FBG-based boiler multi-spectral temperature measurement calibration method and system

Abstract

This invention relates to the field of radiation thermometry technology, and proposes a boiler multispectral temperature measurement calibration method and system based on FBG. The method includes: acquiring the reflection spectrum signal of the boiler heating surface, demodulating the signal to obtain a center wavelength offset sequence; decoupling the center wavelength offset sequence according to the stress compensation coefficient of the sensing point to obtain the stress-compensated temperature value; using the two-dimensional spatial coordinates of the boiler heating surface as a positioning reference, anchoring the stress-compensated temperature value to the corresponding coordinate position, and generating a continuous temperature distribution covering the entire heating surface according to the spatial distance between adjacent coordinate positions and the difference of the anchored temperature value, based on the linear heat transfer relationship; calculating the temperature gradient value and heating rate value of each region, comparing the temperature gradient value and heating rate value with a preset classification threshold, and outputting a classification warning signal. This invention can improve the efficiency of boiler multispectral temperature measurement calibration based on FBG.

Description

Technical Field

[0001] This invention relates to the field of radiation thermometry, and in particular to a boiler multispectral temperature measurement calibration method and system based on FBG. Background Technology

[0002] High-temperature heating surfaces such as boiler water-cooled walls and reheaters are core pressure-bearing components of ultra-supercritical units. Under deep peak-shaving conditions, uneven heat load distribution within the boiler can easily lead to tube rupture and leakage accidents due to localized overheating of the tube walls. Fiber optic grating multispectral sensing, with its advantages of high-temperature resistance and distributed deployment, has become the mainstream technology for comprehensive boiler wall temperature monitoring. This technology calculates tube wall temperature by collecting the wavelength shift of the grating's reflected spectrum, making it a core sensing method for achieving temperature field visualization and overheat warning. The completeness and accuracy of the temperature field data directly determine the effectiveness of identifying overheating risks of heating surfaces.

[0003] Existing multispectral temperature measurement systems for boilers can only acquire temperature data with stress interference from discrete measurement points. They cannot eliminate stress wavelength shifts caused by vibration and pipe wall deformation, nor can they extrapolate temperatures in areas without measurement points based on the geometric characteristics of the heating surface. A complete and continuous temperature field cannot be constructed, leading to distortions in the calculation of temperature gradients and pipe wall heating rates. Ultimately, this results in unreliable early warning and classification results for heating surface overheating risks. Therefore, accurately determining the overheating risk of boiler heating surfaces has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a boiler multispectral temperature measurement calibration method and system based on FBG to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this invention provides a boiler multispectral temperature measurement calibration method based on FBG, comprising: Ⅰ. Obtain the reflection spectrum signal of the sensing point in the fiber grating array of the boiler heating surface, and demodulate the reflection spectrum signal to obtain the center wavelength offset sequence of the sensing point; II. Decouple the center wavelength offset sequence from stress and temperature based on the stress compensation coefficient of the sensing point to obtain the temperature value after stress compensation. III. Using the two-dimensional spatial coordinates of the boiler heating surface as the positioning reference, the stress-compensated temperature value is anchored to the corresponding coordinate position. Based on the spatial distance between adjacent coordinate positions and the difference between the anchored temperature values, the temperature value of the area without measurement points is deduced according to the linear heat transfer relationship, generating a continuous temperature distribution covering the entire heating surface. IV. Based on the continuous temperature distribution, calculate the temperature gradient value and heating rate value of each region, compare the temperature gradient value and heating rate value with the preset grading threshold, and output a graded early warning signal that matches the degree of overheating.

[0006] To address the above problems, the present invention also provides a boiler multispectral temperature measurement and calibration system based on FBG, the system comprising: The reflection spectrum demodulation module is used to acquire the reflection spectrum signal of the sensing point in the fiber grating array of the boiler heating surface, and demodulate the reflection spectrum signal to obtain the center wavelength offset sequence of the sensing point. The stress-temperature decoupling module is used to perform stress-temperature decoupling on the center wavelength offset sequence according to the stress compensation coefficient of the sensing point, so as to obtain the temperature value after stress compensation. The temperature field spatial extrapolation module is used to anchor the stress-compensated temperature value to the corresponding coordinate position using the two-dimensional spatial coordinates of the boiler heating surface as the positioning reference, and extrapolate the temperature value of the area without measurement points according to the spatial distance between adjacent coordinate positions and the difference of the anchored temperature value, generating a continuous temperature distribution covering the entire heating surface. The graded early warning determination module is used to calculate the temperature gradient value and heating rate value of each region based on the continuous temperature distribution, compare the temperature gradient value and heating rate value with the preset graded threshold, and output a graded early warning signal that matches the degree of overheating.

[0007] Compared with the prior art, the present invention has the following beneficial effects: 1. This solution performs stress-temperature decoupling processing on the wavelength offset of the sensing point, accurately eliminates the wavelength offset component caused by stress, and outputs the true pipe wall temperature value without stress interference. The calculation accuracy of single-point temperature measurement data is steadily improved, providing a high-precision basic data source for subsequent temperature field extrapolation.

[0008] 2. This scheme relies on the two-dimensional coordinate grid of the heating surface to complete the temperature anchoring of the measuring points, and combines the linear heat transfer law to deduce the temperature of the blank area to form a continuous temperature field in the whole domain. Based on the unified standard, the regional temperature gradient and heating rate are calculated, and the graded early warning signal matching the degree of overheating is output, so as to effectively improve the accuracy of boiler heating surface overheating risk identification. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating a boiler multispectral temperature measurement calibration method based on FBG, provided in one embodiment of the present invention. Figure 2 This is a functional block diagram of a boiler multispectral temperature measurement and calibration system based on FBG, provided in one embodiment of the present invention.

