Method for measuring focus heat effect and heat damage of Raman probe

Abstract

The invention discloses a method for measuring the focus heat effect and the heat damage of a Raman probe. The method comprises the following steps: S1, establishing a stable laser output system, and configuring a proper optical focusing system; s2, heating the standard sample to a target temperature according to a preset step length, keeping for several minutes at each target temperature, and performing spectral measurement on the standard sample by using a Raman spectrometer; s3, in the Raman spectrum measurement process, carrying out real-time temperature monitoring on the surface of the standard sample, and carrying out normalization processing on collected Raman spectrum data; s4, establishing a quantitative relation model between the relative intensity change characteristic # imgabs0 # of the Raman spectrum and the temperature T; and S5, predicting the temperature T of the standard sample under the current preset parameters according to the model, and adjusting the measurement parameters according to the prediction result, so that the predicted temperature T is lower than the temperature threshold value of the to-be-detected sample. According to the method, the thermal effect of the laser can be predicted before Raman measurement, proper measurement parameters are set in advance, and thermal damage to a measured sample is avoided.

Description

Technical Field

[0001] The present invention relates to the field of laser measurement technology, and in particular to a method for measuring thermal effects and thermal damage of a Raman probe focus. Background Art

[0002] To ensure that the Raman spectrometer obtains the highest possible quality Raman signals without causing ablation of the sample, it is necessary to determine the laser ablation threshold of the sample to be measured to guide the setting of the laser energy. Currently, the following methods can be used to achieve this: 1.1 Laser ablation threshold test method based on Raman spectroscopy Ling Zongcheng et al. proposed a method in patent CN116559145A published on August 8, 2023 as follows: The first step is to determine the ablation test origin on the surface of the sample to be tested and set the test grid based on the ablation reference origin.

[0003] The second step is to obtain the initial microscopic image and initial Raman signal of each test point in the test grid.

[0004] The third step is to select a test point, irradiate the test point with a pulsed laser beam, adjust and record the laser power and spot shape of the surface of the sample to be tested at the test point. During the adjustment process, take a microscopic image of the surface of the sample to be tested at the test point and obtain the Raman spectrum of the surface of the sample to be tested at the test point. Based on the microscopic image and Raman spectrum, judge the damage condition of the sample to be tested, and record the laser power, spot diameter and damage status of the sample to be tested when damage occurs on the surface of the sample to be tested.

[0005] In the fourth step, adjust the position of the sample to be tested, change the test point, and repeat step three until the measurement of all test points in the test grid is completed. Based on the laser power, spot diameter and damage status of the sample to be tested measured at each test point, the laser ablation threshold of the sample to be tested is obtained.

[0006] 1.2 Damage Threshold Prediction Method for Fused Silica Components Patent CN112683865A, published by Zhang Zengming et al. on April 20, 2021, proposes a method for predicting the damage threshold of fused silica components. The method first measures the fluorescence signal and Raman scattering signal of the fused silica component to be tested. Based on the fluorescence signal, the defect fluorescence peak is obtained. Based on the Raman scattering signal, the Raman scattering peak is obtained. The area of ​​the defect fluorescence peak and the area of ​​the Raman scattering peak are then normalized to obtain the ratio of the normalized fluorescence peak area to the Raman scattering peak area. The damage threshold of the fused silica component to be tested is predicted based on this ratio.

[0007] 1.3 Methods for online testing of laser damage to optical components Deng Wenyuan et al. proposed a method in patent CN106066318A published on November 2, 2016, which first generates a laser and irradiates the surface of an optical element. Then, fluorescence imaging is performed on the fluorescence signal excited by the surface of the optical element after the laser is irradiated to form a fluorescence image, and the fluorescence signal excited when the laser is irradiated on the surface of the optical element is collected. Then, the change in fluorescence intensity is obtained according to the fluorescence image, and the damaged area of ​​the optical element is judged according to the change in fluorescence intensity. The spectrum of the fluorescence signal is further analyzed, and the damage of the optical element is further judged according to the change in the spectrum.

