Estimation of the temperature of a steel product

Abstract

The present invention relates to a method for estimating the temperature of a steel product, the method comprising: a calibration step, wherein the intensity at five wavelengths ranging from 0.9 μm to 2.1 μm is recorded for several measurement conditions and a spectral attenuation coefficient is calculated; a measurement step, wherein the intensity at the five wavelengths is recorded and a spectral attenuation coefficient is calculated for several temperatures; and a comparison step, wherein a probability test is performed to estimate the temperature of the steel product.

Description

[0001] This invention relates to a method for estimating the temperature of steel products undergoing cooling. This method is particularly advantageous when water is present on the steel product. For example, in steelmaking, the claimed method can be applied during the secondary cooling process of continuous casting or in the output roller table of a hot rolling mill.

[0002] During the manufacturing process of steel strip, from casting to coiling, the steel undergoes several cooling operations. These operations typically involve spraying water onto the steel. This can result in the formation of a water film on the strip surface.

[0003] Cooling operations typically involve models that adjust cooling power. These models take temperature as input data. Therefore, accurate temperature knowledge during cooling operations is crucial for reliable cooling management.

[0004] The temperature of steel products is typically measured using a pyrometer that measures radiation intensity. However, the measured radiation intensity is affected by the presence of a medium between the product and the pyrometer (such as a layer of water on the product). For example, when there is no interference between the steel surface and the environment, the accuracy is approximately ±10°C. When there is water on the steel surface or when there is fog between the steel and the pyrometer, the measurement error can be as high as 100°C. When there is water on the steel and fog between the steel and the pyrometer, the measurement error can be as high as 200°C.

[0005] Therefore, it is necessary to improve the accuracy of measuring steel temperature when there is water on the steel and / or fog between the steel and the measuring device.

[0006] EP 2889594 discloses a method for accurately measuring the surface temperature of steel materials during water cooling. Radiation in the wavelength bands of 0.7 μm to 0.9 μm, 1.0 μm to 1.2 μm, and 1.6 μm to 1.8 μm is recorded. A pyrometer is used to measure the steel temperature. Furthermore, an optical glass is placed between the pyrometer and the steel, positioned at a defined distance from the steel. The optical glass is also positioned such that during the cooling process, cooling water enters the space between the steel and the optical glass to maintain a stable surface tension. Therefore, the medium between the optical glass and the steel is known. The measured radiation intensity is then corrected using a coefficient related to the gap between the steel and the optical glass. This allows for reduction of temperature measurement errors caused by the absorption or scattering of radiant energy by the water.

[0007] The purpose of this invention is to provide a method for improving the accuracy of temperature measurement of steel strip during cooling operations.

[0008] This objective is achieved by providing the method according to any one of claims 1 to 12.

[0009] Other features and advantages will become apparent from the following detailed description of the invention.

[0010] Figure 1 Implementation of steps Ai to B.ii. of the claimed method is shown.

[0011] Figure 2 An embodiment of the absorption spectrum of water according to wavelength is shown.

[0012] Figure 3 An embodiment of the absorption spectrum of water vapor according to wavelength is shown.

[0013] Figure 4 Several spectral attenuation coefficients C calculated during the calibration procedure are shown. CALIB The implementation method.

[0014] Figure 5 The comparison results of the steel strip temperature measured by thermocouples and estimated by the method required for protection are shown.

[0015] This invention relates to estimating the temperature T of steel products having temperatures ranging from 300°C to 1600°C. REAL The method includes:

[0016] A. Calibration procedure, which includes the following steps:

[0017] i. Measuring with the aid of a sensor from a source with a known temperature (T) REF The intensity (I) of the radiation emitted by a reference object under measurement conditions at five wavelengths (λ) ranging from 0.9 μm to 2.1 μm, wherein one wavelength is from 0.9 μm to 1.35 μm, one wavelength is from 1.35 μm to 1.55 μm, one wavelength is from 1.55 μm to 1.85 μm, one wavelength is from 1.85 μm to 2.05 μm, and one wavelength is from 2.05 μm to 2.1 μm, and the measurement conditions are determined by the emissivity (ε) of the reference object. REF ) and the transmittance (α) of the medium between the reference object and the sensor REF The reference object is a steel product, as characterized by the reference object.