[0010] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0011] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0012] This application provides a boiler multispectral temperature measurement calibration method based on FBG. The execution entity of this FBG-based boiler multispectral temperature measurement calibration method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application embodiment: a server, a terminal, etc. In other words, the FBG-based boiler multispectral temperature measurement calibration method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0013] Reference Figure 1 The diagram shown is a flowchart illustrating a boiler multispectral temperature measurement calibration method based on FBG according to an embodiment of the present invention. In this embodiment, the boiler multispectral temperature measurement calibration method based on FBG includes: Ⅰ. Obtain the reflection spectrum signal of the sensing point in the fiber grating array of the boiler heating surface, and demodulate the reflection spectrum signal to obtain the center wavelength offset sequence of the sensing point; In practice, before acquiring the reflection spectrum signal of the sensing points in the fiber optic grating array of the boiler heating surface, it is necessary to first deploy fiber optic grating sensing arrays at the heating surface locations such as the boiler water-cooled wall, low-temperature reheater, and coal mill outlet pulverized pipe. A multi-channel spectral demodulator is deployed in the electronics room. The demodulator outputs probe light and transmits it to each sensing point via a transmission optical cable, while simultaneously receiving the reflection spectrum signal returned by each measuring point.

[0014] The number of channels per device for the spectral demodulator ranges from 48 to 256, determined based on the total number of measurement points in the project and the available space in the on-site cabinet. A preferred configuration is 64 channels. This range is chosen because: if the number of channels is too low (below 48), the number of measurement points covered by a single device is limited, requiring more demodulation devices and increasing the space occupied by the on-site cabinet and cabling complexity; if the number of channels is too high (above 256), the bandwidth for single-channel data processing is compressed, making it difficult to meet the performance requirements of high-frequency acquisition. Considering the total scale of 1839 measurement points outlined in the technical briefing, a 64-channel configuration can achieve full measurement point coverage with 5 devices while ensuring single-channel acquisition performance, achieving an optimal balance between equipment cost and deployment complexity.

[0015] The demodulator's wavelength scanning range is 1510nm to 1590nm, which matches the operating window of C-band fiber Bragg gratings and can cover the grating wavelength drift requirements across the entire temperature range. This range was chosen because: a wavelength range that is too narrow, such as less than 1510nm to 1590nm, cannot cover the long-wavelength drift limit of the grating at 950℃, leading to signal loss at high-temperature measurement points; a wavelength range that is too wide introduces out-of-band noise, reducing the demodulation signal-to-noise ratio and increasing hardware costs. The 1510nm to 1590nm range can completely cover the grating wavelength shift across the entire temperature range from -60℃ to 950℃, balancing demodulation accuracy and hardware economy.

[0016] The spectral data acquisition frequency is 50Hz to 200Hz, determined based on the dynamic changes in boiler wall temperature, with 100Hz being the preferred choice. This range is chosen because: if the acquisition frequency is too low (below 50Hz), transient fluctuations in wall temperature under deep peak-shaving conditions cannot be captured, easily missing rapid overheating processes; if the acquisition frequency is too high (above 200Hz), the data volume per measurement point increases significantly, and the system's computing load surges when processing thousands of measurement points in parallel, making real-time performance difficult to guarantee. A 100Hz acquisition frequency corresponds to a 10ms sampling interval, which can fully capture the step changes in wall temperature caused by load fluctuations while keeping the data processing volume within the system's computing power capacity, thus adapting to the parallel monitoring needs of a large number of measurement points.

[0017] Regarding the demodulation method of the reflection spectrum signal, in some embodiments of this application, the center wavelength is preferably extracted by using spectral processing combined with a Gaussian peak optimization algorithm.

[0018] After the reflected light signal enters the demodulator, it is first discretized by the beam splitter according to the wavelength dimension. It is then split into continuous wavelength-intensity corresponding data groups according to a fixed wavelength step size. Each wavelength position corresponds to a light intensity sampling value, which completely covers the entire band range of the grating reflection peak.

[0019] The wavelength step size for spectroscopic processing can be selected within the range of 1 pm to 10 pm, with 2 pm being the preferred value. If the wavelength step size is too large, the sampling points become too sparse, making it impossible to accurately locate the spectral peak position, and significantly increasing the demodulation error. If the wavelength step size is too small, the number of sampling points for a single spectrum increases exponentially, prolonging the data processing time and making it difficult to meet the high-frequency acquisition requirements of 100 Hz. A wavelength step size of 2 pm can control the processing time of a single spectrum within 1 ms while ensuring peak positioning accuracy, thus adapting to the real-time requirements of high-frequency acquisition.

[0020] After completing the spectral sampling, peak location calculations are performed on the spectral band corresponding to each sensing point. The local band where the maximum light intensity is located is selected, and the shape of the reflection peak is fitted by a Gaussian function. The wavelength value corresponding to the vertex of the fitted curve is then calculated and used as the current center wavelength of the measuring point.

[0021] The fitting window width for Gaussian peak optimization can be set from 5 to 15 sampling points, with 9 sampling points being the preferred value. If the fitting window is too small, it is susceptible to local noise interference, leading to peak positioning shifts and decreased demodulation stability. If the fitting window is too large, it will include the spectral edge peaks of adjacent measurement points in the fitting range, generating crosstalk errors. A window width of 9 sampling points can completely cover the main lobe region of the single-grating reflection peak while avoiding crosstalk from adjacent peaks, ensuring the accuracy and stability of peak positioning.

[0022] After obtaining the current center wavelength of each measuring point, the initial calibration center wavelength of that measuring point is retrieved from the pre-stored system calibration library, and the difference between the two is used to calculate the wavelength offset of that measuring point.

[0023] The initial calibration ambient temperature for the center wavelength can be controlled within the range of 25℃±5℃, with 25℃ being the preferred standard room temperature condition. Excessive fluctuations in the calibration temperature will cause deviations in the initial wavelength reference, introducing systematic errors into subsequent temperature calculations across the entire temperature range. Conversely, excessively high calibration temperature control precision significantly increases calibration environmental costs, making batch calibration of measurement points uneconomical. 25℃ as the standard room temperature condition ensures the consistency of the reference wavelength across all measurement points while controlling calibration environmental costs.

[0024] Finally, according to the serial arrangement of the measurement points on a single optical fiber, the wavelength offsets of all measurement points in the same channel are arranged sequentially, and then the data of all measurement points are collected according to the channel number to form a complete center wavelength offset sequence, which serves as the input data for subsequent temperature calculation.

[0025] Through the above-mentioned spectral dispersion and Gaussian peak-finding demodulation processing, the center wavelength shift of each measurement point can be accurately extracted, and the discrete light intensity signal can be transformed into a structured wavelength shift sequence, providing standardized input data for subsequent stress decoupling and temperature calculation.