[0008] The above methods mainly have the following problems: 2.1 In many cases, damage threshold prediction relies primarily on the ratio of the fluorescence signal to the Raman scattering signal of a specific material, which is suitable for certain materials (such as fused silica components). However, this method may not be applicable to other types of materials, especially those with weak fluorescence signals or no obvious fluorescence peaks. The acquisition and analysis of fluorescence signals are highly complex, and changes in the surface conditions of optical components (such as contamination or irregular surfaces) can affect the accuracy of the fluorescence signal. In addition, fluorescence imaging may fail under certain conditions (such as high laser power or special materials), resulting in inaccurate detection results.

[0009] 2.2 Current methods, due to performance differences between different devices, such as laser output stability and focusing system accuracy, are also heavily dependent on the physical structural parameters of the device setup, such as the distance from the laser to the test sample and the alignment deviation of the optical components. The laser threshold originally set on a particular device may change significantly after replacing or rebuilding the device. This change makes the original threshold data no longer applicable, requiring recalibration and recalibration, increasing the complexity and time cost of the experiment.

[0010] 2.3 The current method can only test the laser damage threshold of a certain material. It cannot intelligently change the laser power, spot diameter and other parameters before each experiment to predict the laser thermal effect to stimulate sufficient Raman signals without damaging the sample.

[0011] Therefore, it is necessary to provide a method for measuring the thermal effect and thermal damage of the Raman probe focus, which can predict the thermal effect of the laser before Raman measurement and set appropriate Raman measurement parameters in advance to avoid thermal damage to the sample being measured. Summary of the Invention

[0012] The technical problem to be solved by the present invention is to provide a method for measuring the thermal effect and thermal damage of the Raman probe focus, which can predict the thermal effect of the laser before Raman measurement, set appropriate Raman measurement parameters in advance, and avoid thermal damage to the measured sample.

[0013] In order to solve the above technical problems, the present invention provides a method for measuring the thermal effect and thermal damage of the focus of a Raman probe, comprising the following steps: S1: establishing a stable laser output system to ensure that the laser can provide continuously adjustable laser power output, and configuring an appropriate optical focusing system so that the laser can form a uniform light spot on the sample surface; S2: placing a standard sample on a constant temperature heating table, setting the heating temperature range as the room temperature as the starting temperature, heating to the target temperature each time according to a predetermined step length, and maintaining each target temperature for several minutes to ensure that the temperature of the standard sample is stable, and at each target temperature, using a Raman spectrometer to perform spectral measurement on the standard sample; S3: during the Raman spectral measurement process, real-time temperature monitoring is performed on the surface of the standard sample, and the real-time temperature T of the surface of the standard sample is recorded. After the measurement of all temperature points is completed, the collected Raman spectral data is normalized; S4: for the spectrum at each temperature T, point-by-point calculation is performed relative to the base spectrum The relative intensity change , so that the relative intensity changes Modeling and analysis are performed with the corresponding temperature data to establish the relative intensity change characteristics of the Raman spectrum A quantitative relationship model between the temperature T and the sample temperature; S5: predicting the temperature T of the standard sample under the current preset parameters according to the model, and adjusting the measurement parameters according to the prediction results so that the predicted temperature T is lower than the temperature threshold of the sample to be tested.

[0014] Furthermore, in step S2, the heating temperature range is up to 300° C., with a predetermined step length of 8-12° C., and each target temperature is maintained for 4-6 minutes.

[0015] Furthermore, the standard sample in step S2 is single crystal silicon, zinc oxide, graphene or gallium nitride.

[0016] Furthermore, in step S3, during the measurement process, the Raman spectrometer adopts a low-power laser mode and sets a longer integration time to reduce the effect of laser heating on the standard sample.

[0017] Furthermore, in step S3, a thermal imager is used to monitor the real-time temperature T of the surface of the standard sample.

[0018] Furthermore, in step S4: for each standard sample, a Raman spectrum is collected at room temperature as a base spectrum. , and use the following formula to calculate the relative spectrum of each temperature The relative intensity change of: .

