Method for Predicting Physical Properties of Resin, Information Processing Apparatus, and Program

Abstract

Improve the prediction technology of resin physical property values. 【Solution means】A method for predicting resin physical property values executed by an information processing apparatus, including: generating a prediction model based on teacher data with data in a predetermined wavelength region of spectrum data in a resin synthesis process as explanatory variables and resin physical property values as target variables; and predicting resin physical property values based on the prediction model, wherein the predetermined wavelength region is determined using complementary data generated based on the spectrum data.

Description

Technical Field

[0001] The present disclosure relates to a method for predicting resin physical property values, an information processing apparatus, and a program.

Background Art

[0002] In the polymerization process of resin, the physical properties of the resin composition are generally measured in a laboratory according to established procedures. Sampling and physical property inspection operations take several hours, and there may be variations in measurement accuracy by operators. When the physical properties of the resin composition do not meet the quality threshold, corrective measures such as adjusting the polymerization conditions are taken. These corrective measures are costly, labor-intensive, and have problems causing productivity decline. Also, when the physical properties of the resin composition do not meet the quality threshold, the polymerized products may be discarded. Therefore, a technique for generating a quality prediction model for predicting the physical property values of a resin composition by machine learning has been proposed (for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] A certain number of performance data are required as teacher data for building a machine learning model. However, due to the balance with normal operations, it may not be possible to sufficiently secure performance data, and there are problems in practical application. Thus, there is room for improvement in the prediction technology of resin physical property values.

[0005] In view of such circumstances, an object of the present disclosure is to improve the prediction technology of resin physical property values.

Means for Solving the Problems

[0006] (1) A method according to an embodiment of the present disclosure is a method for predicting resin physical property values executed by an information processing apparatus, comprising: generating a prediction model based on teacher data with data in a predetermined wavelength region of spectrum data in a resin synthesis process as explanatory variables and resin physical property values as target variables; predicting resin physical property values based on the prediction model; and the predetermined wavelength region is determined using complementary data generated based on the spectrum data.

[0007] (2) A method according to an embodiment of the present disclosure is the method according to (1), wherein the spectrum data includes data related to a plurality of resin compositions having different composition ratios of raw materials.

[0008] (3) A method according to an embodiment of the present disclosure is the method according to (2), wherein the spectrum data is near-infrared light spectrum data.

[0009] (4) A method according to an embodiment of the present disclosure is the method according to (3), wherein the spectroscopic sensor for measuring the near-infrared light spectrum data is at least one of at least a near-infrared spectroscopic sensor or a Raman spectroscopic sensor.

[0010] (5) A method according to an embodiment of the present disclosure is the method according to (3), wherein the explanatory variables further include the mass ratio of the raw material and the catalyst, and the temperature.

[0011] (6) A method according to an embodiment of the present disclosure is the method according to any one of (1) to (5), wherein the resin physical property value is epoxy monomer equivalent.

[0012] (7) A method according to an embodiment of the present disclosure is the method according to any one of (1) to (6), wherein In the step of generating the prediction model, a plurality of prediction models are respectively generated based on the teacher data by a plurality of machine learning algorithms, In the step of predicting, based on one prediction model selected from the plurality of prediction models according to a predetermined index, the resin physical property value is predicted.

[0013] (8) An information processing apparatus according to an embodiment of the present disclosure is an information processing apparatus including a control unit, wherein the control unit generates a prediction model based on teacher data having data in a predetermined wavelength range of spectrum data in a resin synthesis process as an explanatory variable and a resin physical property value as an objective variable, predicts a resin physical property value based on the prediction model, and the data in the predetermined wavelength range is determined using complementary data generated based on the spectrum data.

[0014] (9) A program according to an embodiment of the present disclosure causes a computer to generate a prediction model based on teacher data having data in a predetermined wavelength range of spectrum data in a resin synthesis process as an explanatory variable and a resin physical property value as an objective variable, predict a resin physical property value based on the prediction model, and execute, wherein the data in the predetermined wavelength range is determined using complementary data generated based on the spectrum data. [Advantages of the Invention]

[0015] According to an embodiment of the present disclosure, the prediction technology of resin physical property values is improved. [Brief Description of the Drawings]

[0016]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Mode for Carrying Out the Invention

[0017] Hereinafter, embodiments of the present disclosure will be described.

