Method for predicting enthalpy value of phase change material

Abstract

The invention relates to a method for predicting an enthalpy value of a phase change material, and the method comprises the steps: carrying out the partial least square regression analysis of a near infrared spectrum of a phase change material sample and a standard enthalpy value, and building a first correction model; then selecting a spectrum fitting residual matrix according to the construction process of the first correction model, and taking the spectrum fitting residual matrix and the enthalpy value matrix as an input signal and a teacher signal of an extreme learning machine respectively to obtain a second correction model; through the combined application of the first correction model and the second correction model, the linear relation and the nonlinear relation between the near infrared spectrum and the enthalpy value of the phase change material sample can be comprehensively considered, the enthalpy value prediction capacity in a wide range is achieved, and the model prediction precision is improved.

Description

Technical Field

[0001] The present disclosure relates to a method for predicting the enthalpy value of a phase change material. Background Art

[0002] Phase change material is a material with energy storage and temperature control properties, and is widely used in aerospace, construction, textiles and other fields. Among them, phase change wax is one of the ideal phase change energy storage materials, usually referring to special wax with a phase change temperature of 20-80°C. It is an excellent organic phase change energy storage material with good heat storage performance, wide raw material sources, low prices, and is non-toxic and non-corrosive. It can be used in many industries such as building materials, solar energy, industrial waste heat utilization, temperature control textiles, electronics, modern agriculture, and health care. The main components of phase change wax are normal alkanes (C n H 2n+2 ), its production raw materials mainly come from two categories, one is petroleum wax, which usually needs to be solvent dewaxed to increase its normal hydrocarbon content in order to avoid affecting the product enthalpy value; the other commonly used raw material is Fischer-Tropsch synthetic wax, which is characterized by a relatively simple composition, and a normal straight-chain alkane content of more than 90%, basically free of cyclic hydrocarbons and aromatic hydrocarbons, and is a high-quality raw material for the production of phase change wax.

[0003] Near infrared spectroscopy has the advantages of rapid and non-destructive analysis. The use of near infrared spectroscopy to determine the physical properties and composition of chemical products is the most extensive and in-depth research topic in the petrochemical field. The near infrared spectral absorbance and enthalpy value of phase change materials can be used to establish a more practical analysis model through the partial least squares method, but this model often improves the robustness of the model at the expense of the model's prediction accuracy. Summary of the invention

[0004] The purpose of the present disclosure is to provide a method for predicting the enthalpy value of a phase change material, so as to achieve fast and accurate prediction over a wide enthalpy value range with strong robustness and high precision.

[0005] In order to achieve the above object, the present disclosure provides a method for predicting the enthalpy value of a phase change material, the method comprising the following steps:

[0006] Acquire near infrared spectra of a plurality of phase change material samples and a standard enthalpy value of each of the phase change material samples;

[0007] Obtaining the absorbance of each characteristic spectrum region in the near-infrared spectrum to form a first characteristic spectrum matrix; and forming an enthalpy value matrix with the standard enthalpy values ​​of all the phase change wax samples;

[0008] Performing partial least squares regression analysis on the first characteristic spectrum matrix and the enthalpy value matrix to establish a first correction model;

[0009] Selecting a first spectrum fitting residual matrix corresponding to the f-th principal factor according to the first correction model, using the first spectrum fitting residual matrix as an input signal of an extreme learning machine, using the enthalpy matrix as a teacher signal of the extreme learning machine, and establishing a second correction model;

[0010] Collecting a sample spectrum of the phase change material sample to be tested; obtaining the absorbance of a characteristic spectrum region of the sample spectrum to be tested; determining a first enthalpy value prediction value according to the absorbance of the characteristic spectrum region of the sample spectrum to be tested and the first calibration model;

[0011] Selecting a second spectrum fitting residual matrix corresponding to the f-th principal factor of the phase change material sample to be tested according to the first calibration model, and determining a second enthalpy value prediction value according to the second spectrum fitting residual matrix and the second calibration model;

[0012] Determine respective prediction weights of the first enthalpy prediction value and the second enthalpy prediction value, and weight the first enthalpy prediction value and the second enthalpy prediction value according to their respective prediction weights to obtain a predicted enthalpy value of the phase change material sample to be tested.

