equipment safety.

[0004] Taking thermal power generation as an example, the uniformity and stability of the temperature field in the boiler furnace are key factors to ensure the continuous and stable production process. If the temperature field is not uniform, it may cause local overheating or overcooling. Local overheating will cause the heated surface pipe to withstand excessive thermal stress, accelerate the aging and damage of the pipe, increase the risk of accidents such as pipe explosion, thereby affecting the normal operation of the power generation unit, and even causing shutdown accidents, causing huge losses to power production; while local overcooling may lead to incomplete combustion, reducing the boiler thermal efficiency, affecting steam production and quality, and thus affecting the stability of the entire power generation system.

[0005] In terms of energy utilization efficiency, accurate understanding of the distribution of the furnace temperature field helps to optimize the combustion process and improve energy utilization efficiency. Through monitoring and analysis of the temperature field, the operating parameters of the burner can be reasonably adjusted, such as the mixing ratio of fuel and air, the angle of the burner nozzle and the wind speed, etc., so that the fuel can be fully and uniformly burned in the furnace, reducing incomplete combustion loss and reducing energy consumption.

[0006] The uniformity and stability of the furnace temperature field are also directly related to the equipment safety and operation and maintenance cost of the power generation boiler that burns high-alkali coal. If the temperature field is not uniform for a long time, in addition to causing slagging and fouling problems, it will also cause uneven heating of the heated surface, exacerbate equipment fatigue and wear, shorten the service life of the boiler, and increase the frequency of maintenance and operation and maintenance costs. SUMMARY

[0007] The purpose of the present application is to provide a boiler furnace temperature field measurement system based on multi-sensor fusion to solve the problems raised in the background.

[0008] To achieve the above object, the application provides a boiler furnace temperature field measurement system based on multi-sensor fusion, which comprises:

[0009] A temperature collection module is arranged for arranging a plurality of infrared temperature measurement sensors and thermocouple sensors in a preset spatial grid of the boiler furnace, and synchronously acquiring real-time temperature data of each spatial grid in the furnace;

[0010] A data fusion module is arranged for time aligning and spatially registering the real-time temperature data collected by the infrared temperature measurement sensors and the thermocouple sensors, and generating a time-space synchronous original temperature data set;

[0011] A noise suppression module is arranged for performing multi-scale decomposition on the original temperature data set based on wavelet transform, extracting high-frequency noise components and performing adaptive filtering, and outputting a temperature data sequence after noise reduction;

[0012] A temperature field reconstruction module is arranged for reconstructing a three-dimensional temperature field distribution model of the furnace according to the temperature data sequence after noise reduction through a radial basis function interpolation algorithm;

[0013] A dynamic compensation module is arranged for calculating a heat conduction lag compensation amount of the temperature field distribution model based on the flue gas flow velocity in the furnace and the burner working condition parameters, and generating a compensated dynamic temperature field model.

[0014] Preferably, the specific way of arranging a plurality of infrared temperature measurement sensors and thermocouple sensors in the temperature collection module is as follows:

[0015] The boiler furnace is divided into a plurality of layer annular monitoring areas according to its geometric structure, and at least three infrared temperature measurement sensors are uniformly arranged along the circumferential direction in each layer annular monitoring area;

[0016] A thermocouple sensor chain is vertically installed at the central axis position of each layer annular monitoring area, and the thermocouple sensor chain is composed of a plurality of thermocouple probes distributed at equal intervals;

[0017] The sampling frequency of the infrared temperature measurement sensors and the thermocouple sensors is dynamically adjusted according to the boiler combustion load, and a first sampling frequency is used when the load is higher than a threshold value, and a second sampling frequency is used when the load is lower than the threshold value.

[0018] Preferably, the specific process of generating a time-space synchronous original temperature data set in the data fusion module is as follows:

[0019] The temperature data collected by the infrared temperature measurement sensors is marked with a time stamp, and is mapped to a unified coordinate system of the furnace through spatial coordinate conversion;

[0020] Linear interpolation is performed on the temperature data collected by the thermocouple sensor chain to generate a continuous temperature profile matched with the spatial resolution of the infrared temperature measurement sensors;

[0021] The sliding time window algorithm is used to align the data collection time points of the two types of sensors, and the window length is adaptively adjusted according to the boiler combustion fluctuation period.

[0022] Preferably, the specific steps of performing adaptive filtering in the noise suppression module include:

[0023] The original temperature data set is subjected to discrete wavelet decomposition to obtain approximate coefficients and detail coefficients at different scales;

[0024] Based on the energy entropy threshold method, a noise dominant frequency band is identified in the detail coefficients, and soft threshold filtering is performed on the coefficients in the frequency band;

[0025] A noise template is constructed based on the statistical characteristics of the historical temperature data of the furnace, and periodic interference components are further suppressed through template matching.

[0026] Preferably, the specific method for reconstructing the three-dimensional temperature field distribution model of the furnace in the temperature field reconstruction module through the radial basis function interpolation algorithm is:

[0027] The noise-reduced temperature data sequence is used as the interpolation node of the radial basis function, and the node weight is solved by the least squares method;

[0028] Dirichlet boundary conditions are introduced at the boundary of the furnace wall to constrain the interpolation process, and the boundary temperature values are obtained from the boiler design parameters;

[0029] A multi-layer grid method is used to accelerate the solution of large-scale radial basis function equation systems, and the number of grid layers is automatically optimized according to the volume of the furnace.

[0030] Preferably, the implementation of the dynamic compensation module for calculating the heat conduction lag compensation amount is:

[0031] The distribution of the flue gas velocity field in the furnace is obtained through computational fluid dynamics simulation, and the local flow velocity vectors of each spatial grid are extracted;

[0032] The mapping relationship between temperature propagation delay time and spatial position is established by combining the operating state of the burner and the fuel characteristic parameters;

[0033] Based on the delay time, the dynamic temperature field model is phase-corrected to compensate for the temperature measurement lag effect caused by thermal inertia.

