A multi-view-based furnace cross-section temperature field reconstruction method and system

Abstract

This invention provides a multi-viewpoint furnace cross-sectional temperature field reconstruction method and system, belonging to the field of boiler combustion monitoring technology. It can significantly alleviate or solve the problems of existing single-viewpoint measurements, such as difficulty in resolving the temperature distribution along the furnace depth, large measurement errors caused by absorption and scattering by flue gas and fly ash, and insufficient accuracy and real-time performance in temperature field reconstruction. The method includes: synchronously acquiring dual-band radiation signals from multiple viewpoints; calculating the colorimetric apparent temperature field as the initial temperature field; acquiring flue gas composition and fly ash concentration in real time to establish a media radiation characteristic model; reconstructing the furnace cross-sectional temperature field based on the radiative transfer equation using an iterative algorithm; and simultaneously correcting the emissivity distribution. This invention has the advantages of strong anti-interference capability, high reconstruction accuracy, good real-time performance, high equipment reliability, and non-contact measurement, and is suitable for combustion monitoring and optimized control of industrial boilers such as those used in thermal power generation and waste incineration.

Description

Technical Field

[0001] This invention belongs to the field of boiler combustion monitoring technology, specifically relating to a method and system for reconstructing the furnace cross-sectional temperature field based on multiple viewpoints. Background Technology

[0002] Accurate sensing and real-time characterization of the combustion state inside the boiler furnace has always been a key technical issue in the operation and control of thermal power generation, waste incineration, and various industrial boilers. The furnace temperature field not only directly reflects the completeness of fuel combustion and the distribution of the flame center, but is also closely related to boiler thermal efficiency, heat transfer effect of heating surfaces, pollutant generation level, and long-term safe and stable operation of the equipment. Therefore, how to continuously, accurately, and online acquire the temperature distribution inside the furnace has become an important research direction in the fields of boiler combustion optimization control, energy saving and consumption reduction, and fault early warning.

[0003] In existing engineering applications, furnace temperature measurement has long relied primarily on contact-based temperature measurement methods such as thermocouples. While these methods are relatively simple in structure and have a long history of application, they essentially only provide localized temperature information near the installation point, failing to reflect the temperature distribution across the entire furnace cross-section or even the entire spatial area. This makes them unsuitable for the field-oriented, distributed, and real-time measurement requirements of modern boiler combustion diagnostics. Especially for large-capacity coal-fired boilers, waste incinerators, and high-temperature industrial furnaces, the furnace environment is typically characterized by complex conditions such as high temperature, high dust, high corrosion, and strong disturbances. Over long-term use, thermocouples are prone to erosion, failure, and drift, resulting in short service lives, high maintenance and replacement costs, and increased downtime for repairs and safety precautions. Furthermore, thermocouples exhibit significant thermal inertia, leading to a relatively slow response time and making it difficult to promptly capture combustion fluctuations and transient temperature changes within the furnace.

[0004] With the development of non-contact temperature measurement technology, optical temperature measurement has gradually become an important technical route for high-temperature furnace measurement. Among them, colorimetric temperature measurement methods, based on the relationship between the ratio of radiation intensity at different wavelengths and temperature, have advantages such as low sensitivity to the absolute value of emissivity and strong anti-interference ability, thus showing good application prospects in high-temperature field measurement. However, existing furnace temperature field measurement schemes based on colorimetric methods still have significant limitations. First, single-viewpoint measurement methods can usually only obtain integrated radiation information along the line of sight, essentially reflecting the projection result, making it difficult to distinguish the true temperature distribution in the depth direction. Therefore, the results are often only two-dimensional projected temperature fields, which cannot accurately characterize the true temperature structure of the furnace cross-section. Second, boiler furnaces generally contain high-temperature flue gas, fly ash particles, and other dispersed media, which absorb and scatter radiation signals, causing the measured signal to deviate from the ideal radiation model and affecting the temperature measurement accuracy. Third, existing temperature field reconstruction algorithms are insufficiently adaptable to complex furnace boundary conditions and media radiation characteristics, often suffering from problems such as large computational load, low reconstruction accuracy, and poor real-time performance, making it difficult to simultaneously meet the needs of online monitoring and engineering feasibility.

[0005] To address this, a method and system for reconstructing the temperature field of the furnace cross section based on multiple viewpoints are proposed. Summary of the Invention

[0006] The present invention aims to solve at least one of the technical problems existing in the prior art, and to provide a method and system for reconstructing the temperature field of the furnace cross section based on multiple viewpoints.

[0007] This invention provides a method for reconstructing the temperature field of a furnace cross-section based on multiple viewpoints, comprising the following steps: S1: Calibrate N colorimetric temperature probes arranged circumferentially along the cross-section of the boiler furnace, where N≥3, and establish a discrete grid model of the furnace cross-section to determine the geometric projection relationship between the line of sight of each colorimetric temperature probe and the discrete grid unit. S2: Control N colorimetric temperature probes to collect radiation images of the furnace interior at the first wavelength λ1 and the second wavelength λ2 in the line of sight of each colorimetric temperature probe, thereby obtaining dual-band radiation images of each colorimetric temperature probe at the first wavelength λ1 and the second wavelength λ2. S3: Based on the calibration results obtained in step S1, the dual-band radiation image is converted into spectral radiation intensity, and the apparent temperature distribution along the line of sight of each colorimetric temperature probe is calculated based on the ratio of the spectral radiation intensity at the first wavelength λ1 and the second wavelength λ2, as the initial temperature field for reconstructing the temperature field of the furnace cross section. S4: Obtain the flue gas composition and fly ash concentration data in the furnace, calculate the absorption coefficient and scattering coefficient of the furnace medium at the first wavelength λ1 and the second wavelength λ2, and establish a medium radiation characteristic model of the furnace cross section. S5: Based on the apparent temperature distribution, the medium radiation characteristic model, and the spectral radiation intensity measured in the line of sight of each of the colorimetric temperature probes, establish a radiation transfer reconstruction model between the measured spectral radiation intensity and the furnace cross-sectional temperature field to be determined, and obtain the furnace cross-sectional temperature field. S6: Based on the preset initial emissivity value, the furnace cross-sectional temperature field, and the spectral radiation intensity, iteratively invert the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2, and correct the furnace cross-sectional temperature field according to the dual-band emissivity ratio constraint, and output the corrected furnace cross-sectional temperature field.

[0008] Further, in step S1, the calibration of the colorimetric temperature probe includes: using a blackbody furnace to perform dual-band radiation intensity calibration on each of the colorimetric temperature probes, establishing a quantitative relationship between the grayscale values ​​of the images acquired by each of the colorimetric temperature probes at the first wavelength λ1 and the second wavelength λ2 and the radiation intensity; determining the optical center position, optical axis direction, focal length and distortion coefficient of each of the colorimetric temperature probes through a standard target, and dividing the cross-section of the furnace to be measured into multiple grid units.