[0026] II. Decouple the center wavelength offset sequence from stress and temperature based on the stress compensation coefficient of the sensing point to obtain the temperature value after stress compensation. In practice, before stress-temperature decoupling, it is necessary to first complete the dual-parameter calibration of each sensing point, obtain the independent stress compensation coefficient of each measuring point and store it in the system calibration database; during the calculation, the compensation coefficient of the corresponding measuring point is called, the stress-induced drift component is separated from the measured center wavelength offset, and the remaining temperature-induced component is converted into the true pipe wall temperature through the calibration curve.

[0027] The number of temperature calibration points set during calibration can be selected from 5 to 10, with 7 being preferred, corresponding to temperature points of 100℃, 200℃, 300℃, 400℃, 500℃, 600℃, and 700℃ respectively.

[0028] With too few calibration points, only first-order linear fitting can be achieved, resulting in significant nonlinear deviations in the high-temperature range and compromising the accuracy of temperature measurement across the entire temperature range. Conversely, with too many calibration points, the calibration cycle for a single measurement point is drastically extended, leading to excessively low efficiency in batch calibration and making it difficult to meet project implementation requirements. Seven calibration points evenly cover the commonly used operating temperature range of the boiler. Through piecewise linear fitting, the fitting error across the entire temperature range can be controlled within ±0.3℃, while simultaneously keeping the calibration time for a single measurement point within a reasonable range, thus balancing calibration accuracy and implementation efficiency.

[0029] The stress loading range during calibration can be set from 0 MPa to 200 MPa, with a preferred loading range of 0 MPa to 150 MPa.

[0030] If the loading range is too small, it cannot cover the peak stress fluctuations in the pipe wall under deep peak-shaving conditions, leading to compensation failure under extreme conditions. If the loading range is too large, the load control accuracy requirements of the calibration equipment are significantly increased, and high stress can easily cause deterioration of the grating structure, reducing the long-term service life of the sensor. The axial stress of the boiler water-cooled wall during normal operation is approximately 50MPa to 120MPa. A loading range of 0MPa to 150MPa can cover most operating stress scenarios, while avoiding high stress damage to the grating structure, thus balancing compensation coverage and sensor reliability.

[0031] The calculation accuracy of the stress compensation coefficient can be controlled within the range of ±1% to ±5%, preferably reaching ±2%.

[0032] When the accuracy is too low, the temperature measurement deviation introduced by stress can reach more than ±2.5℃, which cannot meet the overall temperature measurement accuracy requirements of the system. When the accuracy is too high, the load and temperature control accuracy of the calibration equipment need to be significantly increased, resulting in a significant increase in hardware costs and insufficient cost-effectiveness in the project. An accuracy of ±2% corresponds to a temperature-introduced error of about ±1℃. After adding the system's own error, it can still meet the comprehensive accuracy requirement of ±0.5℃, achieving the best cost-effectiveness in the project.

[0033] By configuring the above calibration parameters, accurate and reliable stress compensation coefficients for single measuring points can be obtained, providing a precise calculation basis for subsequent stress-temperature decoupling and ensuring the accuracy of the compensated temperature.

[0034] Regarding the implementation of stress-temperature decoupling, in some embodiments, this application preferably employs a single-grating dual-parameter linear decoupling method.

[0035] The calibration phase begins with stress-free temperature response testing. The sensing points are placed in a stress-free constant temperature environment, and the temperature is gradually increased according to the set calibration points. Each temperature point is held at the temperature until the temperature field stabilizes, and the center wavelength value at the corresponding temperature is recorded to generate a temperature response offset sequence.

[0036] The temperature fluctuation of the constant-temperature calibration environment can be controlled within the range of ±0.1℃ to ±1℃, with ±0.5℃ being the preferred value. Excessive temperature fluctuation increases the measurement error of the temperature response offset, directly leading to deviations in the sensitivity coefficient calculation; conversely, insufficient temperature fluctuation necessitates the use of a high-precision constant-temperature furnace, significantly increasing equipment procurement costs. A fluctuation of ±0.5℃ can control the calibration error of the sensitivity coefficient within ±1%, while simultaneously controlling the investment cost of calibration equipment.

[0037] Subsequently, a constant temperature stress response test was conducted. The ambient temperature was kept constant, and a graded axial load was applied to the sensing point through a stress loading device. After each load was stabilized, the center wavelength value was recorded to generate a stress response offset sequence.

[0038] The stress loading hold-up stabilization time can be set within the range of 10 to 30 minutes, with 20 minutes being the preferred value. If the hold-up time is too short, the stress transfer between the sensor and the substrate will not reach a stable state, resulting in a smaller measured stress response offset; if the hold-up time is too long, the calibration cycle for a single measurement point will be significantly extended, leading to low batch calibration efficiency. A 20-minute hold-up time ensures that the stress is fully transferred to the grating structure while also maintaining the efficiency of batch calibration.

[0039] Based on two sets of test data, the temperature sensitivity value and stress sensitivity value of a single measuring point were calculated respectively: the temperature sensitivity value is the ratio of the temperature response offset to the corresponding temperature change, reflecting the wavelength drift amplitude caused by a unit temperature change. The stress sensitivity value is the ratio of the stress response offset to the corresponding stress change, reflecting the wavelength drift amplitude caused by a unit stress change. The ratio of the stress sensitivity value to the temperature sensitivity value at the same measuring point is used as the stress compensation coefficient for that measuring point and stored in the system calibration database.

[0040] In actual operation and calculation, the stress compensation coefficient of the corresponding measuring point is retrieved, and the measured center wavelength offset is multiplied by the coefficient to calculate the stress-induced wavelength offset component. After removing the stress-induced offset component from the total center wavelength offset, the remaining portion is the wavelength offset component caused by pure temperature change. Finally, the temperature-induced offset component is substituted into the calibration temperature-wavelength mapping relationship for that measuring point to calculate the true pipe wall temperature value after stress compensation.

[0041] The fitting order of the temperature-wavelength mapping relationship can be selected from the first to the third order, with the second-order polynomial fitting being preferred. First-order linear fitting cannot cover the nonlinear deviation in the high-temperature range, resulting in a large error across the entire temperature range; third-order fitting is prone to overfitting, leading to error fluctuations in the intermediate temperature region. Second-order polynomial fitting can better match the weak nonlinear characteristics of the fiber grating temperature response, resulting in a uniform and minimal overall error distribution across the entire temperature range.