[0018] ii. Calculate the spectral attenuation coefficient C using the measured intensities (I) at the five wavelengths. CALIB ,

[0019]

[0020] Where, P(λ,T) REF ) is based on Planck's law, which states that a blackbody in thermal equilibrium exhibits thermal equilibrium at wavelength (λ) and temperature (T). REF The spectral density of electromagnetic radiation emitted at point )

[0021] iii. For the reference emissivity (ε) REF ) and the transmittance (α) of the medium between the reference object and the sensor REF N CALIB Repeat steps i. and ii. in different combinations to obtain N. CALIB Spectral attenuation coefficient, N CALIB It is an integer greater than 2.

[0022] B. Measurement procedure, which includes the following steps:

[0023] i. Measure the intensity I of the radiation emitted by the steel product at five wavelengths (λ) ranging from 0.9 μm to 2.1 μm.

[0024] ii. For N in the range of 300℃ to 1600℃ T The temperature (Tj) and N are calculated for the five wavelengths. T spectral attenuation coefficient C 计算 Tj, N T It is an integer from 2 to 1300.

[0025]

[0026] in,

[0027] -P(λ,T J () is based on Planck's law, which states that a blackbody in thermal equilibrium exists at wavelength λ and temperature T. J The spectral density of the electromagnetic radiation emitted at that location.

[0028] C. The comparison step includes the following steps:

[0029] i. Execute for C CALIB The most likely C was found among them. 计算 Probability test of Tj

[0030] ii. Set the temperature T of the steel product REAL The estimate is equal to the most likely C. 计算 T J Temperature T J .

[0031] The steps of the process are as follows: Figure 1 As shown in the image.

[0032] Steel products can be of any type, such as strips, hoops, or slabs. The temperature of the steel product is unknown; however, based on the process steps in which the measurements are performed, those skilled in the art know the range within which the temperature should be. For example, in the cooling process of steel strip after hot rolling, the temperature of the steel strip generally falls between 300°C and 1100°C.

[0033] In calibration step Ai, the intensity of radiation emitted by the reference steel product can be measured by any suitable means. These intensities are preferably measured by a hyperspectral imaging device.

[0034] In calibration step Ai, the reference steel product preferably has a composition similar to that of the steel product whose temperature was estimated. Even more preferably, the reference steel product has the same grade as the steel product being analyzed.

[0035] In calibration step Ai, the temperature of the reference steel product can be measured by any means. Preferably, a thermocouple is used to measure this temperature.

[0036] In calibration step A.ii., the spectral attenuation coefficient C can be calculated by dividing the intensity I of each recorded value by P(λ,Tj) at the measurement temperature. CALIB .

[0037] Steps Ai and A.ii were repeated N times. CALIB Each time, the measurement corresponds to a new measurement condition. The measurement conditions are determined by the emissivity (ε) of the reference object. REF ) and the transmittance (α) of the medium between the reference object and the sensor. REF The combination of ) is used to define it. This allows for obtaining various C values ​​for different measurement conditions. CALIB N CALIB The larger the value, the more accurate the estimate.

[0038] In other words, steps i and ii of calibration step A are for the emissivity (ε) of the steel product. REF ) and measured transmittance (α) REF ) repeats in various combinations. ε REF It varies depending on several factors, such as the temperature of the reference steel product and surface properties (e.g., the presence of oil). α REF It depends on the medium between the sensor and the reference steel product, such as the thickness of the water layer on the reference steel product.

[0039] However, for calculating C CALIB , ε REF and α REF The value does not need to be known.

[0040] In measurement step Bi, a hyperspectral imaging device is preferably used to measure the intensity emitted by the steel product. The wavelength at which the intensity is measured in step Bi is the same as the wavelength at which the intensity is measured in step Ai.

[0041] In measurement step B.ii., the number of temperatures (N) used to calculate the spectral attenuation coefficient is calculated. T The larger the value, the more accurate the estimate.

[0042] Furthermore, the intensity of the measurement can be adjusted using a transfer function in both the calibration and measurement steps.