[0013] Optionally, the characteristic spectrum range is 5780-7700cm -1 .

[0014] Optionally, the method further comprises:

[0015] The spectrum fitting residual matrix Ex is determined according to the following formula (1), formula (2) or formula (3):

[0016] Ex=XX(f best ) (1);

[0017] Where X represents the spectral matrix, X(f best ) indicates that the first correction model takes the best principal factor f best The spectral matrix at time ,

[0018] Ex=XX(f best+1～n ) (2);

[0019] Where X(f best+1～n ) indicates that the first correction model takes the best principal factor f best The spectrum matrix of any one of the 1 to n principal factors after , where n is an integer from 2 to 15;

[0020] Ex = X (3);

[0021] Optionally, the number of the spectral fitting residual matrices Ex is one or more.

[0022] Optionally, the method further comprises:

[0023] The regression coefficients of the first correction model and the second correction model are respectively obtained, and the regression coefficients are used as prediction weights corresponding to the first prediction value and the second prediction value.

[0024] Optionally, the method further comprises:

[0025] Multiple predictions are performed based on the second spectral fitting residual matrix and the second correction model, and an average value of the multiple prediction results is taken as the second enthalpy value prediction value.

[0026] Optionally, the method comprises:

[0027] Performing a second-order differential process on each of the near-infrared spectra to obtain a first differential spectrum;

[0028] The first differential absorbance of the characteristic spectrum region in each of the first differential spectra is obtained to form the first characteristic spectrum matrix.

[0029] Optionally, the method comprises:

[0030] Performing second-order differential processing on the spectrum of the sample to be tested to obtain a second differential spectrum;

[0031] Obtaining absorbance of a characteristic spectral region of the second differential spectrum;

[0032] The first enthalpy value prediction value is determined according to the absorbance of the characteristic spectral region of the second differential spectrum and the first calibration model.

[0033] Optionally, the window width of the second-order differential processing is 13 to 25.

[0034] Optionally, the phase change material is phase change wax.

[0035] Optionally, the standard enthalpy value is measured by differential scanning calorimetry.

[0036] Through the above technical scheme, the present invention performs partial least squares regression analysis on the near-infrared spectrum of the phase change material sample and the standard enthalpy value to establish a first correction model; then, according to the construction process of the first correction model, the spectral fitting residual matrix is ​​selected, and the spectral fitting residual matrix and the enthalpy value matrix are used as the input signal and teacher signal of the extreme learning machine respectively to obtain the second correction model. Through the combined application of the first correction model and the second correction model, the linear relationship and nonlinear relationship between the near-infrared spectrum of the phase change material sample and the enthalpy value can be comprehensively considered, and the enthalpy value prediction capability in a wide range is achieved, thereby improving the model prediction accuracy.

[0037] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present disclosure but do not constitute a limitation of the present disclosure. In the accompanying drawings:

[0039] Figure 1 It is a flow chart of a method for predicting the enthalpy value of a phase change material provided in one embodiment of the present disclosure. DETAILED DESCRIPTION

[0040] The specific implementation of the present disclosure is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the present disclosure, and is not used to limit the present disclosure.