[0019] Furthermore, the PLSR modeling analysis is used in step S4, including the following steps: S41: Calculate the covariance matrix of the independent variable X and the dependent variable Y, and extract the first set of principal components through an iterative algorithm. Covariance matrix calculation formula: Extract principal components: Assume that the extracted principal components are t and u, corresponding to the weight vectors w and c of X and Y respectively: S42: Use the extracted principal component t as a new independent variable to perform linear regression modeling on the dependent variable and obtain the regression coefficient b: Build a regression model for the dependent variable: S43: Continue to extract new principal components for the remaining independent variable residuals and perform regression until the predetermined stopping criteria are met, and then update the residuals: in, is the loading vector of the principal component in X: .

[0020] Furthermore, after the model is established in step S4, the temperature T of the current sample is predicted based on the model, and the actual temperature of the current sample is measured to verify the accuracy of the model result.

[0021] Furthermore, after the model is established in step S4, a set of data is recollected in the temperature range of 30°C to 300°C using the same equipment and acquisition parameters, and the temperature prediction values ​​and actual temperature measurement values ​​are compared to verify the accuracy of the model; and a set of data is recollected in the temperature range of 30°C to 300°C using different equipment and acquisition parameters, and the temperature prediction values ​​and actual temperature measurement values ​​are compared to verify the accuracy of the model.

[0022] Furthermore, in step S5, before testing the sample to be tested, the temperature tolerance value of the substance is first queried as the temperature threshold, and then the Raman spectrum of the standard sample at room temperature is tested as the base spectrum. , then calculate the relative intensity change The model predicts the temperature T of the standard sample under the current preset parameters. If the predicted temperature T is greater than or equal to the temperature threshold or the Raman spectrum does not reach the required intensity, the measurement parameters are adjusted until the Raman spectrum reaches the required intensity and the predicted temperature value does not exceed the temperature threshold.

[0023] Compared to existing technologies, the present invention offers the following advantages: The present invention provides a method for measuring focal thermal effects and thermal damage in Raman probes. This method uses a standard sample to predict Raman spectral thermal effects, pre-measures the impact of laser light on sample temperature, and establishes a quantitative relationship model between Raman spectral characteristics and temperature. First, by optimizing parameters such as laser power and integration time before the actual measurement, this method overcomes the problem of sample damage caused by excessively high temperatures due to laser thermal effects in traditional Raman measurements. Second, by establishing a temperature-spectrum relationship model for the standard sample, the present invention ultimately optimizes laser measurement parameters for other samples. This method is highly versatile and can be widely applied to different types of samples, proactively preventing sample damage caused by laser thermal effects. Third, the present invention proposes cross-device applicability through dynamic normalization of a single-device base spectrum. Each device only needs to collect a single base spectrum at room temperature (no full-temperature range data is required). By calculating the relative intensity between the test spectrum and the base spectrum, absolute intensity differences between devices are eliminated. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 Schematic diagram of the measurement process of the thermal effect and thermal damage of the Raman probe focal point in an embodiment of the present invention; Figure 2 This is the original spectrum of the standard sample single crystal silicon in the embodiment of the present invention; Figure 3 This is a spectrum diagram after normalization processing of Raman spectrum data collected from a standard sample of single crystal silicon in an embodiment of the present invention; Figure 4 This is a test set result analysis diagram of the model predicted temperature versus the actual temperature in an embodiment of the present invention; FIG5(a) is a thermal imaging image of the case where the laser causes thermal damage to the paper according to an embodiment of the present invention, and FIG5(b) is an original image of the case where the laser causes thermal damage to the paper according to an embodiment of the present invention; FIG6(a) is a thermal image of the rubber caused by thermal damage by laser in accordance with an embodiment of the present invention, and FIG6(b) is an original image of the rubber caused by thermal damage by laser in accordance with an embodiment of the present invention; Figure 7 This is a box plot of the prediction residuals of the same device in an embodiment of the present invention; Figure 8 This is a box plot of the prediction residuals of different devices in the embodiment of the present invention; Figure 9 This is a Raman spectrum of the paper measured after the parameters in step 2 of Table 5 in the embodiment of the present invention are adjusted; Figure 10 This is a Raman spectrum of the paper measured after the parameters in step 4 of Table 5 are adjusted in the embodiment of the present invention. DETAILED DESCRIPTION

[0025] The present invention will be further described below with reference to the accompanying drawings and examples.