[0042] Through the above linear decoupling process, the stress wavelength interference introduced by pipe wall deformation and vibration can be accurately eliminated, and the true pipe wall temperature without stress deviation can be output, which greatly improves the reliability of temperature measurement data under complex working conditions of deep peak shaving.

[0043] One feasible implementation also includes the following single-grating dual-parameter calibration decoupling implementation path: The specific implementation involves calibrating each fiber Bragg grating sensor separately in a stress-free constant temperature chamber and a constant temperature stress loading stage. The temperature sensitivity coefficient and stress sensitivity coefficient for each measurement point are then fitted, and their ratio is calculated as the stress compensation coefficient specific to that measurement point. During actual measurement, the wavelength offset is multiplied by the compensation coefficient to obtain the stress-induced offset component. This component is then subtracted from the total offset to obtain the temperature value. This solution is suitable for critical heated surface areas with the highest temperature measurement accuracy requirements, such as high-temperature superheaters and reheater outlet tube panels. Its advantages include the highest decoupling accuracy, with single-point temperature errors controllable within ±0.5℃. Its disadvantages include the need for independent calibration of each sensor, resulting in a large workload and a long implementation cycle for batch measurement points.

[0044] III. Using the two-dimensional spatial coordinates of the boiler heating surface as the positioning reference, the stress-compensated temperature value is anchored to the corresponding coordinate position. Based on the spatial distance between adjacent coordinate positions and the difference between the anchored temperature values, the temperature value of the area without measurement points is deduced according to the linear heat transfer relationship, generating a continuous temperature distribution covering the entire heating surface. In practice, before spatial anchoring and extrapolation of temperature values, it is necessary to first construct a two-dimensional coordinate system based on the actual geometric dimensions of the boiler heating surface, match the actual installation positions of all sensing points to the corresponding coordinate nodes, and complete the one-to-one binding of measuring points with spatial positions. Then, based on the anchored discrete temperature points, the temperature of the blank area is extrapolated based on the steady-state heat conduction law of the solid wall, and finally a continuous temperature field of the entire heating surface is generated.

[0045] The step size for dividing the two-dimensional coordinate grid can be selected within the range of 50mm to 200mm, with 100mm being the preferred value. If the step size is too small, the total number of grid nodes increases significantly, and the computational load for temperature extrapolation and visualization rendering increases dramatically, making it difficult to meet the requirements for real-time display. If the step size is too large, a single grid spans multiple heated pipes, which cannot reflect the temperature differences between individual pipes, resulting in insufficient spatial resolution.

[0046] The width of a single tube in the boiler water-cooled wall is approximately 80mm to 120mm. A grid step size of 100mm can be matched with the size of a single tube. Each grid corresponds to a monitoring unit for one tube, balancing the spatial resolution of the temperature field with the computational efficiency of real-time rendering.

[0047] The maximum effective interpolation interval threshold between adjacent measuring points can be set within the range of 200mm to 500mm, with 300mm being the preferred value. If the threshold is too small, the required measuring point density is too high, leading to a significant increase in engineering deployment costs and making it unsuitable for monitoring large-area heated surfaces. If the threshold is too large, the interpolation error in blank areas can reach over 10℃, rendering the extrapolation results worthless. A threshold of 300mm is suitable for conventional measuring point deployment densities, controlling the average interpolation error across the entire area within ±2℃, thus achieving a balance between measuring point cost and extrapolation accuracy.

[0048] The refresh cycle for temperature field visualization can be set within the range of 1 to 10 seconds, with 5 seconds being the preferred value. A refresh cycle that is too short will consume a large amount of system computing resources, squeezing the computing power of other modules such as over-temperature warning and lifespan assessment, and easily causing system lag. A refresh cycle that is too long will fail to capture rapid over-temperature processes in a timely manner, resulting in significant warning lag and rendering real-time monitoring meaningless. The time constant of wall temperature changes caused by boiler combustion adjustments is approximately 3 to 8 seconds; a 5-second refresh cycle can match the dynamic rate of wall temperature change while reserving sufficient computing power to support the parallel operation of multiple modules.

[0049] By configuring the spatial grid and inference parameters as described above, a continuous temperature field with high spatial resolution can be generated under the premise of controlling the computational load, so as to realize an intuitive display of the temperature state of the entire heated surface.

[0050] Regarding the implementation of temperature field extrapolation, in some embodiments, this application preferably employs a global extrapolation method combining coordinate anchoring and bilinear interpolation.

[0051] First, based on the geometric contour dimensions of the boiler heating surface and the total number of sensing points, grid lines are divided at equal intervals in the horizontal and vertical directions to build a two-dimensional spatial coordinate grid for the boiler heating surface.

[0052] Then, based on the actual layout of the sensing points on the boiler heating surface, the corresponding position coordinates of each sensing point are matched from the two-dimensional spatial coordinate grid. The stress-compensated temperature value is then bound to the matched position coordinates one by one to form anchored coordinate points.

[0053] The positioning accuracy of coordinate matching can be controlled within the range of ±5mm to ±20mm, with ±10mm being the preferred value. If the positioning accuracy is too low, the coordinates of the measured point will deviate significantly from the actual position, resulting in misalignment of the temperature field display and inaccurate positioning of abnormal areas. If the positioning accuracy is too high, high-precision 3D scanning mapping is required on-site, significantly increasing implementation costs and workload. A positioning accuracy of ±10mm can meet the positional identification requirements for temperature field visualization while controlling the implementation cost of on-site mapping.

[0054] After coordinate anchoring is completed, based on the row and column arrangement of the anchored coordinate points in the two-dimensional spatial coordinate grid, the spatial straight-line distance and corresponding temperature difference between adjacent coordinate points are extracted. A closed rectangular area is formed with four adjacent anchored coordinate points as vertices, which is used as the sub-region to be deduced.