[0009] Specifically, in step S3, after converting the dual-band radiation information into spectral radiation intensity, the spectral radiation intensity is further preprocessed, and the preprocessing includes at least median filtering and bad pixel correction.

[0010] Specifically, in step S4, obtaining data on flue gas composition and fly ash concentration in the furnace includes obtaining data on CO2, H2O, O2 concentrations and fly ash concentrations in the furnace through a flue gas analyzer.

[0011] Preferably, in step S4, the average temperature of the furnace medium along a preset path and the average concentration of at least one flue gas component along the path are measured using laser absorption spectroscopy or tuned diode laser absorption spectroscopy, as boundary conditions or verification data for radiative transfer calculation.

[0012] Specifically, in step S5, obtaining the furnace cross-sectional temperature field includes: discretizing the radiative transfer reconstruction model using the discrete coordinate method or the finite volume method, establishing a linearized reconstruction equation set, and solving it using algebraic reconstruction technology, synchronous iterative reconstruction technology, or an iterative reconstruction algorithm based on total variation regularization; in each iteration, based on the current iterative temperature field, calculating the predicted radiation intensity along the observation path of each colorimetric temperature probe along the line of sight of the furnace cross-section, comparing the predicted radiation intensity with the measured furnace radiation intensity to obtain the residual, updating the furnace cross-sectional temperature field according to the residual; applying upper and lower temperature physical constraints to the updated furnace cross-sectional temperature field, and stopping the iteration when the residual is less than a preset threshold or the maximum number of iterations is reached.

[0013] Another aspect of the present invention provides a furnace cross-sectional temperature field reconstruction system based on multiple viewpoints, comprising: A multi-viewpoint colorimetric thermography imaging array includes N colorimetric thermography probes arranged circumferentially along the cross-section of the boiler furnace, wherein N≥3, for collecting radiation signals at the first wavelength λ1 and the second wavelength λ2 inside the furnace. A synchronous triggering and control unit is used to control the N colorimetric temperature probes to collect radiation signals; A data transmission network is used to transmit the radiation signals collected by each of the colorimetric temperature probes; The temperature field reconstruction calculation module is used to convert the radiation signal into spectral radiance based on the quantitative relationship between the image grayscale value and radiance intensity established through calibration; to calculate the apparent temperature distribution based on the ratio of spectral radiance intensity at the first wavelength λ1 and the second wavelength λ2 as the initial value for temperature field reconstruction; to establish a medium radiation characteristic model based on the flue gas composition and fly ash concentration data in the furnace, and to solve the temperature field of the furnace cross-section; and to correct the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2 based on the reconstructed temperature field of the furnace cross-section; and... The output and diagnostic unit is used to output the distribution of the temperature field in the furnace cross section and the diagnostic results.

[0014] Furthermore, each of the colorimetric temperature probes includes an optical imaging unit, a dual-channel photodetector, and an emissivity adjustment unit for adjusting the emissivity setpoint.

[0015] Furthermore, the optical imaging unit includes a dual-band narrowband filter and an optical lens. The center wavelength of the dual-band narrowband filter consists of two different wavelengths located in the range of 0.8 to 1.8 μm, and the bandwidth is no greater than 10 nm. The first wavelength λ1 and the second wavelength λ2 are the two center wavelengths of the dual-band narrowband filter. The dual-channel photodetector receives radiation signals at the first wavelength λ1 and the second wavelength λ2 respectively and converts them into electrical signals.

[0016] Specifically, the temperature field reconstruction calculation module includes: The signal preprocessing unit is used to denoise, amplify, and perform analog-to-digital conversion on the dual-channel radiation electrical signals inside the furnace at the first wavelength λ1 and the second wavelength λ2 collected by each of the colorimetric temperature probes, and convert the analog-to-digital converted signals into spectral radiation intensity that characterizes the radiation characteristics of the furnace according to the calibration coefficients. The dielectric radiation characteristics calculation unit is used to calculate the absorption coefficient and scattering coefficient at the first wavelength λ1 and the second wavelength λ2 based on the composition of the flue gas in the furnace and the concentration of fly ash. Temperature field reconstruction unit, used to solve the temperature distribution of the furnace cross section based on the radiative transfer equation and iterative reconstruction algorithm; and The emissivity distribution correction unit is used to invert the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2 based on the preset initial emissivity value, the furnace cross-sectional temperature field output by the temperature field reconstruction unit, and the dual-band spectral radiation intensity, and to correct the furnace cross-sectional temperature field using the dual-band emissivity ratio constraint.

[0017] The beneficial effects of this invention are as follows: This invention acquires multi-directional dual-band radiation information of the furnace cross-section through multi-viewpoint colorimetric thermography imaging, overcoming the problem that single-viewpoint measurement cannot reflect the true temperature distribution of the cross-section. By introducing flue gas composition and fly ash concentration data to establish a medium radiation characteristic model, and combining it with the radiation transfer equation and iterative reconstruction algorithm, the absorption and scattering effects of high-temperature flue gas and particulate media on radiation transfer can be effectively corrected, thereby achieving high-precision, real-time online reconstruction of the furnace cross-section temperature field. At the same time, this invention adopts a dual-band colorimetric thermography and emissivity synchronous correction mechanism, which has the advantages of strong anti-interference ability and high measurement accuracy. In addition, this invention uses a non-contact temperature probe with integrated high-temperature protection and protection mechanisms, which has high equipment reliability and low maintenance, and is suitable for combustion monitoring and optimization control of industrial boilers such as thermal power generation and waste incineration. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the steps of a multi-viewpoint furnace cross-sectional temperature field reconstruction method according to a specific embodiment of the present invention. Figure 2 This is a schematic diagram of the overall structure of a multi-view furnace cross-sectional temperature field reconstruction system according to a specific embodiment of the present invention. Figure 3 The flowchart shows the temperature field reconstruction algorithm of a furnace cross-section temperature field reconstruction method based on multiple viewpoints, which is a specific embodiment of the present invention. Detailed Implementation