[0018] (Overview of the Embodiment)

[0019] Hereinafter, a method for predicting resin physical property values, an information processing apparatus, and a program in embodiments of the present disclosure will be described with reference to the drawings. The prediction target according to the embodiments of the present disclosure is the resin physical property value in the polymerization process of the resin. The resin composition to be targeted may widely include polymers such as homopolymers and copolymers. Further, the resin composition may be a thermoplastic resin or a thermosetting resin. The thermoplastic resin is not particularly limited, and examples thereof include polypropylene (PP), polyethylene (PE), ABS resin, polyvinyl chloride (PVC), acrylic resin, polyester resin, polystyrene resin (PS), urethane resin (PU), polyphenylene sulfide resin (PPS), and the like. The resin physical property value may be an NV value, viscosity, residual monomer concentration, molecular weight, or the like. In the following embodiments, as an example, the case where the resin composition is an epoxy resin will be described.

[0020] In each figure, the same or corresponding parts are denoted by the same reference numerals. In the description of this embodiment, the description of the same or corresponding parts will be omitted or simplified as appropriate.

[0021] First, the outline of this embodiment will be described, and the details will be described later. The method for predicting the resin physical property values in this embodiment is executed by the information processing apparatus 10. The information processing apparatus 10 generates a prediction model based on teacher data having a predetermined wavelength region of spectral data in the resin synthesis process as an explanatory variable and the resin physical property values as an objective variable. Further, the information processing apparatus 10 predicts the resin physical property values based on the prediction model. Here, the predetermined wavelength region is characterized in that it is determined using complementary data generated based on the spectral data.

[0022] Thus, according to this embodiment, complementary data is generated based on the spectral data. Further, a predetermined wavelength region is determined based on such complementary data. Therefore, even when sufficient actual data cannot be secured, the predetermined wavelength region can be determined by generating the complementary data. Also, the prediction technique of the resin physical property values is improved in that the resin physical property values can be predicted with high accuracy by using the spectral data in the predetermined wavelength region as an explanatory variable.

[0023] (Configuration of Information Processing Apparatus) As shown in FIG. 1, the information processing apparatus 10 includes a control unit 11, a storage unit 12, an input unit 13, an output unit 14, and a communication unit 15.

[0024] The control unit 11 includes at least one processor, at least one dedicated circuit, or a combination thereof. The processor is a general-purpose processor such as a CPU (central processing unit) or a GPU (graphics processing unit), or a dedicated processor specialized for specific processing. The dedicated circuit is, for example, an FPGA (field-programmable gate array) or an ASIC (application specific integrated circuit). The control unit 11 executes processing related to the operation of the information processing apparatus 10 while controlling each part of the information processing apparatus 10.

[0025] The storage unit 12 includes at least one semiconductor memory, at least one magnetic memory, at least one optical memory, or a combination of at least two of these. The semiconductor memory is, for example, a RAM (random access memory) or a ROM (read only memory). The RAM is, for example, an SRAM (static random access memory) or a DRAM (dynamic random access memory). The ROM is, for example, an EEPROM (electrically erasable programmable read only memory). The storage unit 12 functions as, for example, a main storage device, an auxiliary storage device, or a cache memory. The storage unit 12 stores data used for the operation of the information processing apparatus 10 and data obtained by the operation of the information processing apparatus 10.

[0026] The input unit 13 includes at least one input interface. The input interface is, for example, a physical key, a capacitive key, a pointing device, or a touch screen provided integrally with a display. The input interface may also be, for example, a sound sensor that receives voice input, or a camera that receives gesture input. The input unit 13 receives an operation for inputting data used for the operation of the information processing apparatus 10. Instead of being provided in the information processing apparatus 10, the input unit 13 may be connected to the information processing apparatus 10 as an external input device. As the connection method, for example, any method such as USB (Universal Serial Bus), HDMI (Registered Trademark) (High-Definition Multimedia Interface), or Bluetooth (Registered Trademark) can be used.

[0027] The output unit 14 includes at least one output interface. The output interface is, for example, a display that outputs information as video, or a speaker that outputs information as audio. The display is, for example, an LCD (liquid crystal display) or an organic EL (electro luminescence) display. The output unit 14 performs display output of data obtained by the operation of the information processing apparatus 10. Instead of being provided in the information processing apparatus 10, the output unit 14 may be connected to the information processing apparatus 10 as an external output device. As the connection method, for example, any method such as USB, HDMI (Registered Trademark), or Bluetooth (Registered Trademark) can be used.