[0055] The maximum area threshold for the sub-region to be simulated can be set within the range of 0.01㎡ to 0.1㎡, with a preferred value of 0.03㎡. If the sub-region area is too large, the heat flow direction within the region may deviate, the linear heat transfer assumption will no longer hold, and the simulation error will increase significantly. If the sub-region area is too small, the number of sub-regions increases exponentially, leading to a substantial increase in computational load, which cannot meet the real-time requirement of a 5-second refresh. A sub-region area of ​​0.03㎡ ensures the validity of the linear heat transfer assumption while keeping the computational load within the limits of real-time rendering.

[0056] For each sub-region to be simulated, the coordinate point corresponding to the maximum temperature value among the four anchored temperature values ​​is selected as the hot end reference point, and the coordinate point corresponding to the minimum temperature value is selected as the cold end reference point. The line connecting the two coordinate points is used as the dominant heat transfer direction line in the sub-region.

[0057] Calculate the distance and temperature difference between the hot end reference point and the cold end reference point along the dominant direction line, and use the quotient of the two as the temperature change per unit distance along the dominant direction line in this sub-region.

[0058] The maximum allowable temperature change per unit distance can be set within the range of 50℃ / m to 200℃ / m, with 100℃ / m being the preferred value. If the allowable value is too high, extreme extrapolation results that do not conform to heat transfer laws may occur, triggering false over-temperature warnings; if the allowable value is too low, it will smooth out the true local thermal deviations, losing information about abnormal operating conditions. A threshold of 100℃ / m conforms to the normal thermal deviation range of the boiler's heating surface, filtering out abnormal extrapolation values ​​while retaining the true characteristics of the heat load distribution.

[0059] Finally, for all coordinate points in the sub-region where no measuring points are set up, measure the projected distance from the point to the hot end reference point along the dominant direction of heat transfer. Multiply the projected distance by the temperature change per unit distance to obtain the temperature decrease at that point. Subtract the temperature decrease from the temperature value of the hot end reference point to get the estimated temperature value of the unset coordinate point.

[0060] After traversing all sub-regions to be simulated and completing the global temperature calculation, the simulated temperature values ​​and the measured temperature values ​​of the anchored coordinate points are integrated into the same spatial coordinate system to obtain a continuous temperature distribution covering the entire heated surface.

[0061] Through the above coordinate anchoring and linear heat transfer extrapolation, a smooth and continuous global temperature field can be generated based on discrete measuring points, filling the temperature data in the monitoring blind spots and providing global data support for subsequent heat load analysis and over-temperature early warning.

[0062] In one feasible implementation, in embodiment III above, a bilinear interpolation global extrapolation scheme can also be used to generate a continuous temperature distribution covering the entire heated surface.

[0063] Specifically, the heated surface is divided into equally spaced rectangular grids, and rectangular units are formed by four adjacent anchoring measurement points as vertices. The temperature value is calculated for any coordinate point within the unit using a bilinear interpolation formula, and the temperature is filled across the entire domain by traversing each unit.

[0064] This scheme is applicable to water-cooled walls and superheater tube panels with uniformly distributed measuring points and regular tube arrays. Its advantages are simple calculation logic, fast processing speed, smooth temperature field transition, and suitability for real-time visualization rendering; its disadvantage is that the interpolation accuracy decreases slightly in the edge regions of sudden changes in heat load.

[0065] In one feasible implementation, in embodiment III above, a triangular mesh linear interpolation scheme can also be used to generate a continuous temperature distribution covering the entire heated surface.

[0066] Specifically, based on the actual location of discrete measuring points, the Delaunay triangulation algorithm is used to generate unstructured triangular meshes. Using the temperatures of the three vertices of each triangle as a reference, the temperature of any point inside the triangle is calculated through planar linear interpolation.

[0067] This solution is suitable for heated surface areas with irregularly distributed measuring points and localized densities or gaps, such as furnace corners and through-wall pipe locations. Its advantages include adaptability to non-uniform measuring point distribution, no need for pre-dividing regular grids, and higher interpolation accuracy in locally densified areas. Its disadvantages include additional computational overhead for grid generation and slightly lower rendering smoothness compared to the rectangular grid solution.

[0068] In one feasible implementation, in embodiment III above, a heat transfer mechanism-assisted deduction scheme can also be used to generate a continuous temperature distribution covering the entire heated surface.

[0069] Specifically, in the interpolation calculation, mechanistic parameters such as pipe wall thermal conductivity, wall thickness, and flue gas side convective heat transfer coefficient are introduced to establish a one-dimensional radial heat transfer model. The temperature on the fire side is deduced by combining the temperature on the back side of the measuring point, and then the temperature distribution of the entire wall surface is generated by combining axial linear interpolation.

[0070] This solution is applicable to high-temperature heated surfaces where sensors can only be installed on the unexposed side, and the actual wall temperature on the fire-facing side needs to be calculated. Its advantages include overcoming installation location limitations and obtaining the full-circumference temperature distribution of the heated surface, more closely reflecting the actual heat transfer conditions. Its disadvantage is the need for accurate thermal parameters; parameter deviations directly affect the calculation accuracy.

[0071] IV. Based on the continuous temperature distribution, calculate the temperature gradient value and heating rate value of each region, compare the temperature gradient value and heating rate value with the preset classification threshold, and output a classification warning signal that matches the degree of overheating. In practice, before making a graded early warning judgment, it is necessary to first calculate the temperature gradient in the spatial dimension and the heating rate in the time dimension based on the continuous temperature field data, region by region, according to the grid unit. The two indicators are then compared with the preset multi-level thresholds one by one to determine the risk level of each area, output the corresponding level of early warning signal and locate the abnormal area.

[0072] The absolute temperature thresholds for tiered early warning systems are set according to the proportion of the pipe material's allowable temperature. The proportion for Level 1 early warning is 80% to 90%, preferably 85%; for Level 2, it is 90% to 95%, preferably 92%; and for Level 3, it is 95% to 100%, preferably 98%. Setting the Level 1 threshold too low will result in excessively frequent warnings, interfering with normal operations; setting it too high will lead to delayed warnings, failing to provide early intervention.

[0073] Taking the permissible temperature of 12Cr1MoV pipe (580℃) as an example, 85% corresponds to 493℃, which is the upper limit for long-term safe operation; 92% corresponds to 533.6℃, ​​where short-term overheating will accelerate material creep damage; and 98% corresponds to 568.4℃, which is close to the material limit, and continuous overheating poses a risk of pipe bursting. This classification setting can achieve risk gradient identification, ensuring both early warning and controlling the false alarm rate.