[0034] Preferably, the data fusion module is also configured with a sensor online calibration unit,

[0035] Based on the reading deviation of the infrared temperature sensor and the thermocouple sensor in the same spatial grid under stable working conditions, a dynamic error correction model is established;

[0036] When the reading of any sensor mutates beyond the preset signal interval, the cross-verification and smooth transition algorithm based on adjacent sensor data is started.

[0037] Preferably, the threshold determination method of the energy entropy threshold method is:

[0038] The energy entropy of each scale detail coefficient is calculated and compared with the adaptive threshold trained according to the historical noise database;

[0039] For the suspected effective signal frequency band with energy entropy lower than the threshold, local signal-to-noise ratio evaluation is introduced for secondary discrimination to avoid effective temperature fluctuation being mis-filtered out.

[0040] Preferably, the temperature field reconstruction module further comprises a model accuracy verification unit,

[0041] for consistency comparison between the reconstructed three-dimensional temperature field distribution model and the adiabatic temperature calculated based on the flue gas composition at the furnace outlet;

[0042] If the deviation exceeds the allowed range, the shape parameter of the radial basis function is automatically adjusted, and the reacquisition of local sensor data is triggered.

[0043] Preferably, the heat conduction lag compensation amount calculation process of the dynamic compensation module is periodically updated, and the update strategy is:

[0044] The burner load change rate is monitored in real time, and when the change rate exceeds the critical value, a new round of computational fluid dynamics simulation is triggered to update the flue gas velocity field distribution;

[0045] Under steady state conditions, the mapping relationship is periodically updated at preset time intervals.

[0046] Compared with the prior art, the present application has the following advantages:

[0047] The temperature acquisition module of the present application can synchronously acquire real-time temperature data of each spatial grid in the furnace by carefully arranging multiple groups of infrared temperature measurement sensors and thermocouple sensors in the preset spatial grid of the boiler furnace. Compared with the traditional temperature measurement method, which often has limited measurement points and can only acquire temperature information of a few points in the furnace, the present application can more comprehensively and meticulously capture temperature changes at different positions in the furnace by arranging multiple sensors, thereby effectively solving the problem of the traditional method that cannot comprehensively reflect the temperature field and providing a rich data basis for subsequent analysis and research of the furnace temperature field.

[0048] The processing of data by the data fusion module and the noise suppression module ensures the reliability of the data. The data fusion module performs time alignment and spatial registration on the real-time temperature data collected by the infrared temperature sensor and the thermocouple sensor, which can eliminate the differences in time and space of different sensors, make the collected data consistent in time and space, and generate a set of original temperature data that is synchronized in time and space. This is like accurately assembling puzzle pieces from different directions to make a complete and orderly picture. The noise suppression module performs multi-scale decomposition on the original temperature data set based on wavelet transform, which can deeply mine the detailed information in the data, accurately extract high-frequency noise components, and perform adaptive filtering. This processing method can effectively remove noise interference in the data, like a fine cleaning of the data, making the data more pure and accurate, providing reliable data support for subsequent temperature field reconstruction and analysis.

[0049] The temperature field reconstruction module uses the radial basis function interpolation algorithm to reconstruct the three-dimensional temperature field distribution model of the furnace according to the denoised temperature data sequence, which has important practical application value. Through this model, the staff can intuitively see the three-dimensional distribution of the temperature in the furnace, clearly understand the temperature level, temperature gradient and temperature change trend in different areas of the furnace. This is like providing a three-dimensional perspective mirror for the staff to understand the internal temperature of the furnace, so that they can have a more comprehensive and in-depth understanding of the combustion state of the furnace. Compared with the traditional two-dimensional temperature representation method, the three-dimensional temperature field distribution model can provide more information, which helps the staff to timely find possible local overheating, subcooling or uneven combustion in the furnace, so as to take corresponding measures to adjust and optimize, and ensure the safe and stable operation of the boiler. BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1 The working principle diagram of the boiler furnace temperature field measurement system based on multi-sensor fusion described in the present application;

[0051] Figure 2 The working principle flow chart of the sensor arrangement method in the temperature collection module;

[0052] Figure 3 The working principle flow chart of the process of generating a set of original temperature data synchronized in time and space by the data fusion module. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0054] Referring to Figure 1 The application provides a multi-sensor fusion-based boiler furnace temperature field measurement system, which comprises a temperature acquisition module, a plurality of infrared temperature measurement sensors and thermocouple sensors are arranged in a preset spatial grid of a boiler furnace, the infrared temperature measurement sensors and the thermocouple sensors synchronously acquire real-time temperature data of each spatial grid in the furnace, and the sampling frequency is dynamically adjusted based on the boiler combustion load during the data acquisition process to ensure data integrity. A data fusion module receives the real-time temperature data acquired by the infrared temperature measurement sensors and the thermocouple sensors, performs time alignment and spatial registration processing on the data, generates a time-space synchronous original temperature data set, and adopts a sliding time window algorithm during the processing to adapt to the combustion fluctuation of the boiler. A noise suppression module performs multi-scale decomposition on the original temperature data set through wavelet transform, extracts high-frequency noise components and performs adaptive filtering, and outputs a temperature data sequence after noise reduction, and the filtering process identifies noise bands based on an energy entropy threshold method. A temperature field reconstruction module reconstructs a three-dimensional temperature field distribution model of the furnace according to the temperature data sequence after noise reduction through a radial basis function interpolation algorithm, and the interpolation process introduces boundary condition constraints and adopts a multi-layer grid method to accelerate the solution. A dynamic compensation module calculates a heat conduction lag compensation amount of the temperature field distribution model based on the flue gas flow speed and the burner working condition parameters in the furnace, generates a compensated dynamic temperature field model, and the compensation amount is obtained by fluid mechanics simulation to obtain a flue gas speed field distribution and establish a delay mapping relationship.