[0019] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] like Figure 1 As shown in the figure, a furnace cross-section temperature field reconstruction method based on multiple viewpoints provided by a specific embodiment of the present invention includes the following steps: S1: Calibrate N colorimetric temperature probes arranged circumferentially along the cross-section of the boiler furnace, where N≥3, and establish a discrete grid model of the furnace cross-section to determine the geometric projection relationship between the line of sight of each colorimetric temperature probe and the discrete grid unit. S2: Control N colorimetric temperature probes to simultaneously acquire radiation images of the furnace interior at the first wavelength λ1 and the second wavelength λ2 along the line of sight of each colorimetric temperature probe, and obtain dual-band radiation images of each colorimetric temperature probe at the first wavelength λ1 and the second wavelength λ2. S3: Based on the calibration results obtained in step S1, the dual-band radiation information is converted into spectral radiation intensity, and the apparent temperature distribution in the line of sight of each colorimetric temperature probe is calculated according to the ratio of spectral radiation intensity at the first wavelength λ1 and the second wavelength λ2, which serves as the initial temperature field for the reconstruction of the furnace cross-section temperature field. S4: Obtain data on flue gas composition and fly ash concentration in the furnace, calculate the absorption coefficient and scattering coefficient of the furnace medium at the first wavelength λ1 and the second wavelength λ2, and establish a model of the medium radiation characteristics of the furnace cross section; the model should at least include the absorption coefficient and scattering coefficient, and if necessary, also include the extinction coefficient, scattering phase function and their relationship with flue gas composition / fly ash concentration / wavelength; S5: Based on the apparent temperature distribution, the medium radiation characteristic model, and the spectral radiation intensity measured in the line of sight of each colorimetric temperature probe, a radiation transfer reconstruction model between the measured spectral radiation intensity and the temperature field of the furnace cross section to be determined is established, and the temperature field of the furnace cross section is obtained by iterative solution. S6: Based on the preset initial emissivity value, furnace cross-sectional temperature field, and spectral radiation intensity, the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2 is iteratively inverted, and the furnace cross-sectional temperature field is corrected according to the dual-band emissivity ratio constraint, outputting the corrected furnace cross-sectional temperature field; the initial emissivity value is the initial setting value of the system; the emissivity distribution is the spatial distribution obtained by iterative inversion; the dual-band emissivity ratio constraint is the constraint condition used to correct the temperature field; the radiation propagation path through the furnace cross-section along the line of sight of each probe; the predicted radiation intensity is the radiation intensity calculated forward based on the current iterative temperature field and medium parameters.

[0021] In one embodiment, N colorimetric temperature probes are arranged circumferentially along the boiler furnace cross-section, preferably evenly, to form multiple different observation perspectives. Each colorimetric temperature probe is synchronously triggered and controlled in a unified timing sequence with the control unit, thereby acquiring dual-band radiation images of different directions inside the furnace at the same time, reducing the impact of combustion fluctuations on the consistency of multi-viewpoint data. The value of N is not less than 3, preferably 4, 6, or 8, to balance cross-sectional coverage, installation feasibility, and reconstruction accuracy.

[0022] Specifically, firstly, the radiation intensity of each colorimetric temperature probe at two wavelengths was calibrated using a high-precision blackbody furnace, and the geometric and distortion parameters of each probe were determined in conjunction with a standard target. Subsequently, dual-band radiation images inside the furnace were acquired under synchronous triggering conditions, and the apparent temperature distribution along each line of sight was calculated after preprocessing. Then, a radiation transmission reconstruction model was established by combining the flue gas composition, fly ash concentration, and furnace medium radiation characteristics, and the temperature field of the furnace cross section was iteratively solved. Finally, the reconstruction results were corrected based on the dual-band emissivity constraint, and the furnace cross section temperature field distribution map and statistical results were output.

[0023] Furthermore, by combining multi-viewpoint observation with dual-band colorimetric thermometry, the limitation of single-viewpoint measurement in obtaining only line-of-sight integral information can be overcome, and the influence of absolute emissivity changes on temperature measurement results can be reduced to a certain extent. At the same time, by combining the absorption and scattering correction of radiation transmission by flue gas and fly ash media in the furnace, the authenticity and accuracy of the temperature field reconstruction of the furnace cross section can be improved.

[0024] Furthermore, the corrected furnace cross-sectional temperature field can be mapped to a preset cross-sectional grid to generate a pseudo-color temperature distribution map, isotherm map, and statistical parameters such as cross-sectional average temperature, maximum temperature, and temperature non-uniformity coefficient. When the temperature exceeds a set threshold or the temperature distribution is abnormal, an alarm signal is output to serve boiler combustion monitoring, combustion optimization control, and operational safety early warning.

[0025] Based on the above basic implementation method, the calibration of the colorimetric temperature probes in step S1 includes: using a high-precision blackbody furnace to perform dual-band radiation intensity calibration on each colorimetric temperature probe, establishing a quantitative relationship between the image grayscale value and radiation intensity acquired by each colorimetric temperature probe at the first wavelength λ1 and the second wavelength λ2; determining the optical center position, optical axis direction, focal length and distortion coefficient of each colorimetric temperature probe through a standard target; and dividing the cross-section of the furnace chamber to be tested into multiple grid units.

[0026] Furthermore, the determination of geometric projection relationships includes: establishing a mapping relationship between the line of sight of each probe and the discrete grid unit of the furnace cross section based on the optical center position, optical axis direction, focal length and distortion coefficient of each colorimetric temperature probe, so that the subsequently acquired dual-band image data can be spatially correlated with each grid unit of the furnace cross section, providing a geometric basis for radiation transfer modeling and fault reconstruction.

[0027] In one specific embodiment, after converting the dual-band radiation information into spectral radiation intensity in step S3, the method further includes preprocessing the spectral radiation intensity, which includes at least median filtering and bad pixel correction; and calculating the apparent temperature distribution along the line of sight of each colorimetric temperature probe based on the ratio of the spectral radiation intensity at the first wavelength λ1 and the second wavelength λ2, including calculation using colorimetric temperature measurement under the gray body assumption.

[0028] In this embodiment, the calibrated and converted spectral radiation intensity image is first subjected to median filtering and bad pixel correction to suppress random noise and reduce the influence of detector defects on the measurement results. Subsequently, under the gray body assumption, the apparent temperature distribution in the line-of-sight direction of each colorimetric temperature probe is calculated using the ratio of the spectral radiation intensity at the first wavelength to the second wavelength, and this apparent temperature distribution is used as the initial temperature field for subsequent iterative reconstruction.

[0029] Furthermore, using the apparent temperature distribution as the initial value for iteration helps to shorten the iteration convergence time of the radiative transfer reconstruction model and improve the stability and real-time performance of solving the temperature field of the furnace cross section.

[0030] In another specific embodiment, in step S4, data on flue gas composition and fly ash concentration in the furnace are acquired, including real-time acquisition of CO2, H2O, O2 concentrations and fly ash concentration data in the furnace using a flue gas analyzer; the absorption coefficient and scattering coefficient at the first wavelength λ1 and the second wavelength λ2 are calculated, including calculations based on the HITRAN spectral database and Mie scattering theory.