[0043] Measuring the radiant intensity at the five wavelengths allows for the use of a reduced number of intensities to describe the spectral shape of the combination of product emissivity and medium transmittance. These five wavelengths include one from 0.9 μm to 1.35 μm, one from 1.35 μm to 1.55 μm, one from 1.55 μm to 1.85 μm, one from 1.85 μm to 2.05 μm, and one from 2.05 μm to 2.1 μm. In fact, within the range from 0.9 to 2.1 μm, as... Figure 1 As shown, the absorption spectrum of water exhibits two peaks, one from 1.35 μm to 1.55 μm and the other from 1.85 μm to 2.05 μm.

[0044] In comparison step C, a probability test is performed to determine the probability of the result in C. CALIB The most likely C was found among them. 计算 T J It can execute commands that are allowed in C. CALIB The most likely C was found among them. 计算 T J Any method.

[0045] This method allows for improved accuracy in temperature estimation of steel products because the estimation takes into account the effects of measurement conditions, such as the presence of water on the steel product.

[0046] Preferably, the cooling treatment is performed during or after hot rolling, and the steel product has a temperature from 300°C to 1100°C, and wherein, in step B, T J The temperature ranges from 300°C to 1100°C. This cooling is typically performed in the output roller conveyor.

[0047] Preferably, the cooling treatment is performed during continuous casting or in the presence of steel products at temperatures ranging from 800°C to 1600°C, and wherein, in step B, T J The temperature range is from 800℃ to 1600℃.

[0048] Preferably, in steps A)i. and B)i., the radiation intensity of eight wavelengths (λ) ranging from 0.9 μm to 2.1 μm is measured, wherein one wavelength is from 0.9 μm to 1.11 μm, one wavelength is from 1.11 μm to 1.15 μm, one wavelength is from 1.15 μm to 1.35 μm, one wavelength is from 1.35 μm to 1.55 μm, one wavelength is from 1.55 μm to 1.85 μm, one wavelength is from 1.85 μm to 2.05 μm, one wavelength is from 2.05 μm to 2.07 μm, and one wavelength is from 2.07 μm to 2.1 μm; and in steps A)ii. and B)ii, the spectral attenuation coefficients of the eight wavelengths are calculated. Measuring the radiation intensity of these eight wavelengths improves the accuracy of the measurement results when there is a medium such as vapor between the steel product and the sensor. Figure 2 As shown, the ranges from 1.11 μm to 1.15 μm and from 2.05 μm to 2.07 μm correspond to the peaks of the vapor absorption spectrum. Furthermore, these two peaks do not match the peaks of the water absorption spectrum.

[0049] Preferably, in steps A)i. and B)i., the radiant intensity at five additional wavelengths ranging from 0.9 μm to 2.1 μm is measured, and in step steps A)ii. and B)ii., the spectral attenuation coefficients of the eight wavelengths and the five additional wavelengths are calculated. The 13 wavelengths are preferably uniformly distributed within the range, meaning that the 13 wavelengths are separated by intervals of 0.1 μm (0.9, 1.0, ..., 2.0, 2.1).

[0050] Preferably, in steps A)i. and B)i., the radiation intensity at 42 additional wavelengths ranging from 0.9 μm to 2.1 μm is measured, and in step steps A)ii. and B)ii., the spectral attenuation coefficients of the 8 wavelengths and the 42 additional wavelengths are calculated.

[0051] Preferably, in steps A)i. and B)i., the radiation intensity at 92 additional wavelengths ranging from 0.9 μm to 2.1 μm is measured, and in step steps A)ii. and B)ii., the spectral attenuation coefficients of the 8 wavelengths and 92 additional wavelengths are calculated.

[0052] Preferably, N CALIB It is an integer from 2 to 1000, and preferably an integer from 20 to 1000.

[0053] Preferably, in step C)i., the probability test includes testing the C that defines the main component. CALIB Dimensionality reduction is performed. Even more preferably, in step C), the dimensionality reduction is performed using principal component analysis.

[0054] Preferably, in step C)i., the probability test includes the C CALIB Projection in the probability model. Even more preferably, in step C)i., the probability model is a Gaussian mixture model.

[0055] Embodiments of the present invention

[0056] To assess the accuracy of the claimed protection method, the temperature of the steel strip being cooled in the output roller conveyor was measured using thermocouples and the temperature was estimated using this method.