[0055] Embodiment 1: Referring to Figure 2 The geometric structure of the boiler furnace is divided into a plurality of layers of annular monitoring regions, and the division is based on the characteristics of the combustion flow inside the furnace and the general law of thermal distribution. At least three infrared temperature measurement sensors are uniformly arranged along the circumferential direction in each layer of annular monitoring regions, and the installation positions of the infrared temperature measurement sensors are optimized by computational fluid dynamics simulation, aiming to completely capture the temperature inhomogeneity on the cross section of the furnace. The optical lens of the infrared temperature measurement sensor faces the central region of the furnace, and the viewing angle range is accurately calculated to avoid mutual shielding and cover the entire annular region. The sensor shell is protected by a water cooling kit to withstand high temperature environment. A thermocouple sensor chain is vertically installed at the central axis position of each layer of annular monitoring regions, and the thermocouple sensor chain is composed of a plurality of thermocouple probes distributed at equal intervals, and the interval of the thermocouple probes is determined according to the expected axial temperature gradient. The thermocouple sensor chain adopts a protective sleeve made of high-temperature resistant alloy material, and the thermocouple probes are fixed to the support structure on the top or bottom of the furnace through flanges, forming a vertical measurement line penetrating through the plurality of layers of annular monitoring regions. The installation of the thermocouple sensor chain ensures that it coincides with the central axis of the furnace, minimizing the measurement error caused by deviation, and the lead of the thermocouple sensor is connected to the data acquisition box through a shielded cable.

[0056] The sampling frequency of the infrared temperature sensor and the thermocouple sensor is dynamically adjusted according to the boiler combustion load, and the real-time data of the boiler combustion load is directly provided by the boiler control system. When the load is higher than the threshold, the first sampling frequency is used, and the value of the first sampling frequency is higher to adapt to the rapidly changing temperature field when the combustion fluctuates sharply; when the load is lower than the threshold, the second sampling frequency is used, and the value of the second sampling frequency is lower to meet the basic monitoring requirements under stable working conditions and reduce the data storage pressure. The sampling frequency switching logic is set with a certain hysteresis interval to prevent frequent switching of the sampling mode when the load fluctuates near the threshold, and the frequency adjustment instruction is executed by the main controller in the data acquisition module. The data acquisition process of the infrared temperature sensor and the thermocouple sensor is kept synchronous, and the synchronous signal is generated by a unified clock source. Each data point is attached with a high-precision time stamp. The division of the spatial grid is not only based on the geometric structure, but also considers the layout of the burner nozzle and the flame shape, and the grid size is relatively dense in the high-temperature flame area and relatively sparse in the outlet area. The infrared temperature sensor uses a model with a specific infrared wavelength to reduce the absorption interference of water vapor and carbon dioxide in the flue gas on the measurement signal, and the periodic calibration of the infrared temperature sensor is carried out on site using a standard blackbody furnace. The thermocouple probe in the thermocouple sensor chain uses K-type or S-type thermocouple, and the cold end compensation of the thermocouple is completed in the data acquisition box, and the compensation accuracy meets the industrial measurement standard. The verification work of the sensor arrangement scheme is carried out during the boiler shutdown maintenance, and the consistency of the actual installation position of the sensor and the design coordinate is confirmed by three-dimensional laser scanning, and the existing deviation is recorded and software corrected during data fusion. The preset spatial grid of the boiler furnace is a three-dimensional grid, and the grid node coordinates are stored in the system database as the reference for spatial registration of all temperature data.

[0057] The power supply of the infrared temperature sensor and the thermocouple sensor is provided by different power sources, the signal transmission line adopts electromagnetic shielding measures, and the data acquisition module has anti-interference capability. The running state of the temperature acquisition module is monitored in real time, the module is provided with self-diagnosis function, and faults such as sensor open circuit, short circuit or signal abnormal attenuation can be detected. When it is detected that an infrared temperature sensor or a thermocouple sensor fails, the system can compensate temporarily according to the redundancy design principle by using the data of the adjacent sensor, and the maintenance alarm information will be sent to the control room. When the geometric structure of the boiler furnace is divided into annular monitoring areas, the number of layers is proportional to the height of the furnace, and the height of each layer is usually matched with the height of the burner group. The division of the annular monitoring area ensures that the difference between the temperature core area and the edge area can be distinguished in the radial direction, and the infrared temperature sensor is uniformly arranged in the circumferential direction of each layer to form a minimum temperature field reconstruction unit. The axial temperature profile data provided by the thermocouple sensor chain and the radial temperature distribution data provided by the infrared temperature sensor complement each other, and constitute the basic framework of spatial temperature information acquisition. The synchronization of data acquisition not only lies in the time stamp, but also the hardware trigger mechanism of the acquisition module ensures that all channels sample and hold at the same time, eliminating the spatial registration error caused by different time.

[0058] The threshold of the sampling frequency dynamic adjustment is set according to the design load of the boiler, and is optimized in combination with historical operation data. The threshold can be set to multiple levels to adapt to more complex load change modes. The specific values of the first sampling frequency and the second sampling frequency are determined according to the Nyquist sampling theorem and the characteristic frequency of the main thermal process of the boiler, so as to avoid the frequency aliasing phenomenon. The arrangement scheme of the infrared temperature sensor and the thermocouple sensor fully considers the harsh working environment inside the boiler furnace, and factors such as high temperature, dust and vibration are considered in the selection and installation structure design of the sensor. The preliminary validity check of the sensor data is completed in the acquisition module, and the check items include signal amplitude range, change rate rationality, etc. Invalid data will be marked and excluded from the subsequent processing process. The communication between the temperature acquisition module and the existing distributed control system of the boiler adopts a standard industrial protocol, realizes the real-time acquisition of the working condition parameters such as the combustion load, and the working state of the module itself is also uploaded to the control system. The preset space grid information of the boiler furnace is configurable, which allows adjustment according to different types and sizes of boilers, and the grid information configuration file is stored in the non-volatile memory. The arrangement of the infrared temperature sensor and the thermocouple sensor forms a three-dimensional sensor network, and the topology structure of this network determines the upper limit of the spatial resolution of the subsequent temperature field reconstruction model. The design of the whole temperature acquisition module follows the reliability and availability requirements of the industrial environment, and has the characteristics of dustproof, waterproof and high temperature resistance, and can work stably in the long-term operation process of the boiler. The continuity and integrity of the data stream are guaranteed by the cache technology and the data packet retransmission mechanism, so that even in the case of temporary communication interruption, the temperature data sequence will not be seriously missing.