[0074] The temperature rise rate thresholds for tiered early warning systems are as follows: Level 1 warning ranges from 2℃ / min to 5℃ / min, preferably 3℃ / min; Level 2 warning ranges from 5℃ / min to 10℃ / min, preferably 8℃ / min; and Level 3 warning ranges from 10℃ / min to 20℃ / min, preferably 15℃ / min. If the threshold is too low, normal peak-shaving temperature fluctuations may be misjudged as abnormal, increasing the number of invalid warnings; if the threshold is too high, rapid overheating processes cannot be detected in time, and the warning will lack advance warning capability.

[0075] When the boiler is operating under normal variable load, the wall temperature rise rate is approximately 1℃ / min to 2℃ / min. 3℃ / min corresponds to abnormal heat load deviation, 8℃ / min corresponds to significant combustion disturbance, and 15℃ / min corresponds to extreme operating conditions. This classification can effectively distinguish between normal fluctuations and abnormal overheating, and the early warning lead time can reach more than 30 seconds.

[0076] The temperature gradient thresholds for tiered early warning systems are as follows: Level 1 warning ranges from 5℃ / m to 10℃ / m, preferably 8℃ / m; Level 2 warning ranges from 10℃ / m to 20℃ / m, preferably 15℃ / m; and Level 3 warning ranges from 20℃ / m to 30℃ / m, preferably 25℃ / m. If the gradient threshold is too low, thermal deviations during normal combustion will frequently trigger warnings; if the threshold is too high, latent thermal stress damage caused by uneven local heat loads will not be identified.

[0077] Under normal combustion conditions, the transverse temperature gradient of the pipe wall is approximately 3℃ / m to 5℃ / m. 8℃ / m corresponds to a localized high heat load, 15℃ / m corresponds to a significant thermal deviation caused by flame deflection, and 25℃ / m corresponds to a severe thermal deviation, which can easily lead to pipe wall fatigue damage. This classification can supplement the identification of thermal deviation risks that absolute temperature cannot cover, expanding the dimensions of early warning.

[0078] By configuring the above-mentioned graded thresholds, it is possible to achieve graded risk identification from three dimensions: absolute temperature, heating rate, and thermal deviation, taking into account both the lead time and accuracy of early warning.

[0079] Regarding the determination method for graded early warning, in some embodiments, this application preferably adopts a three-level early warning mechanism that combines grid cell division with multi-indicator joint determination.

[0080] First, the boiler heating surface is divided into multiple independent monitoring areas according to the grid cells in the two-dimensional spatial coordinate grid. For each area, the temperature difference and spatial straight-line distance of all adjacent coordinate points are extracted. The local temperature gradient value of the corresponding coordinate point pair is obtained by dividing the temperature difference by the spatial straight-line distance. The maximum value of all local temperature gradient values ​​in the area is taken as the final temperature gradient value of the area.

[0081] Then, the continuous temperature field data generated at the previous moment is retrieved, and the temperature data of the two moments are matched one-to-one with the same coordinate point. The temperature change of each coordinate point between the two moments is calculated, and the change is divided by the time interval between the two moments to obtain the local heating rate value of that point. The maximum value of all local heating rate values ​​in the region is taken as the final heating rate value of the region.

[0082] The time window for calculating the heating rate can be set from 1 minute to 10 minutes, with a preferred value of 5 minutes. If the time window is too short, the rate calculation is greatly affected by instantaneous temperature fluctuations, resulting in drastic numerical jumps and failing to reflect the true heating trend. If the time window is too long, the rate calculation lags significantly, failing to capture the rapid heating process in a timely manner. A 5-minute time window smooths out instantaneous fluctuations while ensuring timely identification of the heating trend, adapting to the temperature change characteristics of deep peak-shaving conditions.

[0083] A three-tiered warning system is preset, with each tier corresponding to a set of upper limits for both the temperature gradient and the temperature rise rate. The higher the warning level, the lower the corresponding upper limit value. For each monitoring area, its temperature gradient value is compared sequentially with the upper limits of the three-tiered warning system from highest to lowest, and its temperature rise rate value is also compared sequentially with the upper limits of the three-tiered warning system from highest to lowest. If either the temperature gradient value or the temperature rise rate value of an area reaches the threshold for a certain level, the area is determined to have reached the corresponding warning level.

[0084] If a region simultaneously meets the warning conditions for multiple levels, the highest warning level will be used as the output. The threshold for the number of consecutive triggers for warning determination can be set between 1 and 5 times, with 3 times being the preferred value. Too few triggers make it susceptible to false alarms due to transient noise interference, reducing the reliability of the warning; too many triggers result in a delayed warning response, missing the best intervention opportunity. Triggering a warning only when the threshold conditions are met for three consecutive sampling periods filters out transient noise interference and keeps the warning lag within 30 seconds, balancing false alarm resistance and timely response.

[0085] After the level determination is completed, a warning signal of the corresponding level is output, and the specific location of the abnormal area in the two-dimensional spatial coordinate grid is marked simultaneously and pushed to the operation monitoring interface. The spatial resolution of the warning location can be selected between single grid cell and multi-region merging granularity, with single grid cell positioning being preferred. A single grid cell corresponds to the monitoring granularity of a single pipe, which can accurately locate the over-temperature pipe position while controlling the order of magnitude of the warning information.

[0086] Through the above multi-dimensional joint judgment and processing, it can comprehensively cover three types of risks: absolute overheating, rapid temperature rise, and local thermal deviation, realizing the upgrade from passive alarm to active early warning, and providing accurate risk guidance for operation adjustment.

[0087] In one feasible implementation, in embodiment IV above, a three-level early warning scheme based on dual indicators can also be adopted.

[0088] Specifically, for each region, three indicators are calculated simultaneously: absolute temperature, heating rate, and temperature gradient. If any one indicator reaches a certain warning threshold, a warning for that level is triggered. When multiple indicators are triggered, the highest level is output, and abnormal location and over-temperature amplitude data are pushed simultaneously.