[0041] The present disclosure provides a method for predicting the enthalpy value of a phase change material, such as Figure 1 As shown, the following steps S101 to S107 are included:

[0042] S101, obtaining near-infrared spectra of a plurality of phase change material samples and a standard enthalpy value of each of the phase change material samples;

[0043] S102, acquiring the absorbance of each characteristic spectrum region in the near-infrared spectrum to form a first characteristic spectrum matrix; and forming an enthalpy value matrix with the standard enthalpy values ​​of all the phase change wax samples;

[0044] S103, performing partial least squares regression analysis on the first characteristic spectrum matrix and the enthalpy value matrix to establish a first calibration model;

[0045] S104, selecting a first spectrum fitting residual matrix corresponding to the f-th principal factor according to the first correction model, using the first spectrum fitting residual matrix as an input signal of an extreme learning machine, using the enthalpy matrix as a teacher signal of the extreme learning machine, and establishing a second correction model;

[0046] S105, collecting a sample spectrum of the phase change material sample to be tested; obtaining the absorbance of a characteristic spectrum region of the sample spectrum to be tested; and determining a first enthalpy value prediction value according to the absorbance of the characteristic spectrum region of the sample spectrum to be tested and the first calibration model;

[0047] S106, selecting a second spectrum fitting residual matrix corresponding to the f-th principal factor of the phase change material sample to be tested according to the first calibration model, and determining a second enthalpy value prediction value according to the second spectrum fitting residual matrix and the second calibration model;

[0048] S107: Determine respective prediction weights of the first enthalpy prediction value and the second enthalpy prediction value, and weight the first enthalpy prediction value and the second enthalpy prediction value according to their respective prediction weights to obtain a predicted enthalpy value of the phase change material sample to be tested.

[0049] The present invention divides the correspondence between the near-infrared spectral absorbance and enthalpy value of the phase change material into a linear relationship and a nonlinear relationship, and establishes a correction model step by step through a combination of the two. The linear relationship is fitted and corrected using the partial least squares method, and the nonlinear relationship is fitted and corrected using the extreme learning machine, which has higher prediction accuracy.

[0050] In the present disclosure, the partial least squares regression analysis (PLS) can be performed using conventional selection methods and software in the art. Specifically, the present disclosure uses the interaction test method to determine the optimal number of principal factors f best .

[0051] In the present disclosure, an extreme learning machine (ELM) is a type of machine learning system or method built based on a feedforward neuron network (FNN).

[0052] The present invention discloses an extreme learning machine fitting correction for the nonlinear relationship between the absorbance of a phase change material sample in a characteristic spectral region and the standard enthalpy value, wherein the spectral fitting residual of partial least squares regression is used as the input signal of the extreme learning machine (ELM), and the standard enthalpy value is used as the teacher signal (i.e., the expected output value, also called the supervision signal). The extreme learning machine is subjected to supervised learning (also called teacher learning) to adjust the parameters of the ELM so that the ELM can meet the performance requirements for predicting the nonlinear relationship between the near-infrared spectrum and the enthalpy value of the phase change material sample.

[0053] In one embodiment, the characteristic spectrum range is 5780-7700 cm -1 In this embodiment, the absorbance and standard enthalpy values ​​within the appropriate characteristic spectrum range are selected to construct the model, which can improve the correlation between the absorbance and the standard enthalpy value and simplify the amount of calculation in the model construction process.

[0054] In one embodiment, the method further comprises:

[0055] The spectrum fitting residual matrix Ex is determined according to the following formula (1), formula (2) or formula (3):

[0056] Ex=XX(f best ) (1);

[0057] Where X represents the spectral matrix, X(f best ) indicates that the first correction model takes the best principal factor f best The spectral matrix at time ,

[0058] Ex=XX(f best+1～n ) (2);

[0059] Where X(f best+1～n ) indicates that the first correction model takes the best principal factor f best The spectrum matrix of any one of the 1 to n principal factors after , where n is an integer from 2 to 15;

[0060] Ex=X (3).

[0061] In the present disclosure, the above formula (1), formula (2) or formula (3) are respectively used to obtain three spectral fitting residual matrices, and the three spectral fitting residual matrices obtained are all used to construct a second correction model to obtain three second correction models. According to the verification analysis results of the three second correction models, such as the degree of dispersion, mean square error, etc., the formula corresponding to the second correction model with better prediction effect is selected.