[0059] Example 2: see Figure 3The temperature data collected by the infrared temperature sensor is marked with a time stamp with millisecond-level precision, and the time stamp information comes from a high-stability crystal oscillator clock source within the system. The time stamp is packaged together with the temperature value to form a data frame, which is transmitted to the buffer memory of the data fusion module through industrial Ethernet. Spatial coordinate conversion maps the temperature values measured by the infrared temperature sensor to a unified coordinate system of the furnace, with the geometric center of the boiler furnace as the origin and the three coordinate axes pointing to the width, depth and height directions of the boiler. Spatial coordinate conversion requires the accurate installation position parameters of the infrared temperature sensor in the furnace, which are obtained by total station measurement during the boiler installation and debugging phase and stored in the system configuration file. Linear interpolation is performed on the temperature data collected by the thermocouple sensor chain, and the purpose of linear interpolation is to generate a continuous axial temperature distribution curve between each thermocouple probe of the thermocouple sensor chain. The measurement points of the thermocouple sensor chain are located on the central axis of the furnace, and the spatial resolution is lower than that of the circumferentially arranged infrared temperature sensor. The linear interpolation algorithm inserts multiple virtual measurement points between adjacent thermocouple probes. The interpolated thermocouple data matches the measurement grid of the infrared temperature sensor in spatial distribution, and each spatial grid position corresponding to an infrared temperature sensor can obtain an interpolated temperature data from the thermocouple sensor chain. The linear interpolation algorithm uses a piecewise linear function, and the interpolation nodes are the actual measurement positions of the thermocouple sensor chain. The interpolation calculation is performed in real time at each data acquisition cycle.

[0060] The sliding time window algorithm is used to align the data collection time points of the infrared temperature sensor and the thermocouple sensor. The sliding time window algorithm groups the data samples of the two types of sensors that are close in time into the same processing window. The setting of the window length is related to the fluctuation period of the boiler combustion process, and the combustion fluctuation period is obtained by analyzing the frequency characteristics of temperature changes in historical operation data. The window length is adaptively adjusted according to the boiler combustion fluctuation period. When the combustion is intense, the window length is shortened to capture rapid dynamic changes. When the combustion is stable, the window length is lengthened to improve the data smoothness. The infrared temperature sensor data and the thermocouple sensor data in the sliding time window are unified to the same timestamp sequence through an interpolation algorithm, generating strictly synchronized data pairs. The data fusion module is configured with a sensor online calibration unit, which uses the reading differences of the infrared temperature sensor and the thermocouple sensor in the same spatial grid under stable working conditions. Stable working conditions refer to the operating state in which the main operating parameters such as boiler load, fuel quantity, and air volume remain unchanged, and the furnace temperature distribution is relatively stable. The sensor online calibration unit compares the readings of the infrared temperature sensor and the thermocouple sensor at the same location, calculates the system deviation between them, and the system deviation may be caused by sensor characteristic drift or slight changes in installation position. A dynamic error correction model is established based on long-term statistical system deviation data, and the dynamic error correction model is a polynomial function, and the coefficients are obtained by least squares fitting of historical deviation data.

[0061] When the reading of any sensor suddenly changes beyond the pre-set confidence interval, the cross-validation and smooth transition algorithm based on adjacent sensor data is started. The pre-set confidence interval is set according to the long-term running statistical characteristics of the sensor, and the upper and lower limits of the interval are usually three times the standard deviation of the mean value. The sudden change in reading may indicate that the sensor has a transient failure or is strongly disturbed, and the cross-validation algorithm checks the data of other sensors within a certain spatial range around the sensor with sudden change. If the adjacent sensor data does not show similar mutations, the mutant data is determined as an abnormal value, and the smooth transition algorithm uses the weighted average of the adjacent sensor data to replace the abnormal value. If the adjacent sensor data also shows coordinated changes, it may reflect a sharp change in the real temperature field, and the data is retained but marked for close monitoring. The spatiotemporally synchronized raw temperature data set is stored in the ring buffer of the data fusion module, and the data format uses a floating-point number array, each data point containing spatial three-dimensional coordinates, temperature value, timestamp, and data type identifier. The processing logic of the data fusion module is implemented by a programmable logic device, and the algorithm operation period is strictly synchronized with the data collection period to ensure the real-time nature of data processing. The structural deformation error caused by the thermal expansion of the furnace needs to be corrected in the spatial registration process, and the amount of thermal expansion of the furnace is measured by displacement sensors installed on the furnace shell and used for real-time compensation of coordinates. The calibration coefficients of the sensor online calibration unit are written to the non-volatile memory regularly to prevent data loss due to power failure, and the calibration records are also uploaded to the boiler plant information management system for equipment status analysis.

[0062] Data loss or communication interruption is managed by the exception handling mechanism of the data fusion module, short-time data loss is compensated by the last cycle data retention or linear extrapolation, and long-time interruption triggers system alarm. The data interaction between the data fusion module and the boiler master control system adopts the publish-subscribe mode, and the temperature data is subscribed by other system modules that need data as public information. The execution frequency of the sensor online calibration unit can be configured, usually increasing the calibration frequency at the beginning of the boiler startup and after the load changes greatly, and reducing the frequency during stable operation to save computing resources. The output interface of the data fusion module provides two data streams of the original temperature data set and the temperature data set after preliminary quality control, which are selected for use in different precision requirement application scenarios. The spatial coordinate conversion of the infrared temperature sensor data involves coordinate transformation matrix operation, and the transformation matrix parameters are fixed values after the sensor installation and positioning. The coordinate conversion calculation is performed on the graphics processor to speed up the processing. The linear interpolation of the thermocouple sensor chain not only considers the axial position, but also considers the radial position of the furnace due to the central location of the thermocouple sensor chain, and its data represents the temperature on the axis, which is complementary to the off-line data measured by the infrared temperature sensor. The window moving step of the sliding time window algorithm is less than the window length, so that there is data overlap between adjacent windows, avoiding the occurrence of data discontinuity at the window boundary. The dynamic error correction model of the sensor online calibration unit has a forgetting factor, and the weight of recent data is higher than that of early data, so that the model can track the slow time-varying characteristics of the sensor.