[0026] The present invention provides a method for measuring the thermal effect of a Raman probe focus. This method aims to predict the thermal effect of the laser before Raman measurement, set appropriate Raman measurement parameters in advance, and avoid thermal damage to the sample being measured.

[0027] To achieve the above objectives, this embodiment provides a method for measuring the thermal effect of a Raman probe focal point. This solution first requires determining a standard sample for quantifying temperature. The core criteria for selecting the standard sample material are: the Raman intensity of the material is sensitive to temperature, and the Raman peak intensity of the material must change significantly with temperature and be linearly quantifiable; the material has high thermal conductivity and thermal diffusion capacity. High thermal conductivity ensures rapid heat diffusion, avoids local overheating and nonlinear response, and is conducive to establishing a universal thermal model; the material has structural stability and repeatability, and is a high-purity crystal or homogeneous material to reduce the interference of impurities / defects on the intensity signal; and has a clear optical phonon mode (such as Si's 520cm -1 Peak), for comparison and study. According to the above criteria, the following materials were selected in this embodiment: Table 1 See Figure 1 After selecting the material of the standard sample, the present embodiment provides a method for measuring the focal thermal effect of a Raman probe, comprising the following steps: S1: Establish a stable laser output system, ensure that the laser can provide continuously adjustable laser power output, and configure an appropriate optical focusing system so that the laser can form a uniform spot on the sample surface.

[0028] S2: Place the standard sample on a constant temperature heating table, set the heating temperature range as the room temperature as the starting temperature, raise the temperature to the target temperature each time according to the predetermined step length, and maintain it at each target temperature for several minutes to ensure that the temperature of the standard sample is stable. At each target temperature, use a Raman spectrometer to measure the spectrum of the standard sample; in a preferred embodiment, the heating temperature range is up to 300 ° C, that is, the heating range is from room temperature to 300 ° C. If the ignition point of the sample to be tested is higher than 300 ° C or other situations require spectral data above 300 ° C, the heating range can be adjusted as needed. The predetermined step length is preferably 8-12 ° C, more preferably 10 ° C, and each target temperature is preferably maintained for 4-6 minutes, more preferably 5 minutes. Building a model requires Raman spectra of standard samples at different temperatures. In this embodiment, single crystal silicon is used as the standard sample to collect data, that is, by using a constant temperature table to heat single crystal silicon.

[0029] S3: During the Raman spectrum measurement process, the surface of the standard sample is monitored for real-time temperature, and the real-time temperature T of the standard sample surface is recorded. After the measurement of all temperature points is completed, the collected Raman spectrum data is normalized. Preferably, a thermal imager is used to monitor the real-time temperature T of the surface of the standard sample. During the measurement process, the Raman spectrometer adopts a low-power laser mode and sets a longer integration time to reduce the impact of laser heating on the standard sample. Using a low-power, long integration time acquisition strategy, a total of 33 single-crystal silicon spectra at different temperatures are used, 3 frames are measured for each temperature, and a total of 99 data. Preferably, the laser power: ≤50mW; integration time: ≥10S, which can be adjusted according to the specific measured data quality. The original Raman spectra of single-crystal silicon at different temperatures are shown in the figure below. Figure 2 As shown. The collected Raman spectrum data is normalized to eliminate the influence of laser power fluctuation on the Raman spectrum, ensure the stability and comparability of the spectrum data, and increase the applicability of the model. The normalized spectrum is as follows Figure 3 shown.