[0031] In this embodiment, the radiation characteristics of the medium within the furnace are jointly determined by the absorption characteristics of the flue gas components and the scattering characteristics of fly ash particles. The flue gas components include at least one or more of CO2, H2O, and O2, and the fly ash concentration is used to characterize the scattering and attenuation effects of the dispersed particulate medium in the furnace on radiation transmission. By combining the above medium parameters with a discrete grid of the furnace cross-section, a distribution model of the medium's radiation characteristics suitable for temperature field reconstruction can be established.

[0032] Specifically, the HITRAN spectral database is used to characterize the spectral absorption characteristics of different flue gas components at the first and second wavelengths, and the Mie scattering theory is used to characterize the scattering behavior of fly ash particles to radiation. Based on this, the absorption coefficient and scattering coefficient of the furnace medium at the two wavelengths can be calculated and used as the medium parameter input in the radiative transfer equation.

[0033] Furthermore, the furnace medium radiation characteristic model can be updated with real-time data acquired by the flue gas analyzer, thereby reflecting the dynamic changes in the furnace medium radiation characteristics under different loads, different fuel conditions, or different combustion states.

[0034] In another specific embodiment, step S4 further includes measuring the average temperature of the furnace medium along a preset path and the average concentration of at least one flue gas component in real time using laser absorption spectroscopy or tuned diode laser absorption spectroscopy, as boundary conditions or verification data for radiative transfer calculation.

[0035] Furthermore, the path-average temperature along the preset path in the furnace and the path-average concentration of at least one flue gas component can be used as boundary conditions, parameter correction basis, or independent verification data in radiative transfer calculations to improve the reliability of the calculation results of medium radiation characteristic parameters and temperature field reconstruction results.

[0036] Specifically, the preset path can be set as a representative optical path that passes through the central area of ​​the furnace, the near-wall area, or the area above the burner, according to the boiler structure, the location of the observation hole, and the measurement requirements, so as to obtain the average temperature and component concentration information of the path that is representative of the main combustion area of ​​the furnace.

[0037] In another specific embodiment, in step S5, obtaining the furnace cross-sectional temperature field includes: discretizing the radiation transfer reconstruction model using the discrete coordinate method or the finite volume method, establishing a linearized reconstruction equation set, and solving it using algebraic reconstruction technology, synchronous iterative reconstruction technology, or an iterative reconstruction algorithm based on total variation regularization; in each iteration, based on the current iterative temperature field, calculating the predicted radiation intensity along the observation path of each colorimetric temperature probe along the line of sight of the furnace cross-section, and comparing the predicted radiation intensity with the measured furnace radiation intensity to obtain the residual, updating the furnace cross-sectional temperature field according to the residual; applying upper and lower temperature physical constraints to the updated furnace cross-sectional temperature field, and stopping the iteration when the residual is less than a preset threshold or the maximum number of iterations is reached; the upper and lower temperature physical constraints are used to limit the reconstructed temperature field from exceeding the allowable range of the actual operating conditions of the boiler furnace.

[0038] Furthermore, iterative reconstruction algorithms can employ ART, SIRT, or improved algorithms based on total variation regularization. Among these, improved algorithms based on total variation regularization introduce regularization constraints beyond the data fidelity term, helping to suppress noise amplification while preserving the boundary variation characteristics of the cross-sectional temperature field, thus improving the smoothness and stability of the reconstruction results. ART (Algebraic Reconstruction Technique) is a method that iteratively corrects the reconstruction results line by line using projection data. Its basic idea is to update the image pixel values ​​successively according to the projection equation, making the reconstruction results continuously approximate the actual measurement data. ART is simple to implement and has a fast initial convergence speed, making it particularly suitable for scenarios with incomplete projection data, limited sampling angles, or sparse sampling. However, it is sensitive to noise and projection order; without appropriate constraints, it can easily introduce fluctuations and artifacts into the reconstruction results. SIRT (Simultaneous Iterative Reconstruction Technique) integrates all projection errors in each iteration to perform a comprehensive correction of the image. Compared to ART, SIRT has a smoother update process, better robustness to noise, higher stability of reconstruction results, and can effectively suppress the amplification of local errors. Therefore, it is suitable for applications with high noise levels or high requirements for reconstruction uniformity. However, SIRT's convergence speed is usually slower than ART, often requiring more iterations to achieve the same accuracy. Improved algorithms based on Total Variation (TV) regularization introduce prior constraints into the traditional iterative reconstruction framework. By adding a TV penalty term in addition to the data consistency term, they suppress noise and stripe artifacts in the image while preserving edge and structural information as much as possible. This type of method is particularly suitable for underdetermined reconstruction scenarios such as low-dose, few-angle, and sparse sampling, significantly improving image quality while maintaining reconstruction accuracy. Compared to ART and SIRT, TV-based regularization methods generally have advantages in edge preservation and noise resistance, but their computational complexity is relatively high, and they are sensitive to the selection of regularization parameters, requiring adjustment and optimization based on the specific characteristics of the data.

[0039] Specifically, in each iteration, the predicted radiation intensity of the furnace cross section along the line of sight of each colorimetric temperature probe is first calculated based on the current iteration temperature field and medium radiation characteristic parameters. Then, the predicted radiation intensity is compared with the measured furnace radiation intensity to obtain the residual, and the furnace cross section temperature field is updated accordingly. After the update, physical constraints on the upper and lower temperature limits are further applied to avoid non-physical processes that exceed the actual combustion conditions of the boiler furnace.

[0040] In another specific embodiment, such as Figure 2 As shown, the present invention provides a furnace cross-sectional temperature field reconstruction system based on multiple viewpoints, comprising: The multi-viewpoint colorimetric temperature imaging array includes N colorimetric temperature probes arranged circumferentially along the cross-section of the boiler furnace, where N≥3, for acquiring radiation signals at the first wavelength λ1 and the second wavelength λ2 inside the furnace; a synchronous triggering and control unit for controlling the N colorimetric temperature probes to synchronously acquire radiation signals; a data transmission network for transmitting the radiation signals acquired by each colorimetric temperature probe; a temperature field reconstruction calculation module for converting the radiation signal into spectral radiation intensity based on the quantitative relationship between image grayscale values ​​and radiation intensity established by calibration; calculating the apparent temperature distribution based on the ratio of spectral radiation intensity at the first wavelength λ1 and the second wavelength λ2 as the initial value for temperature field reconstruction; establishing a medium radiation characteristic model based on flue gas composition and fly ash concentration data inside the furnace; solving the furnace cross-sectional temperature field based on the radiation transfer equation and iterative reconstruction algorithm; and correcting the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2 based on the reconstructed furnace cross-sectional temperature field; and an output and diagnostic unit for outputting the distribution of the furnace cross-sectional temperature field and diagnostic results.

