Resin property value prediction apparatus, learning apparatus, and method for predicting resin property value

Patent metadata

Field	Value
Publication number	US-12429378-B2
Application number	US-202318118830-A
Country	US
Kind code	B2
Filing date	Mar 8, 2023
Priority date	Mar 14, 2022
Publication date	Sep 30, 2025
Grant date	Sep 30, 2025

Abstract

To calculate the properties of a resin composition in a short time and with high accuracy, a resin property value prediction apparatus includes an acquisition unit acquiring spectral data obtained by measuring a resin composition by a spectroscopic sensor in a process of performing a resin polymerization reaction in a batch type reaction tank and a calculation unit, using a learned model learned with spectral data obtained by measuring the resin composition in advance and a resin property value as training data, calculating the resin property value of the resin composition having the spectral data acquired by the acquisition unit.

First claim

What is claimed is:

1. A resin property value prediction apparatus comprising:
an acquisition unit acquiring spectral data obtained by measuring a resin composition by a spectroscopic sensor in a process of performing a resin polymerization reaction in a batch type reaction tank; and
a calculation unit, using a learned model learned with spectral data obtained by measuring the resin composition in advance and a resin property value as training data, calculating the resin property value of the resin composition having the spectral data acquired by the acquisition unit,
wherein the spectral data included in the training data is spectral data obtained by performing smoothing processing on spectral data measured by the spectroscopic sensor in advance, and
wherein the calculation unit further includes a preprocessing unit performing smoothing processing on the spectral data acquired by the acquisition unit and calculates the resin property value based on the spectral data with the smoothing processing performed by the preprocessing unit.

2. The resin property value prediction apparatus according to claim 1 , wherein
the spectral data included in the training data is spectral data obtained by performing differentiating processing on spectral data measured by the spectroscopic sensor in advance,
the preprocessing unit further performs differentiating processing on the spectral data acquired by the acquisition unit, and
the calculation unit calculates the resin property value based on the spectral data with the differentiating processing performed by the preprocessing unit.

3. The resin property value prediction apparatus according to claim 2 , wherein
the preprocessing unit performs the smoothing processing on the spectral data acquired by the acquisition unit and the differentiating processing on the spectral data acquired by the acquisition unit independently from each other, and
the calculation unit calculates the resin property value based on any of the spectral data with the smoothing processing performed by the preprocessing unit, the spectral data with the differentiating processing performed by the preprocessing unit, and spectral data with neither the smoothing processing nor the differentiating processing performed by the preprocessing unit.

4. The resin property value prediction apparatus according to claim 3 , wherein the calculation unit performs logarithmic transformation processing on the resin property value independently from the smoothing processing and the differentiating processing and outputs either the resin property value with the logarithmic transformation processing performed or the resin property value without the logarithmic transformation processing performed.

5. The resin property value prediction apparatus according to claim 2 , wherein the preprocessing unit performs the smoothing processing by local regression with a window length of less than 41.

6. The resin property value prediction apparatus according to claim 1 , wherein the acquisition unit acquires spectral data measured after a sampled resin composition is allowed to cool for a certain time.

7. The resin property value prediction apparatus according to claim 1 , wherein the acquisition unit acquires spectral data measured at every certain time while the resin polymerization reaction is being performed in the batch type reaction tank.

8. The resin property value prediction apparatus according to claim 1 , wherein
the resin composition is an acrylic resin, and
the acquisition unit acquires spectral data obtained by measuring the acrylic resin by the spectroscopic sensor.

9. The resin property value prediction apparatus according to claim 1 , wherein the spectroscopic sensor measuring the spectral data acquired by the acquisition unit is at least either a near-infrared (NIR) spectroscopic sensor or a Raman spectroscopic sensor.

10. A method for predicting a resin property value, the method comprising:
acquiring spectral data obtained by measuring a resin composition by a spectroscopic sensor in a process of performing a resin polymerization reaction in a batch type reaction tank; and
using a learned model learned with spectral data obtained by measuring the resin composition in advance and a resin property value as training data, calculating the resin property value of the resin composition having the spectral data acquired at the acquiring,
wherein the spectral data included in the training data is spectral data obtained by performing smoothing processing on spectral data measured by the spectroscopic sensor in advance, and
wherein the calculating further includes a preprocessing performing smoothing processing on the spectral data acquired and calculates the resin property value based on the spectral data with the smoothing processing.

Assignees

• Dainippon Ink & Chemicals
• Univ Meiji

Inventors

• Terauchi Kazutoshi
• Yamaji Toshinori
• Kaneko Hiromasa
• Yamakage Shuto

Classifications

• G01N21/65

  Raman scattering · CPC title

• G01N2021/8411

  Application to online plant, process monitoring · CPC title

• G01N21/01

  Arrangements or apparatus for facilitating the optical investigation · CPC title

• G01N21/359 Primary

  using near infrared light · CPC title

• G01N21/25

  Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands · CPC title