[0062] In a specific implementation, the number of the spectrum fitting residual matrix Ex is one or more. Specifically, one or more principal factors f can be selected, and then one or more spectrum fitting residual matrices Ex are obtained accordingly. For example, f best When is 3, f can be one or more of 1 to 3.

[0063] In a specific implementation, the present disclosure uses an interactive test method to determine the optimal number of principal factors f best .

[0064] In a specific implementation manner, the present disclosure selects the fth principal factor by the following method:

[0065] The spectrum fitting residual matrix E corresponding to each main factor x The and enthalpy value matrices are used as input signals and teacher signals of the extreme learning machine respectively, and the extreme learning machine is corrected to obtain a second correction model corresponding to each principal factor;

[0066] According to the first correction model and the second correction model, the predicted value of the enthalpy value of the phase change material corresponding to each main factor is obtained, and the predicted standard deviation value between the predicted value of the enthalpy value of the phase change material and the true value of the enthalpy value is calculated; the main factor with the smallest predicted standard deviation value or the main factor with a stable predicted standard deviation value is selected as f in the above formulas (1) to (3).

[0067] Specifically, the grid search method can be used to select f in the above formulas (1) to (3). Specifically, the grid search method is used to list all possible main factors f, and then verify them one by one, and select the main factor f with the smallest prediction standard deviation value or the main factor f with a stable prediction standard deviation value.

[0068] In the present disclosure, the appropriate main factor is obtained by back-calculating the predicted standard deviation value as f in the above formulas (1) to (3), and the spectral fitting residual matrix E of multiple main factors can be fully compared. x The model construction results of the enthalpy matrix are used to improve the enthalpy prediction accuracy.

[0069] In one embodiment, the method further comprises:

[0070] The regression coefficients of the first correction model and the second correction model are respectively obtained, and the regression coefficients are used as prediction weights corresponding to the first prediction value and the second prediction value.

[0071] In a specific embodiment, the enthalpy value y of the sample to be tested is un The prediction process includes:

[0072] (1) Determine the first enthalpy prediction value

[0073] x un is the absorbance in the characteristic spectral region of the sample to be tested. The absorbance in the characteristic spectral region of the sample to be tested is input into the established first calibration model to obtain the first enthalpy value prediction value y1.

[0074] (2) Determine the second enthalpy prediction value

[0075] According to the above formula (1), (2) or (3) determined in the process of establishing the second model, the spectral fitting residual matrix corresponding to the f-th principal factor corresponding to the sample to be tested is obtained; the spectral fitting residual matrix of the sample to be tested is input into the second correction model to obtain the second enthalpy value prediction value y2.

[0076] (3) Determine the prediction weights of the first enthalpy value prediction value and the second enthalpy value prediction value

[0077] Perform multiple regression analysis on the first enthalpy value prediction value y1, the second enthalpy value prediction value y1 and the enthalpy value matrix, that is, substitute y1 and y1 into Y=b1×y1+b2×y2+e to solve and obtain regression coefficients b1 and b2, where e is the residual.

[0078] (4) Solve the final enthalpy prediction result

[0079] The first enthalpy prediction value and the second enthalpy prediction value are weighted according to their respective prediction weights to obtain the final enthalpy prediction value y un , that is, y un=b1×y1+b2×y2+e.

[0080] In one embodiment, the method further includes: performing multiple predictions based on the second spectrum fitting residual matrix and the second correction model, and taking an average value of the multiple prediction results as the second enthalpy value prediction value, thereby avoiding errors and improving the accuracy of the prediction results.

[0081] In one embodiment, the method further comprises:

[0082] Performing a second-order differential process on each of the near-infrared spectra to obtain a first differential spectrum;

[0083] Acquire the first differential absorbance of the characteristic spectrum region in each of the differential spectra to form the first characteristic spectrum matrix;

[0084] The first characteristic spectrum matrix and the enthalpy value matrix are subjected to the partial least squares regression analysis to obtain the first correction model.