[0041] Furthermore, the multi-viewpoint colorimetric temperature imaging array, synchronous triggering and control unit, high-speed data transmission network, temperature field reconstruction calculation module, and output and diagnostic unit are connected in sequence to form a complete link from multi-viewpoint radiation signal acquisition, real-time transmission, online calculation to result display and diagnostic alarm.

[0042] Furthermore, the output and diagnostic unit can be connected to the display terminal, alarm device and DCS control system so that the reconstructed furnace cross-sectional temperature field results can be used for boiler combustion status monitoring, operating parameter adjustment and abnormal operating condition linkage alarm.

[0043] In one specific implementation, each colorimetric temperature probe includes a high-temperature protection and installation unit, a protection mechanism unit, an optical imaging unit, a dual-channel photodetector, and an emissivity adjustment unit for adjusting the emissivity setpoint; the high-temperature protection and installation unit includes a high-temperature protective cover, a conduit, and a cooler; the protection mechanism unit includes a pneumatic valve and a control system for automatically shutting off when the cooling gas supply stops or the probe internal temperature exceeds the limit.

[0044] Specifically, the high-temperature protection and installation unit is fixedly connected to the external installation structure through the opening in the furnace wall, so that the colorimetric temperature probe can be stably arranged in the boiler observation hole or measurement hole position; the pneumatic valve and control system in the protection mechanism unit can automatically close the measurement channel when the cooling gas is interrupted or the internal temperature of the probe rises abnormally, thereby avoiding damage to the internal components of the probe by high-temperature flue gas and fly ash.

[0045] Furthermore, the emissivity adjustment unit can pre-set the initial emissivity value based on the material properties of the object being measured or empirical values ​​of the operating conditions, so as to provide a parameter basis for the initial measurement of the system and subsequent emissivity distribution correction.

[0046] In one specific embodiment, the optical imaging unit includes a dual-band narrowband filter and an optical lens. The center wavelength of the dual-band narrowband filter consists of two different wavelengths located in the range of 0.8 to 1.8 μm, and the bandwidth is no greater than 10 nm. The first wavelength λ1 and the second wavelength λ2 are the two center wavelengths of the dual-band narrowband filter. The dual-channel photodetector receives the radiation signals of the first wavelength λ1 and the second wavelength λ2 respectively and converts them into electrical signals.

[0047] In this embodiment, the dual-band narrowband filter works in conjunction with the dual-channel photodetector, enabling each colorimetric temperature probe to simultaneously acquire furnace radiation information at two specific wavelengths; by performing colorimetric temperature measurement using the ratio of radiation intensity at the two wavelengths, the influence of the uncertainty of the absolute value of emissivity on the temperature solution can be reduced.

[0048] Specifically, the first and second wavelengths can be selected based on the measured temperature range, the radiation characteristics of the target material, and the detector response characteristics, as long as they can meet the needs of dual-band colorimetric thermometry and subsequent furnace cross-sectional temperature field reconstruction.

[0049] In one specific implementation, the temperature field reconstruction calculation module includes: a signal preprocessing unit, used to denoise, amplify, and perform analog-to-digital conversion on the dual-channel radiation electrical signals inside the furnace at the first wavelength λ1 and the second wavelength λ2 collected by each colorimetric temperature probe, and convert the analog-to-digital converted signals into spectral radiation intensity characterizing the radiation properties of the furnace according to the calibration coefficients; a medium radiation characteristic calculation unit, used to calculate the absorption coefficient and scattering coefficient at the first wavelength λ1 and the second wavelength λ2 according to the composition of the flue gas and the fly ash concentration in the furnace; a temperature field reconstruction unit, used to solve the temperature distribution of the furnace cross section based on the radiation transfer equation and iterative reconstruction algorithm; and an emissivity distribution correction unit, used to invert the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2 based on the preset initial emissivity value, the furnace cross section temperature field output by the temperature field reconstruction unit, and the dual-band spectral radiation intensity, and to correct the furnace cross section temperature field using the dual-band emissivity ratio constraint.

[0050] In this embodiment, the signal preprocessing unit, the dielectric radiation characteristic calculation unit, the temperature field reconstruction unit, and the emissivity distribution correction unit can be integrated into an industrial computer, edge computing device, or data processing server to achieve real-time processing of multiple dual-band radiation signals and online reconstruction of the furnace cross-section temperature field.

[0051] Specifically, the signal preprocessing unit converts the dual-channel radiation electrical signals collected by each colorimetric temperature probe into spectral radiation intensity. The medium radiation characteristic calculation unit calculates the absorption coefficient and scattering coefficient based on the flue gas composition and fly ash concentration. The temperature field reconstruction unit combines the radiation transfer equation and iterative reconstruction algorithm to obtain the temperature distribution of the furnace cross section. The emissivity distribution correction unit then inverts the emissivity distribution based on the preset initial emissivity value, the temperature field reconstruction results, and the dual-band spectral radiation intensity, and corrects the temperature field of the furnace cross section using the dual-band emissivity ratio constraint.

[0052] In one specific implementation, the system of this invention is applied to a 600MW subcritical coal-fired boiler. The boiler furnace cross-section is 19.2m × 19.2m, and the burners are arranged on the front wall. Four colorimetric temperature probes are installed at four observation holes at a furnace elevation of 28m, forming a four-viewpoint measurement array covering the front wall, rear wall, left wall, and right wall.

[0053] In this embodiment, each colorimetric temperature probe employs a dual-CCD camera structure, equipped with narrow-band filters (8nm bandwidth) with center wavelengths of 900nm and 1000nm, a CCD resolution of 1024×1024 pixels, and a frame rate of 25fps. The probe is equipped with a compressed air cooling system with a cooling airflow of 50Nm³ / h and an air pressure of 0.6MPa.

[0054] Specifically, the system workflow is as follows: Calibration stage: Calibration is performed using a blackbody furnace in the range of 800-1600℃ to establish the grayscale-radiation intensity relationship curve; the geometric parameters of the four probes are determined through a three-dimensional coordinate target, and the reconstruction accuracy calibration results show that the spatial positioning error is <5mm; Real-time measurement: Simultaneously acquire dual-band images from four viewpoints. After preprocessing, first calculate the apparent temperature field of each viewpoint as the initial value for iteration. Medium parameters were obtained: CO2 concentration (approximately 12%), H2O concentration (approximately 8%), and fly ash concentration (approximately 15 g / Nm³) were obtained using a flue gas analyzer installed at the furnace outlet. The absorption coefficients at 900 nm and 1000 nm wavelengths were calculated to be 0.15 m. - ¹ and 0.12m - ¹; Temperature field reconstruction: Iterative reconstruction is performed using the SIRT algorithm, with a grid of 80×80, 50 iterations, and a single reconstruction calculation time of approximately 40ms, which meets the real-time requirements. The results show that the reconstructed cross-sectional temperature field indicates that the temperature is highest in the central region of the furnace (approximately 1450℃) and lower near the wall (approximately 1200℃). The temperature distribution exhibits a clear characteristic of combustion center shift, which is in good agreement with the CFD simulation results, with a relative error of <5%.