[0063] In the cross-validation and smooth transition algorithm, the selection of neighboring sensors is based on spatial distance weighting; the closer the sensor is to the sensor experiencing a sudden change, the higher its data weight. The weighting coefficient is determined based on spatial correlation analysis of historical data. The operational status of the data fusion module is monitored by a hardware watchdog circuit to prevent program crashes or infinite loops. In case of module failure, it can switch to a backup module or send a fault signal to the main control system. The storage format of the raw temperature dataset supports time-series retrieval, facilitating historical data backtracking analysis. The data compression algorithm reduces storage space usage while maintaining accuracy. The operation of the sensor online calibration unit must comply with boiler safety operating procedures. The calibration process must not disturb the boiler combustion stability. Calibration commands are issued by authorized personnel or automatically executed by the system under deemed safe conditions. The software code of the data fusion module adopts a modular design, with coordinate transformation, interpolation calculation, time alignment, and online calibration functions as independent blocks, facilitating individual debugging and upgrade maintenance. The internal data flow of the module uses parallel pipeline processing to improve data throughput, and the real-time operating system ensures that the scheduling of each task thread meets the time limit requirements. The online sensor calibration unit generates a calibration report based on the calibration results. The report includes information such as calibration time, sensor identification, pre-calibration deviation, post-calibration deviation, and changes in calibration model parameters. The data fusion module's hardware platform utilizes an industrial-grade server with redundant power supplies and heat dissipation design, meeting the reliability requirements for long-term continuous operation in power plant environments.

[0064] Example 3: The noise suppression module receives a spatiotemporally synchronized raw temperature dataset from the data fusion module. This raw temperature dataset contains temperature values ​​with spatial coordinates and timestamps provided by infrared temperature sensors and thermocouple sensors. Discrete wavelet decomposition is performed on the raw temperature dataset. The mother wavelet function is used to decompose the temperature signal into approximation coefficients and detail coefficients at different scales. The number of decomposition levels is determined based on the relationship between the main frequency components of the signal and the sampling frequency, ensuring effective separation of noise and true temperature fluctuations in the signal. The discrete wavelet decomposition calculation is performed on a digital signal processor, employing a fast algorithm to improve processing efficiency and meet the system's real-time requirements.

[0065] The energy entropy thresholding method is used to identify noise-dominant frequency bands in detail coefficients. This method quantifies the irregularity of detail coefficients by calculating their energy distribution characteristics at each scale. The formula for calculating energy entropy is:

[0066] ,

[0067] in: Indicates the first The energy entropy value of detail coefficients at each scale It is the first The detail coefficient at each scale It is the first the total number of detail coefficients at a scale, is the total energy of all detail coefficients at a scale, i.e. The energy entropy value reflects the concentration degree of the energy distribution of the coefficients, and the scale with a higher energy entropy value usually corresponds to a noise-dominant frequency band.

[0068] The calculated energy entropy values of each scale are compared with an adaptive threshold value obtained by training a historical noise database. The historical noise database stores the background noise characteristics collected by the boiler under various typical operating conditions. The adaptive threshold value is a classification boundary obtained by machine learning training on a large number of samples in the historical noise database, and the adaptive threshold value is dynamically adjusted with the updating of the database. For scales with energy entropy higher than the adaptive threshold value, it is determined that they are noise-dominant frequency bands, and soft threshold filtering is performed on the detail coefficients of the noise-dominant frequency bands. Soft threshold filtering compares the absolute value of the coefficient with a threshold parameter, and the coefficient less than the threshold is set to zero, and the coefficient greater than the threshold is shrunk to zero. For suspected effective signal frequency bands with energy entropy lower than the adaptive threshold value, a local signal-to-noise ratio evaluation is introduced for secondary discrimination, and the local signal-to-noise ratio evaluation calculates the ratio of signal energy to noise energy in a sliding time window. The local signal-to-noise ratio evaluation can effectively distinguish between weak real temperature fluctuations and residual broadband noise, and avoid misjudging effective temperature dynamic changes as noise and filtering them out. The calculation result of the local signal-to-noise ratio is compared with an empirical threshold value, and the frequency band lower than the threshold value is still regarded as a noise component that needs to be suppressed.

[0069] A noise template is constructed using the statistical characteristics of the historical temperature data of the furnace, and the principal component analysis method is used to extract common periodic interference patterns from the historical data in the construction process of the noise template. The noise template contains regular noise characteristics caused by factors such as combustion oscillation and fan vibration during the operation of the boiler. Through template matching, periodic interference components are further suppressed, and the detail coefficients of the current signal are correlated with the noise template, and the periodic components with high matching degree are selectively attenuated. The adaptive filtering process is iterative, and the residual signal after each filtering is analyzed again to check whether there are still obvious noise components remaining. The number of iterations is controlled according to the preset convergence condition, and the convergence condition can be that the energy change rate of the residual signal is lower than a certain minimum value, or the maximum allowed number of iterations is reached. The filtering parameters are updated according to the statistical characteristics of the current signal in each iteration, achieving dynamic adaptation.