[0085] In one embodiment, the method further comprises:

[0086] Performing second-order differential processing on the spectrum of the sample to be tested to obtain a second differential spectrum;

[0087] Obtaining absorbance of a characteristic spectral region of the second differential spectrum;

[0088] The first enthalpy value prediction value is determined according to the absorbance of the characteristic spectral region of the second differential spectrum and the first calibration model.

[0089] The present disclosure can eliminate the interference of the baseline offset, drift and background of the spectrum by performing second-order differential processing on the near-infrared spectrum and / or the spectrum of the sample to be tested.

[0090] In a specific implementation manner, the window width of the second-order differential processing is 13 to 25. In the present disclosure, the second-order differential processing method can be derived using Savitzky-Golay (SG).

[0091] In one embodiment, the phase change material sample can be a phase change wax. The disclosed method is particularly suitable for predicting the enthalpy value of a phase change wax sample. There is a good linear relationship between the near infrared spectrum absorbance and the enthalpy value of the phase change wax sample, which is conducive to achieving rapid and accurate prediction of phase change wax samples with a wide enthalpy value range.

[0092] In a specific implementation manner, the phase change material samples may be phase change material samples of different types, and the number of samples is at least 100, so as to further improve the accuracy and reliability of the established model.

[0093] In a specific embodiment, the near infrared spectroscopy detection conditions include: the scanning range is 4000-10000cm -1 In the present disclosure, the resolution can be selected according to the model construction accuracy or instrument accuracy, for example, the resolution is 4 cm -1 、8cm -1 The sampling point interval of the near infrared spectrum absorbance is 2 to 16 wave numbers.

[0094] In the present disclosure, the standard enthalpy value is tested by a method conventionally selected in the art, such as differential scanning calorimetry (DSC).

[0095] In the present disclosure, the phase change material samples are divided into calibration set samples and test set samples, wherein the calibration set samples are used for model construction, and the test set samples are used to verify the predicted values ​​and actual values ​​of the constructed model. In the present disclosure, a large number of calibration set samples can be used, and they have good representativeness, that is, the enthalpy values ​​of the calibration set samples can cover the enthalpy values ​​of all predicted phase change wax samples, so that the established model has good generalization ability. Preferably, the number of calibration set samples should account for 2 / 3 to 3 / 4 of the total number of samples, and the remaining 1 / 3 to 1 / 4 are used as test set samples. Preferably, the distribution patterns of the calibration set and test set samples are approximately the same.

[0096] The present disclosure is further described in detail below through examples, but the present disclosure is not limited thereto.

[0097] In the following embodiments and comparative examples, the test conditions for near infrared spectroscopy tests on calibration set samples and test set samples are the same.

[0098] The instrument for near infrared spectroscopy was a Fourier transform near infrared spectrometer produced by Thermo, model Antaris II, and the spectral acquisition conditions were: 0.5 mm cuvette, resolution 8 cm -1 , the acquisition wave number range is 4000~10000cm -1 , cumulative scanning times 64 times, transmission measurement method.

[0099] Example 1

[0100] (1) Determination of phase change wax enthalpy by DSC method

[0101] 100 phase change wax samples from a certain factory were collected and their enthalpy values ​​were determined by DSC method as standard enthalpy values. 75 representative phase change wax samples were selected to form a calibration set, and the remaining samples were used as a test set.

[0102] (2) Constructing the first calibration model and the second calibration model

[0103] All calibration set samples are tested by near infrared spectroscopy to obtain near infrared spectra of all calibration set samples. The near infrared spectrum of each calibration set sample is subjected to second-order differential processing with a window width of 25 to obtain a first differential spectrum.

[0104] For each first differential spectrum, select 5780-7700cm -1 The absorbance in the spectral region is 5780-7700cm -1 The absorbance in the spectral region constitutes the absorbance matrix X, and the standard enthalpy values ​​of all calibration set samples measured in step (1) constitute the enthalpy matrix Y.