[0028] The communication unit 15 includes at least one interface for external communication. The communication interface may be either a wired communication interface or a wireless communication interface. In the case of wired communication, the communication interface may be, for example, a LAN (Local Area Network) interface or a USB (Universal Serial Bus). In the case of wireless communication, the communication interface may be, for example, an interface corresponding to a mobile communication standard such as LTE (Long Term Evolution), 4G (4th generation), or 5G (5th generation), or an interface corresponding to short-range wireless communication such as Bluetooth (registered trademark). The communication unit 15 receives data used in the operation of the information processing apparatus 10 and transmits data obtained by the operation of the information processing apparatus 10.

[0029] The functions of the information processing apparatus 10 are realized by executing the program according to the present embodiment on a processor corresponding to the information processing apparatus 10. That is, the functions of the information processing apparatus 10 are realized by software. The program causes a computer to execute the operations of the information processing apparatus 10, thereby causing the computer to function as the information processing apparatus 10. That is, the computer functions as the information processing apparatus 10 by executing the operations of the information processing apparatus 10 according to the program.

[0030] In this embodiment, the program can be recorded on a computer-readable recording medium. The computer-readable recording medium includes a non-transitory computer-readable medium, and for example, a magnetic recording device, an optical disk, a magneto-optical recording medium, or a semiconductor memory. The distribution of the program is performed, for example, by selling, transferring, or lending a portable recording medium such as a DVD (digital versatile disc) or a CD-ROM (compact disc read only memory) on which the program is recorded. Also, the distribution of the program may be performed by storing the program in the storage of an external server and transmitting the program from the external server to another computer. Also, the program may be provided as a program product.

[0031] Some or all of the functions of the information processing apparatus 10 may be realized by a dedicated circuit corresponding to the control unit 11. That is, some or all of the functions of the information processing apparatus 10 may be realized by hardware.

[0032] (Operation of the information processing apparatus) With reference to FIG. 2, the operation of the information processing apparatus 10 according to this embodiment will be described.

[0033] Step S10: The control unit 11 of the information processing apparatus 10 acquires the spectral data in the resin synthesis process.

[0034] Any method can be adopted for acquiring the spectral data. For example, the control unit 11 may acquire the spectral data by receiving the spectral data from an external database or the like via the communication unit 21 and the network. Also, the method of measuring the spectral data may be either offline measurement or online measurement. The spectral data according to this embodiment may be, for example, near-infrared light spectral data. Also, the spectroscopic sensor for measuring the near-infrared light spectral data may be at least one of at least a near-infrared (NIR) spectroscopic sensor or a Raman spectroscopic sensor.

[0035] Step S20: The control unit 11 performs preprocessing on the spectral data acquired in Step S1. The amount of spectral data is huge, and if all the spectral data is used as explanatory variables of the prediction model, problems such as overfitting may occur. Therefore, by preprocessing, a predetermined wavelength region that has a large influence on the prediction result is selected from the spectral data. Any method such as a genetic algorithm described later can be adopted for the selection of the predetermined wavelength region.

[0036] Step S30: The control unit 11 generates a prediction model based on the training data with the data in the predetermined wavelength region of the spectral data in the resin synthesis process as explanatory variables and the resin physical property values as target variables. In other words, the control unit 11 generates a prediction model based on the preprocessed spectral data as explanatory variables and the training data with the resin physical property values as target variables.

[0037] Note that the explanatory variables are not limited to the data in the predetermined wavelength region of the spectral data described above. The explanatory variables may further include the mass ratio of the raw material and the catalyst, and the temperature. In this case, in Step S10, the control unit 11 also acquires information on the mass ratio of the raw material and the catalyst, and the temperature. Any method can be adopted for the acquisition of this information.

[0038] Step S40: The control unit 11 predicts the resin physical property values from the spectral data based on the generated prediction model. The control unit 11 may output the prediction result by the output unit 14.

[0039] In the step of generating the prediction model in step S30, a plurality of prediction models may be generated based on the training data by a plurality of machine learning algorithms. As algorithms for such a plurality of prediction models, PLS (Partial Least Squares Regression), Ridge, Lasso, EN (ElasticNet), SVR (Support Vector Regression) (non-linear), RF (Random Forest), GP_0 (Gaussian Process Regression_0) (linear), GP_1-10 (Gaussian Process Regression_1-10) (non-linear), XGBoost (eXtreme Gradient Boosting), Light gbm (Light Gradient Boosting Machine), GBDT (Gradient Boosting Decision Tree), etc. can be adopted. In this case, in the step of prediction in step S40, the resin physical property value may be predicted using one prediction model selected based on a predetermined index from the plurality of prediction models. The predetermined index is, for example, R2. When the predetermined index is R2, the control unit 11 predicts the resin physical property value using the prediction model with the R2 value closest to 1. Note that the predetermined index is not limited to R2, and for example, the predetermined index may be an error rate. Thus, in the present embodiment, the control unit 11 may select an optimal algorithm with high prediction accuracy as the algorithm of the prediction model.