[0055] To aid in a better understanding of the invention, a more comprehensive and specific embodiment of the invention is described, in which, as follows: Figure 3 As shown, the present invention provides a method for reconstructing the temperature field of a furnace cross-section based on multiple viewpoints, comprising the following steps: S1: System Calibration and Initialization A high-precision blackbody furnace was used to calibrate the dual-band radiation intensity of each colorimetric temperature probe, and a quantitative relationship between the detector output gray value and the radiation intensity was established. The geometric parameters (optical center position, optical axis direction, focal length) and distortion coefficient of each probe are determined using a standard target. A two-dimensional discrete mesh model of the furnace cross-section is established, and the area to be measured is divided into M×M pixel units.

[0056] S2: Multi-view synchronous image acquisition The synchronous trigger control unit issues a data acquisition command, and N colorimetric temperature probes simultaneously acquire dual-band radiation images inside the furnace. Acquire 2N digital images, denoted as I1 (n) (x,y) and I2 (n) (x,y), where n=1,2,...,N represents the probe number, and (x,y) is the image pixel coordinate.

[0057] S3: Radiation Intensity Calculation and Preprocessing Based on the calibration coefficients, the image grayscale values ​​are converted into spectral radiance:

[0058] in, and They represent the first Each probe at image pixel coordinates Location, wavelength is and Spectral radiance at that time; and They represent the first The original image grayscale values ​​obtained by one probe in two spectral bands; and The first One probe in , Calibration slope coefficient under the band; and These are the calibration offset coefficients for the corresponding wavebands; and These are the nonlinear response correction coefficients for the two channels, which can be taken as 1 when the detector response is approximately linear. These are the pixel coordinates of the image. Number the probe.

[0059] Median filtering and bad pixel correction are applied to the radiation intensity image to eliminate the effects of random noise and detector defects.

[0060] S4: Preliminary Calculation of Colorimetric Temperature Field For the dual-band radiation intensity image at each viewpoint, the apparent temperature distribution along the line of sight is calculated using a colorimetric thermometry formula:

[0061] Among them, under the gray body assumption The formula simplifies to:

[0062] in, Indicates the first Each probe at pixel coordinates The apparent temperature obtained along the line of sight; It is Planck's second radiation constant; and Wavelength and Spectral radiance at that location; and These are the center wavelengths of the dual-band narrowband filter, and they are different; and The target being measured is in and Emissivity in the band; This represents the natural logarithm. Under the gray body assumption, it can be considered that... Therefore, the emissivity ratio term can be ignored, and the formula can be simplified accordingly. S5: Obtaining the radiation characteristic parameters of the medium Real-time data on CO2, H2O, O2 concentrations and fly ash concentrations inside the furnace are obtained using a flue gas analyzer. Based on the HITRAN spectral database and Mie scattering theory, the medium absorption coefficients in the λ1 and λ2 bands were calculated. and scattering coefficient ; in, Indicates wavelength The absorption coefficient of the medium is used to characterize the ability of media such as flue gas and fly ash to absorb radiant energy. Indicates wavelength The scattering coefficient of the medium is used to characterize the scattering effect of the medium on the direction of radiation propagation and energy distribution; all of the above parameters vary with the flue gas composition, particle concentration, temperature and selected wavelength. Establish a distribution model of the medium radiation characteristics of the furnace cross section.

[0063] S6: Solving the radiative transfer equation and reconstructing the temperature field Establish the radiative transfer equation for the furnace cross-section:

[0064] in, Indicates wavelength Below, position Along the direction The intensity of the propagating spectral radiation; Represents the spatial coordinates along the radiation propagation path; Indicates the extinction coefficient, which satisfies ; Indicates the absorption coefficient; Represents the scattering coefficient; Indicates position Location, temperature is The blackbody spectral radiation intensity at that time; This represents the scattering phase function, used to describe the scattering of radiation from the incident direction. Scattered towards the observation direction Angular distribution pattern; For the direction of observation, The incident direction is within the integration range; Indicates the range of solid angles across the entire space; The radiative transfer equation is discretized using the Discrete Coordinate Method (DOM) or the Finite Volume Method (FVM). Establish a linearized system of reconstruction equations:

[0065] in, To reconstruct the coefficient matrix, its elements comprehensively characterize the geometric relationship between each measurement optical path and the discrete grid cell of the cross section, the radiation absorption and scattering characteristics of the medium, and the discrete calculation weights. Let be the cross-sectional temperature vector to be solved; The measured radiation intensity vector is composed of multi-viewpoint dual-band measurement data; The iterative reconstruction algorithm is used to solve the problem. Initialize temperature field (The apparent temperature field from step S4 can be used as the initial value). k-th iteration: Calculate the composite radiation intensity under the current temperature field.

[0066] Calculate the residuals:

[0067] Update the temperature field: ,in, Indicates the initial value of the temperature field; Indicates the first The temperature vector obtained from the next iteration; This represents the composite radiation intensity vector calculated from the current temperature field in the forward direction; This represents the residual vector between the measured radiation intensity and the calculated radiation intensity; This represents the relaxation factor, which is used to control the step size update in each iteration and improve the stability of the solution. and These represent the physical lower limit and physical upper limit of the temperature field, respectively; Indicates the iterative convergence threshold; Indicates the number of iterations; Determine the convergence condition: If If the maximum number of iterations is reached, the iteration will stop.

[0068] S7: Emittance Field Synchronization Correction Based on the reconstructed temperature field and measured radiation intensity, the emissivity distribution of the inverted medium is retrieved:

[0069] in, Indicates the first The wavelength obtained in the next iteration Emission rate; This represents the measured total spectral radiance. This represents the radiation component introduced by scattering from the medium; Indicates the medium absorption coefficient; Indicates path location Location, temperature is The blackbody spectral radiation intensity at that time; Indicates the extinction coefficient; Indicates the total optical path along the line of sight; and For intermediate path variables along the optical path integral; The transmission attenuation factor represents the transmission attenuation factor of radiation propagating in a medium. By utilizing the constraint of the emissivity ratio of the two bands, the temperature field calculation results are corrected, eliminating the error introduced by the emissivity assumption.