[0057] A. Calibration Steps

[0058] During the calibration process, the intensity of radiation emitted by the steel strip, which was cooled in the output roller conveyor, was recorded using a hyperspectral imaging device. More precisely, the intensity was recorded at 256 wavelengths ranging from 1.1 μm to 2.1 μm. These 256 wavelengths are uniformly distributed within this range. The temperature of the steel strip was measured using thermocouples.

[0059] The calibration procedure consisted of six tests lasting between 5 and 15 minutes, during which the radiation intensity was measured between 10 and 100 times per second. Thus, thousands of measurement conditions were recorded. These conditions occurred during cooling (in the presence of water droplets, mist, and water) and before / after cooling (in the absence of water / mist), and were applied to steel strips with various degrees of oxidation.

[0060] Then, several spectral attenuation coefficients C are calculated using the spectral density of electromagnetic radiation emitted by a blackbody in thermal equilibrium, the temperature measured by a thermocouple, and the transfer function of the hyperspectral imaging device. CALIB C CALIB Each of them has 256 values, one value corresponding to each of the 256 wavelengths. Figure 4 C was drawn in CALIB .

[0061] B. Measurement

[0062] During the measurement step, the intensity of radiation emitted by the steel strip, which was cooled in the output roller conveyor, was recorded using a hyperspectral imaging device. More precisely, the intensity at the same 256 wavelengths as in the calibration step was recorded.

[0063] Furthermore, 261 temperatures ranging from 300°C to 1600°C are defined as possible temperatures for the steel strip being measured, therefore N T =261. These temperatures are separated by 5°C intervals: (300°C, 305°C, 310°C...1595°C, 1600°C).

[0064] Then, for each of the previously defined 261 temperatures, the measured radiant intensity at the 256 wavelengths is divided by the value of Planck's law and multiplied by the transfer function. Thus, the 261 spectral attenuation coefficients C are calculated. 计算 T J Each of the 261 spectral attenuation coefficients has 256 values, one value corresponding to each measured wavelength and associated temperature.

[0065] C. Comparison

[0066] The probability test in this implementation includes 10 steps.

[0067] 1) For all spectral attenuation coefficients C obtained during the calibration step CALIB Principal component analysis (PCA) was performed to obtain three principal PCA components (PC1, PC2, and PC3). Then, these three principal components were used to analyze each C in the database. CALIB Approximation is performed.

[0068] 2) Project the spectral attenuation coefficients into the PCA space defined by the three principal components previously defined.

[0069] 3) Project the density of points in the PCA space from step 2) into the Gaussian mixture model.

[0070] 4) The 261 C 计算 T J Each of the components is projected into the PCA space defined by the three previously defined principal components.

[0071] 5) For 261 C 计算 T J For each of them, an approximate spectral attenuation coefficient is reconstructed based on its coordinates in the PCA space defined in 4), and a first likelihood factor (Like1) is defined by comparing the original spectral attenuation with its reconstruction based on the PCA coefficients.

[0072] 6) The 261 C 计算 T J Each of the projections is assigned to the Gaussian mixture model defined in 3), and a second likelihood factor (Like2) is determined. Like2 represents the local density of the learning data projected into the PCA space.

[0073] Therefore, the two likelihoods (Like1 and Like2) are related to the 261 Cs. 计算 T J Each of them is associated with something else.

[0074] 7) Then, for the 261 Cs 计算 T J Each of the following is a product of Like1 and Like2.

[0075] Finally, C will be the product of (7) and (7). 计算 T J The associated temperature is limited to the temperature of the steel strip.

[0076] The comparison between the temperature obtained by thermocouple measurement and the temperature estimated by the claimed protection is plotted on... Figure 5 In the diagram, dotted lines represent temperatures estimated using the previously described method, while continuous lines represent temperatures measured by thermocouples. It is clear that the temperatures given by this method are reliable.

