Granted Patent B2

US 12,488,239 · App. 17/774,889 · Granted Dec 2, 2025

Inventor: Kyohei Hanaoka (Tokyo, JP)

G06N3/08G16C20/30G16C20/70

Abstract

Claims (60)

1 . An input data generation system, comprising:

at least one processor,

wherein the at least one processor is configured to:

receive at least an input of:

first molecular graph data specifying a molecular graph corresponding to a first monomer, wherein the first molecular graph data includes node information specifying a node of the molecular graph and edge information specifying an edge of the molecule graph;

second molecular graph data specifying a molecular graph corresponding to a second monomer, wherein the second molecular graph data includes node information specifying a node of the molecular graph, and edge information specifying an edge of the molecule graph; and

mixing rate data indicating a first mixing rate of the first monomer and a second mixing rate of the second monomer in a polymer alloy generated by mixing the first monomer and the second monomer;

generate synthetic molecular graph data by:

combining the node information of the first molecular graph data and the node information of the second molecular graph data to generate a node set data;

combining the edge information of the first molecular graph data and the edge information of the second molecular graph data to generate an edge set data; and

combining the node set data and the edge set data;

convert the synthetic molecular graph data into a feature vector;

generate input data for a neural network having a graph as its input:

by reflecting the mixing rate data on elements of the feature vector corresponding to the node information and the edge information of the first molecule graph data; and

by reflecting the mixing rate data on elements of the feature vector corresponding to the node information and the edge information of the second molecule graph data;

input the input data to the neural network to generate a prediction result of thermal properties, mechanical properties, or adhesiveness of the polymer alloy; and

output the prediction result.

2 . The input data generation system according to claim 1 ,

wherein the at least one processor is configured to multiply, add or connect the mixing rate of each of the first molecule monomer and second molecule monomer to the vector corresponding to the node information of the first molecular graph data and the second molecular graph data.

3 . The input data generation system according to claim 1 ,

wherein the at least one processor is configured to multiply, add or connect the mixing rate of each of the first molecule monomer and second molecule monomer to the vector corresponding to the edge information of the first molecular graph data and the second molecular graph data.

4 . The input data generation system according to claim 1 ,

wherein the at least one processor is further configured to:

receive, as the first molecular graph data and the second molecular graph data, bondable node information specifying nodes of a molecular graph that can be bonded to each other;

generate additional edge information regarding an edge bonding two nodes among nodes indicated by the bondable node information included in the first molecular graph data and nodes indicated by the bondable node information included in the second molecular graph data;

generate the synthetic molecular graph data by adding the additional edge information; and

generate the input data by reflecting the mixing rate of each of the first monomer and the second monomer on a vector corresponding to the additional edge information in the feature vector.

5 . The input data generation system according to claim 1 ,

wherein the first monomer and the second monomer are monomers, and

the mixing rate data indicates a mixing rate of each of the first monomer and the second monomer in a polymer alloy generated based on the first monomer and the second monomer.

6 . An input data generation method executed by a computer including at least one processor, the method comprising:

receiving at least an input of:

first molecular graph data specifying a molecular graph corresponding to a first monomer, wherein the first molecular graph data includes node information specifying a node of the molecular graph and edge information specifying an edge of the molecule graph;

second molecular graph data specifying a molecular graph corresponding to a second monomer, wherein the second molecular graph data includes node information specifying a node of the molecular graph, and edge information specifying an edge of the molecule graph; and

mixing rate data indicating a first mixing rate of the first monomer and a second mixing rate of the second monomer in a polymer alloy generated by mixing the first monomer and the second monomer;

generating synthetic molecular graph data by:

combining the node information of the first molecular graph data and the node information of the second molecular graph data to generate a node set data;

combining the edge information of the first molecular graph data and the edge information of the second molecular graph data to generate an edge set data; and

combining the node set data and the edge set data;

converting the synthetic molecular graph data into a feature vector;

generating input data for a neural network having a graph as its input:

by reflecting the mixing rate data on elements of the feature vector corresponding to the node information and the edge information of the first molecule graph data; and

by reflecting the mixing rate data on elements of the feature vector corresponding to the node information and the edge information of the second molecule graph data;

inputting the input data to the neural network to generate a prediction result of thermal properties, mechanical properties, or adhesiveness of the polymer alloy; and

outputting the prediction result.

7 . A non-transitory computer-readable storage medium storing an input data generation program causing a computer to execute:

receiving at least an input of:

first molecular graph data specifying a molecular graph corresponding to a first monomer, wherein the first molecular graph data includes node information specifying a node of the molecular graph and edge information specifying an edge of the molecule graph;

second molecular graph data specifying a molecular graph corresponding to a second monomer, wherein the second molecular graph data includes node information specifying a node of the molecular graph, and edge information specifying an edge of the molecule graph; and

mixing rate data indicating a first mixing rate of the first monomer and a second mixing rate of the second monomer in a polymer alloy generated by mixing the first monomer and the second monomer;

generating synthetic molecular graph data by:

combining the node information of the first molecular graph data and the node information of the second molecular graph data to generate a node set data;

combining the edge information of the first molecular graph data and the edge information of the second molecular graph data to generate an edge set data; and

combining the node set data and the edge set data;

converting the synthetic molecular graph data into a feature vector;

generating input data for a neural network having a graph as its input:

by reflecting the mixing rate data on elements of the feature vector corresponding to the node information and the edge information of the first molecule graph data; and

by reflecting the mixing rate data on elements of the feature vector corresponding to the node information and the edge information of the second molecule graph data;

inputting the input data to the neural network to generate a prediction result of thermal properties, mechanical properties, or adhesiveness of the polymer alloy; and

outputting the prediction result.

Assignments (3)

Priority Claims (1)

Continuity (1)

References Cited (70)