[0070] S8: Results Output and Visualization The reconstructed temperature field is mapped onto the furnace cross-section grid to generate a pseudo-color temperature distribution map. Calculate statistical parameters such as cross-sectional average temperature, maximum temperature, and temperature non-uniformity coefficient; An alarm signal is triggered when the temperature exceeds the set threshold or when the temperature distribution is abnormal. The iterative reconstruction algorithm described in step S6 employs an improved algorithm based on total variation (TV) regularization:

[0071] in, This represents the data fidelity term, used to characterize the L2 error between the reconstruction results and the measured data; This represents the total variation regularization parameter, used to balance the accuracy of data fitting with the smoothness of the solution; Represents the temperature field The total variation.

[0072] In this embodiment, a boiler furnace cross-sectional temperature field reconstruction system based on a multi-viewpoint colorimetric pyrometer includes: A multi-viewpoint colorimetric thermography imaging array, consisting of N (N≥3) colorimetric thermography probes, arranged circumferentially along the furnace cross-section, each probe comprising: High-temperature protection and installation unit: includes a high-temperature protective cover, conduit, and cooler. The high-temperature protective cover is fixed to the furnace wall opening via conduit, with the probe placed inside the cover; the cooler achieves cooling through compressed air or water cooling to ensure stable operation of the equipment in ambient temperatures up to 70°C (operating temperature 0~70°C, additional cooling is required beyond this range); all metal hoses are made of 304 stainless steel. Protection mechanism unit: Built-in pneumatic valve and control system. When the cooling air supply stops or the probe internal temperature exceeds the limit, the pneumatic valve automatically closes, effectively protecting the pyrometer from damage; Optical imaging unit: It adopts dual-band narrowband filters (center wavelengths λ1 and λ2, bandwidth ≤10nm, spectral range can cover 0.8~1.8μm, suitable for materials such as metals, graphite, and ceramics) and optical lenses to acquire radiation signals of two specific wavelengths inside the furnace. Dual-channel photodetector: Employs a high-sensitivity, fast-response detector (response time t95 can be as low as 10ms) to receive radiation signals in the λ1 and λ2 bands respectively and convert them into electrical signals; Emissivity adjustment unit: Located at the rear of the device, it is equipped with a decimal switch (tens and units digits) and a locking nut for manually setting the emissivity value (adjustment range 0.05~1.00, step size 0.01) to adapt to different test materials; The synchronous triggering and control unit is used to achieve synchronous signal acquisition from N probes, ensuring the time consistency of data from each viewpoint; A high-speed data transmission network transmits multiple signals to a data processing server in real time.

[0073] The temperature field reconstruction calculation module includes: Signal preprocessing module: Denoises, amplifies, and performs analog-to-digital conversion on the dual-channel signal, and converts it into spectral radiance based on calibration coefficients; Medium radiation characteristics calculation module: Based on real-time data of furnace flue gas composition (CO2, H2O, fly ash concentration), calculate the absorption coefficient and scattering coefficient of high-temperature medium in the λ1 and λ2 bands; Temperature field reconstruction module: Based on the radiative transfer equation (RTE) and algebraic reconstruction technique (ART) or regularized reconstruction algorithm, solve for the temperature distribution of the furnace cross section; Emissivity distribution correction module: Combines the preset initial emissivity value and the reconstructed temperature field to iteratively correct the measurement results; The visualization and diagnostic unit is used to display the pseudo-color distribution map of the cross-sectional temperature field, isotherm map, temperature statistics, and provides an alarm function for abnormal combustion.

[0074] A further improvement of the present invention is that the center wavelength of the dual-band narrowband filter can be selected from two different wavelengths in the range of 1.5 to 1.8 μm.

[0075] Specifically, N colorimetric pyrometer probes are evenly arranged around the furnace cross-section, with an angle of 360° / N between adjacent probes to ensure full coverage observation of the cross-sectional area.

[0076] In summary, the embodiments disclosed herein have at least the following technical effects: This invention employs multiple colorimetric temperature probes arranged circumferentially along the boiler furnace cross-section for simultaneous multi-viewpoint measurement. This enables the acquisition of dual-band radiation information of the furnace cross-section from multiple observation directions, overcoming the problem that existing single-viewpoint measurements can only obtain integrated information along the line of sight and are difficult to reflect the true temperature distribution of the cross-section, thereby improving the completeness and accuracy of the furnace cross-section temperature field reconstruction. This invention employs a dual-band colorimetric thermometry method to determine the apparent temperature by the ratio of radiation intensity at two wavelengths. This method has the advantages of low dependence on the absolute value of emissivity and strong anti-interference ability. At the same time, combined with synchronous correction of emissivity distribution, it can further reduce the error introduced by the emissivity assumption and improve the accuracy of temperature solution. This invention acquires the composition of flue gas and fly ash concentration in the furnace in real time, and establishes a medium radiation characteristic model based on the HITRAN spectral database and Mie scattering theory. It incorporates the absorption and scattering effects of high-temperature flue gas and fly ash particles on radiation transmission into the temperature field reconstruction process, which can effectively reduce the measurement error caused by the complex furnace medium environment and improve the authenticity and reliability of the reconstruction results. This invention establishes a temperature field reconstruction model based on the radiative transfer equation, and combines the initial value of apparent temperature, iterative reconstruction algorithm and physical constraints to solve the temperature field of the furnace cross section. It can balance reconstruction accuracy and computational efficiency and meet the needs of online monitoring. This invention employs a colorimetric temperature probe structure that integrates high-temperature protection, cooling, and protection mechanisms. It can operate stably in the high-temperature and high-dust environment of boiler furnaces and automatically protects itself in case of abnormal cooling or probe overheating. Therefore, it has the advantages of high equipment reliability, non-contact measurement, and low maintenance. It is suitable for combustion monitoring and optimization control of industrial boilers such as thermal power generation and waste incineration plants.