[0030] S4: For each temperature T spectrum, calculate the relative value of the base spectrum point by point. The relative intensity change , so that the relative intensity changes Modeling and analysis are performed with the corresponding temperature data to establish the relative intensity change characteristics of the Raman spectrum The quantitative relationship model between the Raman spectrum and the temperature T is collected at room temperature for each standard sample as the base spectrum. , and use the following formula to calculate the relative spectrum of each temperature The relative intensity change of: In this step, First, for each single crystal silicon sample, a Raman spectrum is collected at room temperature (e.g., 25°C) as a base spectrum. , multiple measurements need to be averaged to reduce noise. And the following formula is used to calculate the relative value of each temperature to the base spectrum The relative intensity change of: The modeled data all use the relative intensities calculated in this step.

[0031] Then, use PLSR modeling: The core idea of ​​Partial Least Squares Regression (PLSR) is to find new orthogonal projection directions (principal components) that maximize the covariance between the projected dependent and independent variables, thereby establishing a predictive model. PLSR considers the correlation between the dependent and independent variables during the dimensionality reduction process, aiming to maximize predictive performance while reducing dimensionality.

[0032] The main steps of the algorithm are as follows: S41: Extracting principal components First, calculate the covariance matrix of the independent variable X and the dependent variable Y, and then extract the first set of principal components through an iterative algorithm (such as the NIIPALS algorithm). Covariance matrix calculation formula: Extract principal components: Assume that the extracted principal components are t and u, corresponding to the weight vectors w and c of X and Y respectively: S42: Regression Modeling The extracted principal component t is used as a new independent variable to perform linear regression modeling on the dependent variable and obtain the regression coefficient b: Build a regression model for the dependent variable: S43: Repeated Iteration Continue to extract new principal components for the remaining independent variable residuals and perform regression until the predetermined stopping criteria are met, and then update the residuals: in, is the loading vector of the principal component in X: .

[0033] like Figure 4 As shown, the final model has high linearity for the test set results of actual temperature and low error .

[0034] Furthermore, the temperature T of the current standard sample is predicted according to the model, and the actual temperature of the current standard sample is measured to verify the accuracy of the model result.

[0035] Verification Analysis 1: Same Device Using the same equipment and acquisition parameters, a new set of data was collected in the temperature range of 30°C to 300°C to verify the accuracy of the model. .

[0036] Table 2 Practical application model prediction Figure 5(a) shows a thermal image of the paper after laser thermal damage, and Figure 5(b) shows the original image of the paper after laser thermal damage. Table 2 shows that the model predicted the temperature when thermal damage occurred: 134.06°C; Figure 5(a) shows that the measured temperature was 134.3°C. The predicted relative standard deviation (RSD) was only 0.12%.

[0037] Figure 6(a) shows a thermal image of the rubber after laser thermal damage, and Figure 6(b) shows the original image of the rubber after laser thermal damage. Table 2 shows that the model predicts the temperature of the rubber after thermal damage: 296.9°C; Figure 6(a) shows that the measured temperature is 297.1°C. The predicted relative standard deviation (RSD) is only 0.05%.

[0038] Figure 7 The following is a box plot of the residual error in the model's temperature prediction for single-crystal silicon spectra collected using the same equipment. The following information is shown: 1. Overall deviation trend The median of the lowest temperature range (20-70℃) is closest to 0 because the data contains basis functions and has a large deviation; the interquartile range (IQR) of the high temperature range is short and relatively stable.

[0039] In general, the deviation is within 4°C.

[0040] 2. Error dispersion 20–70°C: The box height is the largest and the IQR is large, indicating that the low-temperature model is relatively unstable, with the error fluctuating between approximately -2.3°C and -3°C.

[0041] 70–120°C: The cabinet height and IQR are small, and the error fluctuates between approximately -1.5°C and 3.1°C.

[0042] 120–170°C: The box height and IQR are small, and the residual is approximately -1.3–3.8°C.

[0043] 170–230°C: The IQR becomes wider again, but the error is within ±3.5°C.

[0044] 230–290°C: Minimum box height, IQR, and fluctuation.

[0045] Verification Analysis 2: Different Devices Using different equipment and acquisition parameters, a new set of data was collected in the temperature range of 30℃ to 300℃ to verify the accuracy of the model. .

