Granted Patent B2

US 12,399,223 · App. 18/555,121 · Granted Aug 26, 2025

Inventors: Yuji Kurauchi (Tokyo, JP); Shimpei Takemoto (Tokyo, JP); Yoshishige Okuno (Tokyo, JP)

Assignee: Resonac Corporation

G01R31/367G01R31/3842G01R31/392

Abstract

Claims (35)

1. An analyzing device comprising:

a processor; and

a memory storing program instructions that cause the processor to:

acquire lifetime data from cycle measurement data of a target battery;

calculate, by factoring a relationship between a voltage and a current capacity calculated from the cycle measurement data of the target battery, factor intensity transition data indicating a change in intensity of each factor affecting the current capacity and factor data representing a relationship between a voltage and a current capacity of each factor; and

output the lifetime data, the factor data, and the factor intensity transition data.

2. The analyzing device as claimed in claim 1 , wherein the relationship between the voltage and the current capacity is calculated based on a QV curve or a dQ/dV curve generated from the cycle measurement data.

3. The analyzing device as claimed in claim 2 , wherein the processor performs the factoring by non-negative matrix factorization.

4. A predicting device comprising:

a processor; and

a memory storing program instructions that cause the processor to:

calculate, by factoring a relationship between a voltage and a current capacity calculated from cycle measurement data of a target battery up to a predetermined cycle, factor intensity transition data indicating a change up to the predetermined cycle in intensity of each factor affecting the current capacity;

generate a feature from the factor intensity transition data of each factor; and

predict a lifetime characteristic of the target battery by inputting the generated feature into a model that has learned a relationship between a feature of each factor and lifetime data that are generated based on cycle measurement data for training up to the predetermined cycle.

5. The predicting device as claimed in claim 4 , wherein the program instructions further cause the processor to output a prediction value of the lifetime characteristic and factor data representing a relationship between the voltage and the current capacity of a predetermined factor calculated from the cycle measurement data for training.

6. The predicting device as claimed in claim 5 , wherein the program instructions further cause the processor to:

acquire a contribution state to the prediction of the lifetime characteristic of the target battery for each generated feature by analyzing the model; and

extract the feature whose contribution state satisfies a predetermined condition from among the generated features to identify a corresponding factor,

wherein the processor outputs the prediction value of the lifetime characteristic of the target battery, the factor data related to the identified factor, and the contribution state corresponding to the identified factor.

7. An analyzing method performed by a computer, comprising:

acquiring lifetime data from cycle measurement data of a target battery;

calculating, by factoring a relationship between a voltage and a current capacity calculated from the cycle measurement data of the target battery, factor intensity transition data indicating a change in intensity of each factor affecting the current capacity and factor data representing a relationship between a voltage and a current capacity of each factor; and

outputting the lifetime data, the factor data, and the factor intensity transition data.

8. A predicting method performed by a computer, comprising:

calculating, by factoring a relationship between a voltage and a current capacity calculated from cycle measurement data of a target battery up to a predetermined cycle, factor intensity transition data indicating a change up to the predetermined cycle in intensity of each factor affecting the current capacity;

generating a feature from the factor intensity transition data of each factor; and

predicting a lifetime characteristic of the target battery by inputting the generated feature into a model that has learned a relationship between a feature of each factor and lifetime data that are generated based on cycle measurement data for training up to the predetermined cycle.

9. A non-transitory computer-readable storage medium having stored therein a program for causing a computer to perform:

acquiring lifetime data from cycle measurement data of a target battery;

calculating, by factoring a relationship between a voltage and a current capacity calculated from the cycle measurement data of the target battery, factor intensity transition data indicating a change in intensity of each factor affecting the current capacity and factor data representing a relationship between a voltage and a current capacity of each factor; and

outputting the lifetime data, the factor data, and the factor intensity transition data.

10. A non-transitory computer-readable storage medium having stored therein a program for causing a computer to perform:

calculating, by factoring a relationship between a voltage and a current capacity calculated from cycle measurement data of a target battery up to a predetermined cycle, factor intensity transition data indicating a change up to the predetermined cycle in intensity of each factor affecting the current capacity;

generating a feature from the factor intensity transition data of each factor; and

predicting a lifetime characteristic of the target battery by inputting the generated feature into a model that has learned a relationship between a feature of each factor and lifetime data that are generated based on cycle measurement data for training up to the predetermined cycle.

Assignments (2)

Priority Claims (1)

Continuity (1)

References Cited (10)