[0105] The absorbance matrix X and the enthalpy matrix Y were subjected to partial least squares regression analysis (PLS regression analysis) to obtain the first calibration model of linear fitting. All the main factors were determined in the PLS regression analysis process, and the optimal number of main factors f was determined by the interaction test method. best In this embodiment, the number of main factors selected is 4, f best is the fourth principal factor.

[0106] According to formula (1), the spectrum fitting residual matrix (Ex = XX (4) ), and then use the spectrum fitting residual matrix and enthalpy value matrix as the input signal and teacher signal of the extreme learning machine respectively, and use the ELM algorithm to perform nonlinear correction on the spectrum fitting residual matrix and enthalpy value matrix to obtain a second correction model of nonlinear fitting.

[0107] (3) Use the test set samples to verify the prediction results of the first calibration model and the second calibration model

[0108] A near-infrared spectrum test was performed on each sample in the test set to obtain the spectrum of the sample to be tested. After the second-order differential processing with a window width of 25, the second differential spectrum was obtained; the wave number of the second differential spectrum was intercepted to be 5780-7700cm -1 The absorbance of the characteristic spectral region is used as the absorbance to be measured; the absorbance of the characteristic spectral region of the test set sample is input into the first calibration model established to obtain the first enthalpy value prediction value y1; the spectrum fitting residual matrix corresponding to the fourth principal factor of the test set sample is selected according to the first calibration model, and the spectrum fitting residual matrix of the sample to be tested is input into the second calibration model to obtain the second enthalpy value prediction value y2, and the first enthalpy value prediction value y1, the second enthalpy value prediction value y1 and the enthalpy value matrix are subjected to multiple regression analysis, that is, y1 and y1 are substituted into Y = b1 × y1 + b2 × y2 + e to solve, and the regression coefficients b1 and b2 are obtained, where e is the residual. It is calculated that b1 = 0.3481 and b2 = 0.6554.

[0109] The first enthalpy prediction value and the second enthalpy prediction value are weighted according to their respective prediction weights to obtain the final enthalpy prediction value y un , that is, y un =b1×y1+b2×y2+e. The comparison results of the predicted enthalpy values ​​of the test set samples and the standard enthalpy values ​​determined by the DSC method are shown in Table 1.

[0110] The calculation formula of relevant statistical parameters is as follows:

[0111]

[0112] in, is the predicted value of the i-th sample, y i (i=1,2,…,n) is the true value of the i-th sample, and n is the number of samples in the test set.

[0113] Among them, deviation = measured value - predicted value.

[0114] Table 1

[0115]

[0116] Comparative Example 1

[0117] The difference from Example 1 is that only the first step of partial least squares regression analysis (PLS) is performed to construct the first calibration model, and the step of constructing the second calibration model is not performed, and the optimal principal factor f is best The model was used for prediction under the condition of =4. The comparison results of the predicted enthalpy values ​​of the test set samples with the standard enthalpy values ​​determined by the DSC method are shown in Table 2.

[0118] Table 2

[0119]

[0120] Comparative Example 2

[0121] The difference from Example 1 is that the correction model is established by using the extreme learning machine (ELM) algorithm alone, and the ELM algorithm used in this embodiment is a pre-trained ELM algorithm, and the model is used for prediction under the optimal parameter conditions. The comparison results of the enthalpy value prediction values ​​of the test set samples with the standard enthalpy values ​​determined by the DSC method are shown in Table 3.

[0122] Table 3

[0123]

[0124]

[0125] According to the relevant data of the above embodiments and comparative examples, it can be seen that the SEP of the phase change material enthalpy value prediction model obtained by the method provided by the present invention is smaller, indicating that compared with the single linear partial least squares phase change material enthalpy value prediction method (Comparative Example 1) and the single nonlinear extreme learning machine phase change material enthalpy value prediction method (Comparative Example 2), the method provided by the present invention has better prediction accuracy for the phase change material enthalpy value.

[0126] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings; however, the present disclosure is not limited to the specific details in the above embodiments. Within the technical concept of the present disclosure, a variety of simple modifications can be made to the technical solution of the present disclosure, and these simple modifications all fall within the protection scope of the present disclosure.