[0046] Table 3 Prediction of different device application models Figure 8The following is a box plot of the residual error in the model's temperature prediction for single-crystal silicon spectra collected by different devices. The following information can be seen: 1. Overall deviation trend The median of the lowest temperature range (20–70°C) is closest to 0 because the data contain basis functions and have a large deviation.

[0047] Overall, the deviation is within 4.3℃.

[0048] 2. Error dispersion 20–70°C: The box height and IQR are large, indicating that the low-temperature model is relatively unstable, with the error fluctuating between approximately -1.9°C and 3°C.

[0049] 70–120°C: The box height and IQR are small, and most of the errors fluctuate between approximately 0°C and 3°C.

[0050] 120–170°C: The box height and IQR are the largest, indicating that the model in this temperature range is the most unstable, with residuals of approximately -2.3–4°C.

[0051] 170–230°C: Error within ±3.1°C.

[0052] 230–290°C: Minimum box height, IQR, and fluctuation.

[0053] Comprehensive performance summary table: The following table shows the key indicators of the prediction results of the two devices.

[0054] Table 4 Prediction of different device application models As shown in Table 4, all three sets of data show excellent fitting effects, the coverage of the explanatory variables on the target variables is close to perfect, and the absolute value of the error is also low.

[0055] Between new data and modeled data for the same device and The difference is minimal, and the model's generalization ability on new data from the original device is stable.

[0056] The adaptability of the model to new equipment data is slightly reduced, but the fitting effect is still excellent and the error is within an acceptable range.

[0057] S5: Predict the temperature T of the standard sample under the current preset parameters based on the model, and determine whether the measurement parameters need to be adjusted based on the prediction results so that the predicted temperature T is lower than the temperature threshold of the sample to be tested. Specifically, before testing the sample to be tested, the temperature tolerance value of the substance is first queried as the temperature threshold, and then the Raman spectrum of the standard sample at room temperature is tested as the base spectrum. , then calculate the relative intensity change The established model is used to predict the temperature T of the standard sample. If the predicted temperature T is greater than or equal to the temperature threshold or the Raman spectrum does not reach the required intensity, the measurement parameters (such as laser power, integration time, objective lens, focal length, etc.) are adjusted until the Raman spectrum reaches the required intensity and the predicted temperature value is lower than the temperature threshold. This process may involve multiple dynamic adjustments.

[0058] The following uses writing paper as an example to illustrate the practical application of the model.

[0059] Basic ignition point range: 130-180℃ Model application and parameter adjustment process: Preset parameters: laser power = 300mW (785nm), integration time = 3S Model predicted temperature: 226.6°C (far exceeding the lower limit of the ignition point range) Dynamically adjust parameters: Table 5 Parameter adjustment process Figure 9 This is the Raman spectrum of the paper measured after the parameters are adjusted in step 2. The signal-to-noise ratio is low. Figure 10 The image shows the Raman spectrum after adjusting the parameters within the safe temperature range, which provides high-quality data while ensuring that the paper is not damaged.

[0060] In summary, the measurement method provided by the present invention has the following technical effects: (1) Effectively prevent the measured sample from being damaged by laser heating: By using standard samples such as single crystal silicon to establish a relationship model between laser power and temperature before the formal measurement, the thermal effect of the laser can be predicted in advance, and parameters such as laser power, integration time, and the distance between the sample and the laser can be optimized to ensure that the heat generated by laser heating does not exceed the temperature threshold of the measured sample during the actual measurement process, thereby effectively preventing the measured sample from being damaged by the laser thermal effect.

[0061] (2) Breakthrough of material applicability limitations: Replace the fluorescence dependence of the target material with standardized signals such as single crystal silicon, and support pre-test temperature detection of weakly fluorescent / non-fluorescent materials.

[0062] (3) Plug and play devices, zero-cost adaptation: The same model supports any device, and new devices can be used by simply uploading the base spectrum.

[0063] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make some modifications and improvements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the definition of the claims.