[0089] This solution is applicable to general scenarios of daily operation monitoring, taking into account multiple risks such as absolute over-temperature, rapid temperature rise, and thermal deviation. Its advantages are comprehensive coverage of risk dimensions, low false negative rate, and clear warning levels; its disadvantage is that the logic triggered by a single indicator can lead to a higher frequency of warnings, requiring operators to verify them.

[0090] In one feasible implementation, in embodiment IV above, a multi-factor weighted comprehensive early warning scheme can also be adopted.

[0091] Specifically, weighting coefficients are assigned to the absolute temperature value, heating rate, and temperature gradient, and a comprehensive risk score is calculated using weighted averages. Warning levels are then categorized based on the score range. This scheme is suitable for mature generating units with extensive operating experience where a reduction in false alarm rates is needed, minimizing operational interference from invalid warnings. Its advantages include the ability to adapt to the operating characteristics of different generating units through weight adjustments, reducing the false alarm rate by 40%. Its disadvantages include the need for weighting coefficients to be calibrated based on historical data, requiring a certain data accumulation period for new generating units.

[0092] like Figure 2 The diagram shown is a functional block diagram of a boiler multispectral temperature measurement and calibration system based on FBG according to an embodiment of the present invention.

[0093] The FBG-based boiler multispectral temperature measurement and calibration system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the FBG-based boiler multispectral temperature measurement and calibration system 100 may include a reflectance spectrum demodulation module 101, a stress-temperature decoupling module 102, a temperature field spatial extrapolation module 103, and a graded early warning judgment module 104. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0094] In this embodiment, the functions of each module / unit are as follows: The reflection spectrum demodulation module 101 is used to acquire the reflection spectrum signal of the sensing point in the fiber grating array of the boiler heating surface, and demodulate the reflection spectrum signal to obtain the center wavelength offset sequence of the sensing point. The stress-temperature decoupling module 102 is used to perform stress-temperature decoupling on the center wavelength offset sequence according to the stress compensation coefficient of the sensing point, so as to obtain the temperature value after stress compensation. The temperature field spatial extrapolation module 103 is used to anchor the stress-compensated temperature value to the corresponding coordinate position using the two-dimensional spatial coordinates of the boiler heating surface as the positioning reference, and extrapolate the temperature value of the area without measurement points according to the spatial distance between adjacent coordinate positions and the difference of the anchored temperature value, and generate a continuous temperature distribution covering the entire heating surface. The graded early warning determination module 104 is used to calculate the temperature gradient value and heating rate value of each region according to the continuous temperature distribution, compare the temperature gradient value and heating rate value with the preset graded threshold, and output a graded early warning signal that matches the degree of overheating.

[0095] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0096] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0097] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0098] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0099] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A boiler multispectral temperature measurement calibration method based on FBG, characterized in that, The method includes: Ⅰ. Obtain the reflection spectrum signal of the sensing point in the fiber grating array of the boiler heating surface, and demodulate the reflection spectrum signal to obtain the center wavelength offset sequence of the sensing point; II. Decouple the center wavelength offset sequence from stress and temperature based on the stress compensation coefficient of the sensing point to obtain the temperature value after stress compensation. III. Using the two-dimensional spatial coordinates of the boiler heating surface as the positioning reference, the stress-compensated temperature value is anchored to the corresponding coordinate position. Based on the spatial distance between adjacent coordinate positions and the difference between the anchored temperature values, the temperature value of the area without measurement points is deduced according to the linear heat transfer relationship, generating a continuous temperature distribution covering the entire heating surface. IV. Based on the continuous temperature distribution, calculate the temperature gradient value and heating rate value of each region, compare the temperature gradient value and heating rate value with the preset grading threshold, and output a graded early warning signal that matches the degree of overheating.

2. The boiler multispectral temperature measurement calibration method based on FBG as described in claim 1, characterized in that, The process of acquiring the reflection spectral signal of the sensing point in the fiber optic grating array of the boiler heating surface and demodulating the reflection spectral signal to obtain the center wavelength offset sequence of the sensing point includes: The reflection spectrum signal of the sensing point in the fiber grating array of the boiler heating surface is acquired, and the reflection spectrum signal is split according to wavelength to generate a light intensity signal group corresponding to each wavelength position. Based on each of the sensing points, the wavelength position corresponding to the light intensity peak value is selected from the light intensity signal group as the current center wavelength of the sensing point. By comparing the current center wavelength with the initial center wavelength of the sensing point, the wavelength offset of the sensing point is obtained. The wavelength offsets are collected according to the order of the measurement points in the fiber grating array to obtain the center wavelength offset sequence of the sensing points.

3. The boiler multispectral temperature measurement calibration method based on FBG as described in claim 1, characterized in that, The step of performing stress-temperature decoupling on the center wavelength offset sequence based on the stress compensation coefficient of the sensing point to obtain the stress-compensated temperature value includes: The sensing point was placed in a stress-free constant temperature environment and a constant temperature stress loading environment, respectively. The center wavelength offset was recorded under the two environments, and temperature response offset and stress response offset were generated respectively. For each of the sensing points, the ratio of the temperature response offset to the corresponding temperature change is taken as the temperature sensitivity value of the sensing point, and the ratio of the stress response offset to the corresponding stress change is taken as the stress sensitivity value of the sensing point. The ratio of the stress sensitivity value to the temperature sensitivity value at the same measuring point is used as the stress compensation coefficient of the sensing point, and the center wavelength offset of the sensing point is multiplied by the stress compensation coefficient of the corresponding measuring point to obtain the stress-induced offset component of the sensing point. Remove the stress-induced offset component from the center wavelength offset of the sensing point, and retain the temperature-induced offset component of the sensing point. The temperature-induced offset component is compared with the calibrated temperature wavelength mapping relationship of the sensing point, and the corresponding temperature value is extracted as the stress-compensated temperature value of the sensing point.

4. The boiler multispectral temperature measurement calibration method based on FBG as described in claim 1, characterized in that, The step of using the two-dimensional spatial coordinates of the boiler's heating surface as a positioning reference to anchor the stress-compensated temperature value to the corresponding coordinate position includes: Based on the geometric contour dimensions of the boiler heating surface and the total number of sensing points, grid lines are divided at equal intervals in the horizontal and vertical directions to obtain the two-dimensional spatial coordinate grid of the boiler heating surface. Based on the actual placement of the sensing points on the boiler heating surface, the corresponding position coordinates of the sensing points are matched from the two-dimensional spatial coordinate grid. The stress-compensated temperature value of the sensing point is bound to the matched position coordinates to obtain the anchored coordinate point of the position coordinates.