[0127] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.

[0128] In addition, various embodiments of the present disclosure may be arbitrarily combined, and as long as they do not violate the concept of the present disclosure, they should also be regarded as the contents disclosed by the present disclosure.

Claims

1. A method for predicting the enthalpy value of a phase change material, characterized in that: The method comprises the following steps: Acquire near infrared spectra of a plurality of phase change material samples and a standard enthalpy value of each of the phase change material samples; Obtaining the absorbance of each characteristic spectrum region in the near-infrared spectrum to form a first characteristic spectrum matrix; and forming an enthalpy value matrix with the standard enthalpy values ​​of all the phase change wax samples; Performing partial least squares regression analysis on the first characteristic spectrum matrix and the enthalpy value matrix to establish a first correction model; Selecting a first spectrum fitting residual matrix corresponding to the f-th principal factor according to the first correction model, using the first spectrum fitting residual matrix as an input signal of an extreme learning machine, using the enthalpy matrix as a teacher signal of the extreme learning machine, and establishing a second correction model; Collecting a sample spectrum of the phase change material sample to be tested; obtaining the absorbance of a characteristic spectrum region of the sample spectrum to be tested; determining a first enthalpy value prediction value according to the absorbance of the characteristic spectrum region of the sample spectrum to be tested and the first calibration model; Selecting a second spectrum fitting residual matrix corresponding to the f-th principal factor of the phase change material sample to be tested according to the first calibration model, and determining a second enthalpy value prediction value according to the second spectrum fitting residual matrix and the second calibration model; Determine respective prediction weights of the first enthalpy prediction value and the second enthalpy prediction value, and weight the first enthalpy prediction value and the second enthalpy prediction value according to their respective prediction weights to obtain a predicted enthalpy value of the phase change material sample to be tested.

2. The method according to claim 1, wherein: The characteristic spectrum range is 5780-7700 cm -1 .

3. The method according to claim 1, wherein: The method further includes: The spectrum fitting residual matrix Ex is determined according to the following formula (1), formula (2) or formula (3): Ex=X-X(f best ) (1); Where X represents the spectral matrix, X(f best ) indicates that the first correction model takes the best principal factor f best Spectral matrix at ; Ex=X-X(f best+1～n ) (2); Where X(f best+1～n ) indicates that the first correction model takes the best principal factor f best The spectrum matrix of any one of the 1 to n principal factors after , where n is an integer from 2 to 15; Ex = X (3); Optionally, the number of the spectral fitting residual matrices Ex is one or more.

4. The method according to claim 1, wherein: The method further includes: The regression coefficients of the first correction model and the second correction model are respectively obtained, and the regression coefficients are used as prediction weights corresponding to the first enthalpy value prediction value and the second enthalpy value prediction value.

5. The method according to claim 1, wherein: The method further includes: Multiple predictions are performed based on the second spectral fitting residual matrix and the second correction model, and an average value of the multiple prediction results is taken as the second enthalpy value prediction value.

6. The method according to claim 1, wherein: The method includes: Performing a second-order differential process on each of the near-infrared spectra to obtain a first differential spectrum; The first differential absorbance of the characteristic spectrum region in each of the first differential spectra is obtained to form the first characteristic spectrum matrix.

7. The method according to claim 6, wherein: The method includes: Performing second-order differential processing on the spectrum of the sample to be tested to obtain a second differential spectrum; Obtaining absorbance of a characteristic spectral region of the second differential spectrum; The first enthalpy value prediction value is determined according to the absorbance of the characteristic spectral region of the second differential spectrum and the first calibration model.

8. The method according to claim 6 or 7, wherein: The window width of the second-order differential processing is 13-25.

9. The method according to claim 1, wherein: The phase change material is phase change wax.

10. The method according to claim 1, wherein: The standard enthalpy value is measured by differential scanning calorimetry.

Images