Claims

1. A method for measuring the thermal effect and thermal damage of a Raman probe focus, characterized in that: The method includes the following steps: S1: Establish a stable laser output system to ensure that the laser can provide continuously adjustable laser power output, and configure an appropriate optical focusing system so that the laser can form a uniform spot on the sample surface; S2: Place the standard sample on a constant temperature heating table, set the heating temperature range with room temperature as the starting temperature, and raise the temperature to the target temperature in predetermined steps each time. Maintain each target temperature for several minutes to ensure the temperature of the standard sample is stable. At each target temperature, perform spectral measurement on the standard sample using a Raman spectrometer. S3: During the Raman spectrum measurement process, the surface temperature of the standard sample is monitored in real time, and the real-time temperature T of the surface of the standard sample is recorded. After the measurement of all temperature points is completed, the collected Raman spectrum data is normalized; S4: For each temperature T spectrum, calculate the relative value of the base spectrum point by point. The relative intensity change , so that the relative intensity changes Modeling and analysis are performed with the corresponding temperature data to establish the relative intensity change characteristics of the Raman spectrum Quantitative relationship model between temperature T; S5: Predicting the temperature T of the standard sample under the current preset parameters according to the model, and adjusting the measurement parameters according to the prediction result so that the predicted temperature T is lower than the temperature threshold of the sample to be tested.

2. The measuring method according to claim 1, wherein In step S2, the heating temperature range is up to 300° C., with a predetermined step length of 8-12° C., and each target temperature is maintained for 4-6 minutes.

3. The measuring method according to claim 1, wherein The standard sample in step S2 is single crystal silicon, zinc oxide, graphene or gallium nitride.

4. The measuring method according to claim 1, wherein In step S3, during the measurement process, the Raman spectrometer adopts a low-power laser mode and sets a longer integration time to reduce the effect of laser heating on the standard sample.

5. The measuring method according to claim 1, wherein: In step S3, a thermal imager is used to monitor the real-time temperature T of the surface of the standard sample.

6. The measuring method according to claim 1, wherein: In step S4: for each standard sample, a Raman spectrum is collected at room temperature as a base spectrum. , and use the following formula to calculate the relative spectrum of each temperature The relative intensity change of: .

7. The measuring method according to claim 1, wherein: The PLSR modeling analysis is used in step S4, including the following steps: S41: Calculate the covariance matrix of the independent variable X and the dependent variable Y, and extract the first set of principal components through an iterative algorithm. Covariance matrix calculation formula: Extract principal components: Assume that the extracted principal components are t and u, corresponding to the weight vectors w and c of X and Y respectively: S42: Use the extracted principal component t as a new independent variable to perform linear regression modeling on the dependent variable and obtain the regression coefficient b: Build a regression model for the dependent variable: S43: Continue to extract new principal components for the remaining independent variable residuals and perform regression until the predetermined stopping criteria are met, and then update the residuals: in, is the loading vector of the principal component in X: 。 8. The measuring method according to claim 1, wherein: After the model is established in step S4, the temperature T of the current standard sample is predicted based on the model, and the actual temperature of the current standard sample is measured to verify the accuracy of the model result.

9. The measuring method according to claim 8, wherein: After the model is established in step S4, a set of data is recollected in the temperature range of 30°C to 300°C using the same equipment and acquisition parameters, and the temperature prediction values ​​and actual temperature measurement values ​​are compared to verify the accuracy of the model; and a set of data is recollected in the temperature range of 30°C to 300°C using different equipment and acquisition parameters, and the temperature prediction values ​​and actual temperature measurement values ​​are compared to verify the accuracy of the model.

10. The measuring method according to claim 1, wherein: In step S5, before testing the sample to be tested, the temperature tolerance value of the substance is first queried as the temperature threshold, and then the Raman spectrum of the standard sample at room temperature is tested as the base spectrum. , then calculate the relative intensity change The model predicts the temperature T of the standard sample under the current preset parameters. If the predicted temperature T is greater than or equal to the temperature threshold or the Raman spectrum does not reach the required intensity, the measurement parameters are adjusted until the Raman spectrum reaches the required intensity and the predicted temperature value does not exceed the temperature threshold.

Images