5. The boiler multispectral temperature measurement calibration method based on FBG as described in claim 4, characterized in that, The step of extrapolating the temperature value of the unmeasured area based on the difference between the spatial distance between adjacent coordinate positions and the anchored temperature value, according to the linear heat transfer relationship, includes: Based on the row and column arrangement of the anchored coordinate points in the two-dimensional spatial coordinate grid, the spatial straight-line distance between the anchored coordinate points and adjacent coordinate points is extracted to generate the distance value of adjacent point pairs; Extract the difference between the anchored temperature values ​​carried between the anchored coordinate point and the adjacent coordinate point to generate the temperature difference value of the adjacent point pair; The distance value and the temperature difference value of the adjacent point pairs are associated accordingly, and the closed region enclosed by the adjacent point pairs with the distance value and the temperature difference value is taken as the sub-region to be deduced.

6. The boiler multispectral temperature measurement calibration method based on FBG as described in claim 5, characterized in that, The method of extrapolating the temperature value of the unmeasured area based on the difference between the spatial distance between adjacent coordinate positions and the anchored temperature value according to the linear heat transfer relationship also includes: For each of the sub-regions to be extrapolated, the anchored coordinate point where the maximum temperature value is located is selected from the anchored temperature values ​​as the hot end reference point, and the anchored coordinate point where the minimum temperature value is located is selected as the cold end reference point. Connect the coordinates of the hot end reference point and the cold end reference point in the two-dimensional spatial coordinate grid to obtain the dominant heat transfer direction line in the sub-region to be deduced. Extract the difference between the distance spanned by the hot end reference point and the cold end reference point along the dominant direction of heat transfer and the anchored temperature value. Based on the heat transfer relationship in the sub-region to be simulated, where the heat decreases uniformly along the dominant heat transfer direction line, the quotient of the difference and the span distance is taken as the temperature change per unit distance in the sub-region to be simulated along the dominant heat transfer direction line.

7. The boiler multispectral temperature measurement calibration method based on FBG as described in claim 6, characterized in that, The simulation of temperature values ​​in areas without measured points generates a continuous temperature distribution covering the entire heated surface, including: For the unmarked coordinate points in the sub-region to be simulated, measure the projected distance from the unmarked coordinate points to the hot end reference point along the dominant heat transfer direction; The product of the projected distance and the temperature change per unit distance is used as the temperature decrease of the unmarked coordinate point relative to the hot end reference point. The difference between the anchored temperature value carried at the hot end reference point and the temperature decrease is the estimated temperature value of the unmarked coordinate point. The predicted temperature values ​​of the unmarked coordinate points in all the sub-regions to be predicted are combined with the anchored temperature values ​​carried at the anchored coordinate points into the same spatial coordinate system to obtain a continuous temperature distribution covering the entire heated surface.

8. The boiler multispectral temperature measurement calibration method based on FBG as described in claim 7, characterized in that, The step of calculating the temperature gradient and heating rate values ​​for each region based on the continuous temperature distribution includes: The boiler heating surface is divided into multiple regions according to the grid cells in the two-dimensional spatial coordinate grid. For each region, the temperature difference and spatial straight-line distance between all adjacent coordinate points in the region are extracted. The quotient of the temperature difference divided by the spatial straight-line distance is used as the local temperature gradient value of the coordinate point pair, and the maximum value of all local temperature gradient values ​​in the region is taken as the temperature gradient value of the region. Retrieve the continuous temperature distribution of the region obtained at the previous moment, and map the continuous temperature distribution at the previous moment to the continuous temperature distribution at the current moment according to the same coordinate point; Extract the temperature change and time interval between two time points for each coordinate point, and use the quotient of the temperature change divided by the time interval as the local heating rate value at the coordinate point. Take the maximum value of all the local heating rate values ​​in the region as the heating rate value of the region.

9. The boiler multispectral temperature measurement calibration method based on FBG as described in claim 8, characterized in that, The step of comparing the temperature gradient value and the heating rate value with a preset grading threshold and outputting a graded early warning signal matching the degree of overheating includes: Three warning levels are preset, and each warning level corresponds to a set of upper limit values ​​for gradient and upper limit values ​​for heating rate. The upper limit values ​​for gradient and heating rate decrease as the warning level increases. The temperature gradient value of each region is compared with the upper limit of the gradient of the three-level warning level one by one, and the heating rate value is compared with the upper limit of the heating rate of the three-level warning level one by one. When the temperature gradient value of the area is greater than the upper limit of the gradient value of the current warning level, or the temperature rise rate value of the area is greater than the upper limit of the temperature rise rate of the current warning level, the area is determined to have reached the current warning level. The warning levels are determined step by step from high to low. If the region meets the conditions of any of the warning levels, the warning signal corresponding to the warning level is output, and the spatial position of the region in the two-dimensional spatial coordinate grid is located.

10. A boiler multispectral temperature measurement and calibration system based on FBG, characterized in that, The system for implementing the FBG-based boiler multispectral temperature measurement calibration method of claim 1 includes: The reflection spectrum demodulation module is used to acquire the reflection spectrum signal of the sensing point in the fiber grating array of the boiler heating surface, and demodulate the reflection spectrum signal to obtain the center wavelength offset sequence of the sensing point. The stress-temperature decoupling module is used to perform stress-temperature decoupling on the center wavelength offset sequence according to the stress compensation coefficient of the sensing point, so as to obtain the temperature value after stress compensation. The temperature field spatial extrapolation module is used to anchor the stress-compensated temperature value to the corresponding coordinate position using the two-dimensional spatial coordinates of the boiler heating surface as the positioning reference, and extrapolate the temperature value of the area without measurement points according to the spatial distance between adjacent coordinate positions and the difference of the anchored temperature value, generating a continuous temperature distribution covering the entire heating surface. The graded early warning determination module is used to calculate the temperature gradient value and heating rate value of each region based on the continuous temperature distribution, compare the temperature gradient value and heating rate value with the preset graded threshold, and output a graded early warning signal that matches the degree of overheating.

Images