[0070] ​The processing flow of the noise suppression module is integrated on a dedicated real-time signal processor, and the algorithm is optimized to reduce the computational complexity to ensure that it can process high-frequency sampled temperature data streams. Memory management uses a block processing strategy to divide large temperature data sets into small blocks for sequential processing, reducing the instantaneous demand for memory resources. The historical noise database has an online update mechanism. When the boiler is significantly modified or the fuel characteristics change significantly, the database can trigger a retraining process to ensure the accuracy of the noise model. The adaptive threshold training process of the energy entropy threshold method uses a support vector machine algorithm, and the training features include the energy entropy, variance, skewness, kurtosis, and other statistical quantities of each scale detail coefficient. The trained model parameters are stored in the non-volatile memory of the noise suppression module and loaded into the memory for real-time calculation when the system starts. Noise template matching calculation involves signal convolution operation, which converts time domain signal to frequency domain for matching through fast Fourier transform to improve calculation speed. The output of the noise suppression module is the denoised temperature data sequence, which maintains the same data structure and time-space alignment properties as the original temperature data set. The denoising effect is internally evaluated by comparing the energy spectrum density of the signals before and after filtering, and the evaluation results are recorded in the system log but do not affect the real-time data output. The module has a self-monitoring function that can detect boundary effects or reconstruction errors that may occur during wavelet decomposition and use methods such as symmetric extension to mitigate the effects of these effects. The entire noise suppression process aims to preserve the real thermodynamic process information in the boiler furnace while maximizing the removal of sensor electronic noise, transmission interference, and environmental background noise, providing a high-quality data foundation for subsequent temperature field reconstruction. The denoised temperature data sequence is sent to the temperature field reconstruction module for further processing, and the data transmission interface uses a high-throughput parallel bus protocol.

[0071] In the embodiment 4, the temperature field reconstruction module receives the denoised temperature data sequence from the noise suppression module, which contains spatial coordinate information and corresponding temperature measurement values. The denoised temperature data sequence is taken as the interpolation nodes of the radial basis function, and each data point constitutes a node in the interpolation network, with the node position determined by its three-dimensional coordinates in the unified coordinate system of the boiler furnace. The node weights are solved by the least squares method, which constructs an overdetermined equation set, and the solution of the equation set minimizes the sum of squares of the differences between the calculated values and the measured values of the reconstructed temperature field at the nodes. The Tikhonov regularization term is introduced in the solving process to improve the ill-conditioned problem and prevent the weight coefficient from having excessive numerical fluctuations. The Dirichlet boundary condition is introduced at the furnace wall boundary to constrain the interpolation process, which specifies the known temperature values of the furnace wall. The boundary temperature values are taken from the boiler design parameters, including the thermal conductivity of the furnace material, the design working temperature, and the design working condition of the cooling system. The application of the boundary condition is realized by adding additional constraint equations to the radial basis function interpolation equation set, ensuring that the reconstructed temperature field is consistent with the physical settings at the boundary.

[0072] A multi-layer grid method is used to accelerate the solution of large-scale radial basis function equation sets. The multi-layer grid method divides the calculation grid into different levels of grid series with different densities. The number of grid layers is automatically optimized according to the furnace volume, and more grid layers are used for boilers with larger furnace volumes to balance the calculation accuracy and efficiency. The solution process quickly captures the macro trend of the temperature field on the coarse grid and finely corrects the local features on the fine grid, improving the convergence speed through information transmission and correction between grids. The temperature field reconstruction module includes a model accuracy verification unit that compares the reconstructed three-dimensional temperature field distribution model with the adiabatic temperature calculated based on the furnace outlet flue gas composition. The furnace outlet flue gas composition is measured in real time by a gas analyzer installed on the flue, and the adiabatic temperature is calculated based on the heat balance of the combustion reaction. The comparison process calculates the deviation between the average temperature of the entire three-dimensional temperature field distribution model at the furnace outlet area and the calculated adiabatic temperature. If the deviation exceeds the allowed range, the model accuracy verification unit automatically adjusts the shape parameter of the radial basis function, which controls the width of the radial basis function and affects the smoothness of the interpolation result. The adjustment process finds the shape parameter value that minimizes the deviation based on an optimization algorithm, which can use the gradient descent method or the simplex method. The model accuracy verification unit triggers the re-collection of local sensor data, and the re-collection instruction is sent to the temperature collection module, giving priority to the infrared temperature sensor and the thermocouple sensor corresponding to the area with larger deviation.

[0073] The radial basis function interpolation algorithm of the temperature field reconstruction module uses a Gaussian function as the kernel function, and the width parameter of the Gaussian function is associated with the average distance of the data points. The linear equation system generated by the interpolation calculation is solved iteratively using the preconditioned conjugate gradient method, and the preconditioning matrix is constructed according to the eigenvalue distribution of the coefficient matrix to improve the convergence. The three-dimensional temperature field distribution model is output in the form of a voxel grid, each voxel contains the coordinates of the center point and the reconstructed temperature value, and the grid resolution can be configured to meet different application requirements. The allowed range threshold of the model accuracy verification unit is dynamically set according to the type of the boiler, the type of fuel and the operating standard, and the threshold is usually expressed as a percentage interval of the inverse calculated adiabatic temperature. The historical record of the comparison results is stored in the system database for analyzing the long-term accuracy trend of the temperature field reconstruction model, and the data record contains information such as timestamp, deviation value, adjusted shape parameter, etc. During the local sensor data reacquisition process, the temperature acquisition module temporarily increases the sampling frequency of the sensors in the specified area to obtain more intensive temperature data for local model correction.

[0074] The calculation task of the temperature field reconstruction module is deployed on a high-performance computing node, and the computing node is equipped with large-capacity memory to store the large coefficient matrix of the interpolation equation system. The module provides an application programming interface for the boiler control system to call, and the control system can request the temperature distribution of a specific section or the temperature time series of a certain point in space. The parameter configuration of the radial basis function interpolation algorithm, including the type of basis function, regularization coefficient, convergence tolerance, etc., is stored in a configuration file, allowing offline adjustment. The calculation period of the model accuracy verification unit is associated with the operating condition of the boiler, and the verification frequency is increased during rapid load changes and reduced during stable operating conditions. The reacquired sensor data is preprocessed by the data fusion module and the noise suppression module, and then input into the temperature field reconstruction module for local reconstruction, and the local reconstruction result is integrated with the global temperature field through a data fusion algorithm. The temperature field reconstruction module has a running state self-checking function, which periodically checks the memory usage, the completion status of the calculation task, and the communication connection with other modules. Referring to Table 1, the typical configuration range of the key parameters of the radial basis function interpolation algorithm in the temperature field reconstruction module is shown, which affects the accuracy and computational efficiency of the reconstruction model:

[0075] Table 1: Parameter configuration of radial basis function interpolation algorithm

[0076]