[0040] (Example) Hereinafter, examples according to the present embodiment will be described. As data for the examples, at an experimental scale of 600 g, three product numbers (product number A, product number B, and product number C) were each synthesized in 4 lots or 2 lots according to predetermined raw materials and reaction conditions, and spectral data (here, NIR spectrum) and resin physical property values (here, epoxy monomer equivalent data) were obtained. The NIR spectrum was measured every minute. On the other hand, the epoxy monomer equivalent data was obtained by periodically sampling and analyzing.

[0041] FIG. 3 shows an overview diagram of a dataset of an example related to the accuracy verification of the prediction method according to the present embodiment. As shown in FIG. 3, for part number A, the dataset includes data of four lots, lot #1 to #4. Also, for part numbers B and C, the dataset includes data of two lots, lot #1 to #2 respectively.

[0042] For such a dataset, the above-described preprocessing is executed. The overview of the preprocessing is shown by the flowchart of FIG. 4.

[0043] Step S21: The control unit 11 determines whether the number of data of the spectral data of a certain part number is equal to or greater than a predetermined value. In this embodiment, the number of data corresponds to the number of lots for each part number. For example, the predetermined value may be 4 (lots). The control unit 11 determines whether sufficient data can be ensured for each part number. If the number of data of the spectral data is equal to or greater than the predetermined value, the process proceeds to step S22. On the other hand, if the number of data of the spectral data is less than the predetermined value, the process proceeds to step S22. Specifically, for part number A, the preprocessing is not executed. On the other hand, for part numbers B and C, the preprocessing is executed.

[0044] Step S22: The control unit 11 executes noise reduction processing to remove noise from the spectral data. Here, the spectral data may include information directly used for calculating the resin physical property values and unnecessary information (noise) that causes errors in calculating the resin physical property values. An example of the noise reduction processing is, for example, smoothing processing. In other words, the control unit 11 performs noise reduction on the spectral data by performing smoothing processing on the acquired spectral data. Specifically, the control unit 11 may perform smoothing processing by approximating to a polynomial. Preferably, the control unit 11 may perform smoothing by local regression. In this case, by varying the local width (kernel width), necessary information may be lost, or noise may not be completely removed. Therefore, the control unit 11 may regress the spectral data by selecting an optimal local width. Note that, as a typical method of the smoothing processing as described above, the Savitzky-Golay method (hereinafter also referred to as the SG method) based on the least squares method is known. The control unit 11 may execute noise reduction processing by the SG method.

[0045] Step S23: The control unit 11 selects a predetermined wavelength region based on the spectral data. Any method can be adopted for the selection of the predetermined wavelength region. For example, among the spectral data, a wavelength region that has a large influence on the prediction result may be selected by a genetic algorithm or the like. The parameter settings when using the genetic algorithm may be, for example, as follows. · Number of individuals in the genetic algorithm: 100 · Number of generations in the genetic algorithm: 100 · Selection region for wavelength selection: 1 to 10 · Wavelength width for wavelength selection: 50 · Method: supervised learning: partial least squares regression (PLS) · Number of executions: 10 times

[0046] Step S24: When the number of pieces of spectrum data is less than a predetermined value, the control unit 11 generates complementary data based on the acquired spectrum data. Any method can be adopted as the method for generating the complementary data. For example, the control unit 11 may generate complementary data by a method such as linear complementation or non-linear complementation based on the acquired spectrum data. For example, when the number of actually measured values is about 10, the control unit 11 generates about 300 pieces of complementary data.

[0047] Step S25: The control unit 11 executes noise reduction processing for removing noise from the spectrum data and the complementary data. An example of the noise reduction processing is, for example, smoothing processing. In other words, the control unit 11 performs noise reduction on the spectrum data and the complementary data by performing smoothing processing on the acquired spectrum data and the complementary data. Specifically, the control unit 11 may perform smoothing processing by approximating with a polynomial. Preferably, the control unit 11 may perform smoothing by local regression. In this case, by varying the local width (kernel width), necessary information may be lost or noise may not be completely removed. Therefore, the control unit 11 may regress the spectrum data and the complementary data by selecting an optimal local width. Note that the control unit 11 may execute noise reduction processing by the SG method.