Claims

1. A method for estimating the temperature T of steel products having temperatures ranging from 300°C to 1600°C. REAL The method, the method comprising: A. a calibration step comprising the following steps: i. the intensity I of the radiation emitted by a reference object having a known temperature T REF at 5 wavelengths λ ranging from 0.9 µm to 2.1 µm, with one wavelength from 0.9 µm to 1.35 µm, one wavelength from 1.35 µm to 1.55 µm, one wavelength from 1.55 µm to 1.85 µm, one wavelength from 1.85 µm to 2.05 µm, and one wavelength from 2.05 µm to 2.1 µm, the measurement conditions being characterized by the emissivity ε REF of the reference object and the transmissivity α REF of the medium between the reference object and the sensor, wherein the reference object is a steel product, ii. using the measured intensities I at the 5 wavelengths to calculate the spectral attenuation coefficients C CALIB , where P(λ, T) is the spectral density of electromagnetic radiation emitted by a black body in thermal equilibrium at wavelength λ and at temperature T REF REF at temperature T​ iii. for the reference emissivity ε REF and the transmissivity a of the medium between the reference and the sensor REF N different combinations of the N CALIB repeating the steps i. and ii. for N CALIB spectral attenuation coefficients, N CALIB is an integer greater than 2, B. a measurement step comprising the following steps: i. measuring the intensity I of the radiation emitted by the steel product at 5 wavelengths l of the range from 0.9 pm to 2.1 pm, ii. N T temperature Tj and for the 5 wavelengths T N 计算 Tj, N T is an integer from 2 to 1300, wherein, - P( l, Tj) is the spectral density of the electromagnetic radiation emitted by a black body in thermal equilibrium at wavelength l and at temperature Tj, based on Planck's law, C. a comparison step comprising the following steps: i. performing a probability test for the C CALIB most probable C 计算 Tj is found among the C Tj, ii. the temperature T of the steel product is estimated to be equal to the most probable C REAL estimated to be equal to the most probable C 计算 the temperature Tj of Tj.

2. The method of claim 1, wherein, the cooling treatment is performed during or after hot rolling and the steel product has a temperature from 300°C to 1100°C, and wherein in step B, Tj ranges from 300°C to 1100°C.

3. The method of claim 1, wherein, the cooling treatment is performed during or after continuous casting and the steel product has a temperature from 800°C to 1600°C, and wherein in step B, Tj ranges from 800°C to 1600°C.

4. The method according to any one of claims 1 to 3, in steps A i. and B i., the intensity of the radiation is measured at 8 wavelengths λ ranging from 0.9 pm to 2.1 pm, wherein, one wavelength from 0.9 pm to 1.11 pm, one wavelength from 1.11 pm to 1.15 pm, one wavelength from 1.15 pm to 1.35 pm, one wavelength from 1.35 pm to 1.55 pm, one wavelength from 1.55 pm to 1.85 pm, one wavelength from 1.85 pm to 2.05 pm, one wavelength from 2.05 pm to 2.07 pm, and one wavelength from 2.07 pm to 2.1 pm, and in steps A ii. and B ii., the spectral attenuation coefficients of the 8 wavelengths are calculated.

5. The method of claim 4, wherein, in steps A i. and B i., the intensity of the radiation is measured at 5 additional wavelengths of the range from 0.9 pm to 2.1 pm, and in stepwise steps A ii. and B ii., the spectral attenuation coefficients of the 8 wavelengths and the 5 additional wavelengths are calculated.

6. The method of claim 4, wherein, in steps A i. and B i., the intensity of the radiation is measured at 42 additional wavelengths of the range from 0.9 pm to 2.1 pm, and in stepwise steps A ii. and B ii., the spectral attenuation coefficients of the 8 wavelengths and the 42 additional wavelengths are calculated.

7. The method of claim 4, wherein, in steps A i. and B i., the intensity of the radiation is measured at 92 additional wavelengths of the range from 0.9 pm to 2.1 pm, and in stepwise steps A ii. and B ii., the spectral attenuation coefficients of the 8 wavelengths and the 92 additional wavelengths are calculated.

8. The method of any one of claims 1 to 3, wherein, N CALIB is an integer from 2 to 1000.

9. The method of claim 8, wherein, N CALIB is an integer from 20 to 1000.

10. The method of any one of claims 1 to 3, wherein, In step C i. the probability test comprises a test of the C CALIB Dimensionality reduction is performed.

11. The method of claim 10, wherein, in step C i., the dimensionality reduction is performed with principal component analysis.

12. The method of any one of claims 1 to 3, wherein, In step C i. the probability test comprises the C CALIB Projection in a probabilistic model.

13. The method of claim 12, wherein, in step C i., the probabilistic model is a Gaussian mixture model.

Images