[0077] The bias calculation of the model accuracy verification unit adopts the weighted average method, and the temperature bias at different positions in the furnace outlet area is given different weights according to its flow characteristics. The weight coefficient is determined based on the residence time distribution of the flue gas obtained by computational fluid dynamics simulation, and the weight of the area with long residence time is higher. The output data format of the temperature field reconstruction module supports multiple industry standards, such as VTK format for scientific visualization and HDF5 format for large data storage and analysis. The model accuracy verification unit is integrated with the boiler combustion optimization system, and when persistent large deviations are detected, not only are the reconstruction model parameters adjusted, but also calibration suggestions are sent to the optimization system. The calculation process of the temperature field reconstruction module is optimized for real-time requirements, and the processing delay is reduced through algorithm parallelization, memory pre-allocation, and calculation pipeline technology. The configuration interface of the module allows users to adjust parameters according to the specific characteristics of the boiler, and the interface provides parameter sensitivity analysis to help users understand the impact of parameter changes on the reconstruction results. The entire temperature field reconstruction process aims to generate a three-dimensional temperature field distribution that is physically reasonable and numerically accurate, providing spatial temperature information for boiler combustion state monitoring and optimization. The reconstruction results are transmitted to the monitoring system through an industrial network and displayed in real time to the operators in the form of three-dimensional cloud maps, isothermal surfaces, temperature profiles, and other forms.

[0078] Example 5: The dynamic compensation module receives a three-dimensional temperature field distribution model from the temperature field reconstruction module, which is a static temperature field reconstructed based on current sensor data. The flue gas velocity field distribution inside the furnace is obtained through computational fluid dynamics simulation, which is based on the accurate three-dimensional geometric model of the boiler furnace, the current burner operation state, the fuel characteristic parameters, and the boundary condition settings. The computational fluid dynamics simulation runs on a dedicated high-performance computing cluster, and the simulation software solves the Navier-Stokes equation and the energy equation to simulate the flow and heat transfer process of the flue gas inside the furnace. The simulation calculation generates physical field data such as flue gas velocity vector, pressure, temperature, etc. for each spatial grid node inside the furnace. The local flow velocity vector of each spatial grid is extracted from the computational fluid dynamics simulation results, which is a three-dimensional vector containing speed and direction information. The spatial grid division is consistent with the grid system used by the temperature field reconstruction module, ensuring that the data can be directly corresponded. The local flow velocity vector data is stored in the memory array of the dynamic compensation module, and the array index corresponds to the grid coordinates one by one. Combined with the burner operation state and fuel characteristic parameters, a mapping relationship between temperature propagation delay time and spatial position is established. The burner operation state information comes from the boiler control system, including which burners are in operation, the fuel quantity of each, the air volume, etc. The fuel characteristic parameters include the low calorific value, composition analysis, density, specific heat capacity, and other physical and chemical properties of the fuel. The temperature propagation delay time represents the time required for temperature information to propagate from the furnace combustion area to the sensor installation location, and the delay time is inversely proportional to the local flue gas flow rate. The mapping relationship is established through an empirical model, with the local flow rate, fuel characteristics, and burner position as inputs, and the delay time from each grid point to the upstream reference point as output.

[0079] The dynamic temperature field model is phase-corrected based on the delay time, and the phase-correction algorithm time-shifts the static three-dimensional temperature field distribution model along the flue gas flow direction. For each spatial position inside the furnace, the temperature value at an earlier time for the corresponding upstream point is found according to its delay time, and the temperature value at the earlier time is used to replace the current measurement value. The phase-correction process compensates for the temperature measurement lag effect caused by thermal inertia, making the dynamic temperature field model closer to the real instantaneous combustion state. The heat conduction lag compensation calculation process of the dynamic compensation module is periodically updated, and the update strategy is dynamically adjusted according to the boiler operating conditions. The burner load change rate is monitored in real time, which is obtained by calculating the change amplitude of the fuel quantity per unit time. When the change rate exceeds the critical value, a new round of computational fluid dynamics simulation is triggered to update the flue gas velocity field distribution. The critical value is set according to the dynamic characteristics of the boiler, and usually corresponds to the working condition of large fluctuations in load. In steady-state conditions, the burner load change rate is below the critical value, and the update strategy updates the mapping relationship regularly according to the pre-set time interval, which is determined according to the thermal inertia time constant of the boiler, for example, updated every 30 minutes.

[0080] The initialization and execution process of the computational fluid dynamics simulation is automated, and the simulation input parameters are automatically obtained from the real-time database. The simulation grid uses unstructured grids, and local encryption is performed near the burner nozzle and in the corner area to capture complex flow details. The simulation results are verified by comparing the furnace outlet parameters obtained by simulation with the measured data to ensure the reliability of the simulation results. The fuel characteristic parameters are in an online database, and when different coal types are blended in the coal-fired power plant, the database is automatically updated to the industrial analysis data of the current coal fed into the furnace. The delay time mapping relationship model is periodically retrained with historical operation data, so that the model can adapt to changes in the state of the boiler equipment. The phase correction algorithm is implemented using digital filtering technology, and the filter is designed as an all-pass filter that only delays the phase of the signal without changing the amplitude. The output of the dynamic compensation module is the compensated dynamic temperature field model, which outputs in the form of a time series, with each time step corresponding to a three-dimensional temperature field that has been compensated for lag. The dynamic temperature field model is sent to the boiler combustion optimization system through a high-speed data interface for real-time adjustment of control parameters such as air-coal ratio and burner inclination. The module has a cache mechanism inside to store historical temperature field data and velocity field data for a certain period of time for backtracking analysis and model verification.

[0081] The critical value of the update strategy is not fixed, but is self-adaptively adjusted according to factors such as seasonal changes and equipment aging. The adjustment algorithm analyzes the relationship between historical load change data and the dynamic response characteristics of the boiler to optimize the setting of the critical value. When triggering a new round of computational fluid dynamics simulation, the system will evaluate the current occupation of computing resources, and if the resources are tight, it will appropriately reduce the simulation grid resolution or use a simplified model to ensure the real-time performance of the calculation. There is bidirectional communication between the dynamic compensation module and the boiler main control system, with the main control system providing real-time operation instructions to the dynamic compensation module, and the dynamic compensation module feeding back the dynamic characteristics of the temperature field to the main control system. The module has perfect fault diagnosis and processing logic, and when the computational fluid dynamics simulation does not converge or the data is abnormal, it will automatically switch to a backup simplified lag model and issue a maintenance alert.