[0048] Step S26: The control unit 11 selects a predetermined wavelength region based on the spectrum data and the complementary data. Any method can be adopted for the selection of the predetermined wavelength region. For example, among the spectrum data and the complementary data, a wavelength region having a large influence on the prediction result may be selected by a genetic algorithm or the like. The parameter settings when using the genetic algorithm may be, for example, as follows. · Number of individuals in the genetic algorithm: 100 · Number of generations of the genetic algorithm: 100 · Selection region for wavelength selection: 1 to 10 · Wavelength width for wavelength selection: 50 · Method: Teacher-assisted learning: Partial Least Squares Regression (PLS) · Number of executions: 10 times

[0049] Step S27: The control unit 11 deletes the complementary data from the data set. In other words, the control unit 11 uses the acquired spectral data for the learning of the prediction model described later, and does not use the complementary data for the learning of the prediction model described later. That is, the control unit 11 deletes the amplified complementary data and arranges the explanatory variables so that the data is composed only of the measured values.

[0050] Step S28: The control unit 11 merges the data with other product number data having a data number equal to or greater than a predetermined value. Specifically, in this embodiment, the control unit 11 merges the data of product number B with the data of product number A. Also, the control unit 11 merges the data of product number C with the data of product number A. In this way, the spectral data in the teacher data may include data related to a plurality of resin compositions having different composition ratios of raw materials. By doing so, the prediction accuracy can be improved as described later.

[0051] Referring to FIGS. 5 and 6, the prediction results and prediction accuracy by the prediction model generated using only the data of product number B as teacher data are shown. Here, Ridge is selected as the algorithm of the prediction model. The horizontal axis in FIG. 5 is the label number, and the vertical axis is the value of the epoxy monomer equivalent. The label number is an identification number associated with each sampling of each lot in the synthesis process. Also, the horizontal axis in FIG. 6 is the actual value, and the vertical axis is the predicted value corresponding to the actual value. The R2 value of the prediction model calculated from the results in FIG. 6 is 0.641. From these results, even using the prediction model with only the data of product number B as teacher data, the epoxy monomer equivalent data, which is the target variable, can be predicted with a certain prediction accuracy.

[0052] Referring to FIGS. 7 and 8, the prediction results and prediction accuracy by the prediction model generated using the teacher data obtained by merging the data of part number B and the data of part number A are shown. Here, Ridge is selected as the algorithm of the prediction model. The horizontal axis in FIG. 7 is the label number, and the vertical axis is the value of the epoxy monomer equivalent. Also, the horizontal axis in FIG. 8 is the actual value, and the vertical axis is the predicted value corresponding to the actual value. The R2 value of the prediction model calculated from the results in FIG. 8 is 0.829. From these results, by using the teacher data obtained by merging the data of part number B and the data of part number A, it is possible to predict the epoxy monomer equivalent data, which is the target variable, with a higher prediction accuracy than using only the data of part number B as the teacher data.

[0053] Referring to FIGS. 9 and 10, the prediction results by the prediction model generated using only the data of part number C as the teacher data are shown. Here, PLS is selected as the algorithm of the prediction model. The horizontal axis in FIG. 9 is the label number, and the vertical axis is the value of the epoxy monomer equivalent. Also, the horizontal axis in FIG. 10 is the actual value, and the vertical axis is the predicted value corresponding to the actual value. The R2 value of the prediction model calculated from the results in FIG. 10 is 0.957. From these results, even by using the prediction model with only the data of part number C as the teacher data, it is possible to predict the epoxy monomer equivalent data, which is the target variable, with a certain prediction accuracy.

[0054] Referring to FIGS. 11 and 12, the prediction results by the prediction model generated using the teacher data obtained by merging the data of part number C and the data of part number A are shown. Here, lasso is selected as the algorithm of the prediction model. The horizontal axis in FIG. 11 is the label number, and the vertical axis is the value of the epoxy monomer equivalent. Also, the horizontal axis in FIG. 12 is the actual value, and the vertical axis is the predicted value corresponding to the actual value. The R2 value of the prediction model calculated from the results in FIG. 12 is 0.979. From these results, by using the teacher data obtained by merging the data of part number C and the data of part number A, it is possible to predict the epoxy monomer equivalent data, which is the target variable, with a higher prediction accuracy than using only the data of part number C as the teacher data.