[0077] 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 multi-view based reconstruction method of a temperature field of a furnace cross section, characterized in that, Includes the following steps: S1: Calibrate N colorimetric temperature probes arranged circumferentially along the cross-section of the boiler furnace, where N≥3, and establish a discrete grid model of the cross-section of the furnace to determine the geometric projection relationship between the detection line of each colorimetric temperature probe and the discrete grid unit. S2: Control N colorimetric temperature probes to collect radiation images of the furnace interior at the first wavelength λ1 and the second wavelength λ2 in the line of sight of each colorimetric temperature probe, thereby obtaining dual-band radiation images of each colorimetric temperature probe at the first wavelength λ1 and the second wavelength λ2. S3: Based on the calibration results obtained in step S1, the dual-band radiation image is converted into spectral radiation intensity, and the apparent temperature distribution along the line of sight of each colorimetric temperature probe is calculated based on the ratio of the spectral radiation intensity at the first wavelength λ1 and the second wavelength λ2, as the initial temperature field for reconstructing the temperature field of the furnace cross section. S4: Obtain the flue gas composition and fly ash concentration data in the furnace, calculate the absorption coefficient and scattering coefficient of the furnace medium at the first wavelength λ1 and the second wavelength λ2, and establish a medium radiation characteristic model of the furnace cross section. S5: Based on the apparent temperature distribution, the medium radiation characteristic model, and the spectral radiation intensity measured in the line of sight of each of the colorimetric temperature probes, establish a radiation transfer reconstruction model between the measured spectral radiation intensity and the furnace cross-sectional temperature field to be determined, and obtain the furnace cross-sectional temperature field. S6: Based on the preset initial emissivity value, the furnace cross-sectional temperature field, and the spectral radiation intensity, iteratively invert the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2, and correct the furnace cross-sectional temperature field according to the dual-band emissivity ratio constraint, and output the corrected furnace cross-sectional temperature field.

2. The multi-view based furnace cross-sectional temperature field reconstruction method according to claim 1, characterized in that, In step S1, the calibration of the colorimetric temperature probes includes: using a blackbody furnace to perform dual-band radiation intensity calibration on each colorimetric temperature probe, establishing a quantitative relationship between the image grayscale value and radiation intensity acquired by each colorimetric temperature probe at the first wavelength λ1 and the second wavelength λ2; determining the optical center position, optical axis direction, focal length and distortion coefficient of each colorimetric temperature probe using a standard target, and dividing the furnace cross-section to be measured into multiple grid units.

3. The multi-view based furnace cross-sectional temperature field reconstruction method according to claim 1, characterized in that, In step S3, after converting the information of the dual-band radiation image into spectral radiation intensity, the spectral radiation intensity is further preprocessed, which includes at least median filtering and bad pixel correction.

4. The multi-view based furnace cross section temperature field reconstruction method according to claim 1, characterized in that, In step S4, obtaining data on flue gas composition and fly ash concentration in the furnace includes obtaining data on CO2, H2O, O2 concentrations and fly ash concentrations in the furnace using a flue gas analyzer.

5. The furnace cross-sectional temperature field reconstruction method based on multiple viewpoints according to claim 1, characterized in that, In step S4, the average temperature of the furnace medium along a preset path and the average concentration of at least one flue gas component along the path are measured using laser absorption spectroscopy or tuned diode laser absorption spectroscopy to serve as boundary conditions or verification data for radiative transfer calculation.

6. The furnace cross-sectional temperature field reconstruction method based on multiple viewpoints according to any one of claims 1 to 5, characterized in that: In step S5, obtaining the furnace cross-sectional temperature field includes: discretizing the radiative transfer reconstruction model using the discrete coordinate method or the finite volume method, establishing a linearized reconstruction equation set, and solving it using algebraic reconstruction technology, synchronous iterative reconstruction technology, or an iterative reconstruction algorithm based on total variation regularization; in each iteration, based on the current iterative temperature field, calculating the predicted radiation intensity along the observation path of each colorimetric temperature probe along the line of sight of the furnace cross-section, comparing the predicted radiation intensity with the measured furnace radiation intensity to obtain the residual, updating the furnace cross-sectional temperature field according to the residual; applying upper and lower temperature physical constraints to the updated furnace cross-sectional temperature field, and stopping the iteration when the residual is less than a preset threshold or the maximum number of iterations is reached.

7. A furnace cross-sectional temperature field reconstruction system based on multiple viewpoints, characterized in that, include: A multi-viewpoint colorimetric thermography imaging array includes N colorimetric thermography probes arranged circumferentially along the cross-section of the boiler furnace, wherein N≥3, for collecting radiation signals at the first wavelength λ1 and the second wavelength λ2 inside the furnace. A synchronous triggering and control unit is used to control the N colorimetric temperature probes to collect radiation signals; A data transmission network is used to transmit the radiation signals collected by each of the colorimetric temperature probes; The temperature field reconstruction calculation module is used to convert the radiation signal into spectral radiance based on the quantitative relationship between the image grayscale value and radiance intensity established through calibration; to calculate the apparent temperature distribution based on the ratio of spectral radiance intensity at the first wavelength λ1 and the second wavelength λ2 as the initial value for temperature field reconstruction; to establish a medium radiation characteristic model based on the flue gas composition and fly ash concentration data in the furnace, and to solve the temperature field of the furnace cross-section; and to correct the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2 based on the reconstructed temperature field of the furnace cross-section; and... The output and diagnostic unit is used to output the distribution of the temperature field in the furnace cross section and the diagnostic results.

8. The furnace cross-sectional temperature field reconstruction system based on multiple viewpoints according to claim 7, characterized in that, Each of the colorimetric temperature probes includes an optical imaging unit, a dual-channel photodetector, and an emissivity adjustment unit for adjusting the emissivity setpoint.

9. The furnace cross-sectional temperature field reconstruction system based on multiple viewpoints according to claim 8, characterized in that, The optical imaging unit includes a dual-band narrowband filter and an optical lens. The center wavelength of the dual-band narrowband filter is two different wavelengths located in the range of 0.8 to 1.8 μm, and the bandwidth is no greater than 10 nm. The first wavelength λ1 and the second wavelength λ2 are the two center wavelengths of the dual-band narrowband filter. The dual-channel photodetector receives radiation signals at the first wavelength λ1 and the second wavelength λ2 respectively and converts them into electrical signals.

10. The furnace cross-sectional temperature field reconstruction system based on multiple viewpoints according to claim 7, characterized in that: The temperature field reconstruction calculation module includes: The signal preprocessing unit is used to denoise, amplify, and perform analog-to-digital conversion on the dual-channel radiation electrical signals inside the furnace at the first wavelength λ1 and the second wavelength λ2 collected by each of the colorimetric temperature probes, and convert the analog-to-digital converted signals into spectral radiation intensity that characterizes the radiation characteristics of the furnace according to the calibration coefficients. The dielectric radiation characteristics calculation unit is used to calculate the absorption coefficient and scattering coefficient at the first wavelength λ1 and the second wavelength λ2 based on the composition of the flue gas in the furnace and the concentration of fly ash. Temperature field reconstruction unit, used to solve the temperature distribution of the furnace cross section based on the radiative transfer equation and iterative reconstruction algorithm; and The emissivity distribution correction unit is used to invert the emissivity distribution of the furnace medium at the first wavelength λ1 and the second wavelength λ2 based on the preset initial emissivity value, the furnace cross-sectional temperature field output by the temperature field reconstruction unit, and the dual-band spectral radiation intensity, and to correct the furnace cross-sectional temperature field using the dual-band emissivity ratio constraint.

Images