[0082] It should be noted that, in the present document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0083] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A boiler furnace temperature field measurement system based on multi-sensor fusion, characterized in that, include: The temperature acquisition module is used to arrange multiple sets of infrared temperature sensors and thermocouple sensors in a preset spatial grid within the boiler furnace to synchronously acquire real-time temperature data of each spatial grid within the furnace. The data fusion module is used to perform time alignment and spatial registration of real-time temperature data collected by infrared temperature sensors and thermocouple sensors to generate a spatiotemporally synchronized raw temperature dataset. The noise suppression module is used to perform multi-scale decomposition of the original temperature dataset based on wavelet transform, extract high-frequency noise components and perform adaptive filtering, and output the noise-reduced temperature data sequence. The temperature field reconstruction module is used to reconstruct the three-dimensional temperature field distribution model of the furnace based on the denoised temperature data sequence using the radial basis function interpolation algorithm. Specifically, it includes using the denoised temperature data sequence as the interpolation nodes of the radial basis function, and solving the node weights using the least squares method. A Dirichlet boundary condition constraint interpolation process is introduced at the boundary of the furnace wall, and the boundary temperature value is taken from the boiler design parameters. A multi-layer grid method is used to accelerate the solution of large-scale radial basis function equations, and the number of grid layers is automatically optimized according to the furnace volume. The dynamic compensation module is used to calculate the heat conduction lag compensation of the temperature field distribution model based on the flue gas flow velocity in the furnace and the burner operating parameters, and to generate the compensated dynamic temperature field model. The dynamic compensation module calculates the heat conduction hysteresis compensation amount as follows: The velocity field distribution of flue gas inside the furnace is obtained through computational fluid dynamics simulation, and the local velocity vector of each spatial grid is extracted. By combining the burner's operating status and fuel characteristic parameters, a mapping relationship between temperature propagation delay time and spatial location is established; Phase correction is performed on the dynamic temperature field model based on the delay time to compensate for the temperature measurement hysteresis effect caused by thermal inertia. The calculation process for the heat conduction hysteresis compensation of the dynamic compensation module is periodically updated, and its update strategy is as follows: Real-time monitoring of burner load change rate; when the change rate exceeds the critical value, a new round of computational fluid dynamics simulation is triggered to update the flue gas velocity field distribution. Under steady-state conditions, the mapping relationship is updated periodically according to a preset time interval.

2. The boiler furnace temperature field measurement system based on multi-sensor fusion according to claim 1, characterized in that, The specific method for arranging multiple sets of infrared temperature sensors and thermocouple sensors in the temperature acquisition module is as follows: Based on the geometry of the boiler furnace, it is divided into several layers of annular monitoring areas, and at least three infrared temperature sensors are evenly deployed in each annular monitoring area along the circumference. A thermocouple sensor chain is vertically installed at the central axis position of each layer of the annular monitoring area. The thermocouple sensor chain consists of multiple thermocouple probes that are evenly distributed. The sampling frequency of the infrared temperature sensor and the thermocouple sensor is dynamically adjusted according to the boiler combustion load. When the load is higher than the threshold, the first sampling frequency is used, and when the load is lower than the threshold, the second sampling frequency is used.

3. The boiler furnace temperature field measurement system based on multi-sensor fusion according to claim 2, characterized in that, The specific process of generating the spatiotemporally synchronized original temperature dataset in the data fusion module is as follows: The temperature data collected by the infrared temperature sensor is time-stamped and mapped to the unified coordinate system of the furnace through spatial coordinate transformation. Linear interpolation is performed on the temperature data collected by the thermocouple sensor chain to generate a continuous temperature profile that matches the spatial resolution of the infrared temperature sensor. A sliding time window algorithm is used to align the data acquisition time points of the two types of sensors, and the window length is adaptively adjusted according to the boiler combustion fluctuation cycle.

4. The boiler furnace temperature field measurement system based on multi-sensor fusion according to claim 3, characterized in that, The specific steps for performing adaptive filtering in the noise suppression module include: Discrete wavelet decomposition was performed on the original temperature dataset to obtain approximation coefficients and detail coefficients at different scales; Based on the energy entropy thresholding method, noise-dominant frequency bands in detail coefficients are identified, and soft thresholding is applied to the coefficients of this frequency band. A noise template is constructed using the statistical characteristics of historical furnace temperature data, and periodic interference components are further suppressed through template matching.

5. A boiler furnace temperature field measurement system based on multi-sensor fusion according to claim 1, characterized in that, The data fusion module is also equipped with an online sensor calibration unit. A dynamic error correction model is established based on the reading deviation between infrared temperature sensors and thermocouple sensors within the same spatial grid under stable operating conditions. When the reading of any sensor suddenly exceeds the preset confidence interval, the cross-validation and smooth transition algorithm based on the data of neighboring sensors is activated.

6. A boiler furnace temperature field measurement system based on multi-sensor fusion according to claim 4, characterized in that, The threshold determination method of the energy entropy threshold method is as follows: Calculate the energy entropy of detail coefficients at each scale and compare it with an adaptive threshold trained based on a historical noise database; For suspected effective signal frequency bands with energy entropy below the threshold, a local signal-to-noise ratio assessment is introduced for secondary discrimination to avoid the false filtering of effective temperature fluctuations.

7. A boiler furnace temperature field measurement system based on multi-sensor fusion according to claim 1, characterized in that, The temperature field reconstruction module also includes a model accuracy verification unit. Used to compare the consistency between the reconstructed three-dimensional temperature field distribution model and the adiabatic temperature calculated from the composition of the flue gas at the furnace outlet; If the deviation exceeds the allowable range, the shape parameters of the radial basis function will be automatically adjusted, and the local sensor data will be reacquired.

Images