[0055] As described above, the information processing apparatus 10 according to the present embodiment generates complementary data based on spectral data. Further, the information processing apparatus 10 determines a predetermined wavelength region based on the complementary data. According to such a configuration, even when sufficient performance data cannot be secured, the predetermined wavelength region can be determined by generating complementary data. Further, by using the spectral data of the predetermined wavelength region as an explanatory variable, the resin physical property value prediction technology is improved in that the resin physical property value can be predicted with high accuracy.

[0056] Although the present disclosure has been described based on the drawings and examples, it should be noted that those skilled in the art may make various modifications and alterations based on the present disclosure. Therefore, it should be noted that these modifications and alterations are included in the scope of the present disclosure. For example, the functions included in each component or each step, etc. can be rearranged so as not to be logically contradictory, and a plurality of components or steps, etc. can be combined into one or divided.

[0057] Also, for example, in step S22 or step S25, the control unit 11 may perform a differentiation process in addition to or instead of the smoothing process. By performing the differentiation process by the control unit 11, each spectral information can be extracted from the overlapping peaks, and the baseline can be corrected. When the control unit 11 performs the differentiation process, the above-described prediction model is generated based on the spectral data on which the differentiation process has been performed. Note that performing the differentiation process may increase the noise in the spectrum. Therefore, depending on the spectral data, there are cases where it is better to perform the differentiation process and cases where it is better not to perform the differentiation process. Therefore, the control unit 11 may perform a smoothing process on the spectral data, further perform a differentiation process separately and independently, and generate a prediction model using the optimal spectral data.

[0058] For example, in step S22 or step S25, in addition to smoothing and / or differentiating the spectral data, the control unit 11 may perform a logarithmic conversion process on the resin physical property value. For example, when the relationship between the spectral data and the resin physical property value follows the Arrhenius-type physical law, it is effective to perform a logarithmic conversion process on the resin physical property value. On the other hand, when the resin physical property value is logarithmically converted, the prediction error of the resin physical property value may increase. That is, when the logarithmic conversion process of the resin physical property value is performed, there are cases where the accuracy is improved and cases where it is not. Therefore, the control unit 11 may output either the resin physical property value on which the logarithmic conversion process has been performed or the resin physical property value on which the logarithmic conversion process has not been performed as the resin physical property value. Note that the logarithmic conversion process of the resin physical property value is performed independently of the smoothing process and the differentiation process.

Explanation of Signs

[0059] 10 Information processing apparatus 11 Control unit 12 Storage unit 13 Input unit 14 Output unit 15 Communication unit

Claims

1. A method for predicting resin physical property values executed by an information processing apparatus, comprising: generating a prediction model based on teacher data with data in a predetermined wavelength region of spectral data in a resin synthesis process as explanatory variables and resin physical property values as target variables; predicting resin physical property values based on the prediction model; including: the predetermined wavelength region is determined using complementary data generated based on the spectral data; the explanatory variables further include the mass ratio of raw materials and catalyst, and temperature.

2. The method according to claim 1, wherein the spectral data includes data related to a plurality of resin compositions having different composition ratios of raw materials.

3. The method according to claim 1 or 2, wherein the spectral data is near-infrared light spectral data.

4. The method according to claim 3, wherein the spectroscopic sensor for measuring the near-infrared light spectral data is at least one of at least a near-infrared spectroscopic sensor or a Raman spectroscopic sensor.

5. The method according to claim 1, wherein the resin physical property value is epoxy monomer equivalent.

6. The method according to claim 1, wherein in the step of generating the prediction model, a plurality of prediction models are generated based on the teacher data by a plurality of machine learning algorithms; in the step of predicting, the resin physical property value is predicted based on one prediction model selected from the plurality of prediction models based on a predetermined index.

7. An information processing apparatus comprising a control unit, wherein the control unit: generates a prediction model based on teacher data with data in a predetermined wavelength region of spectral data in a resin synthesis process as explanatory variables and resin physical property values as target variables; predicts resin physical property values based on the prediction model; the data in the predetermined wavelength region is determined using complementary data generated based on the spectral data; the explanatory variables further include the mass ratio of raw materials and catalyst, and temperature.

8. Causing a computer to generate a prediction model based on teacher data with data in a predetermined wavelength region of spectral data in a resin synthesis process as explanatory variables and resin physical property values as target variables; predict resin physical property values based on the prediction model; and execute, wherein the data in the predetermined wavelength region is determined using complementary data generated based on the spectral data. The explanatory variables further include the mass ratio of the raw material to the catalyst and the temperature, and the program.

Images