# US 12,181,529 B2 · Learning apparatus, estimation apparatus, learning method, estimation method, recording medium having recorded thereon learning program, and recording medium having recorded thereon estimation program > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,181,529 B2 - Document ID: US12181529B2 - Status: Granted - Kind: B2 - Application number: 17/359,638 - Filed: 2021-06-28 - Published: 2021-10-21 - Granted: 2024-12-31 - Earliest claimed: 2018-12-28 - Assignees: Yokogawa Electric Corporation - Applicants: Yokogawa Electric Corporation - Inventors: Soichiro Torai - Web page: https://patentgrep.com/patent/US12181529B2 ## CPC - G01R 31/367: Software therefor, e.g. for battery testing using modelling or look-up tables - G06N 3/04: Architecture, e.g. interconnection topology - G01R 31/3842: combining voltage and current measurements - G06N 3/0499: Feedforward networks - G06N 3/088: Non-supervised learning, e.g. competitive learning - G06N 3/0895: Weakly supervised learning, e.g. semi-supervised or self-supervised learning - G06N 3/09: Supervised learning - H01M 10/44: Methods for charging or discharging (circuits for charging H02J7/00) - H01M 10/48: Accumulators combined with arrangements for measuring, testing or indicating the condition of cells, e.g. the level or density of the electrolyte (constructional details of current conducting connections for detecting conditions inside cells or batteries, e.g. details of voltage sensing terminals, H01M50/569) - H02J 7/80: including monitoring or indicating arrangements - G06N 20/00: Machine learning - H01M 2220/20: Batteries in motive systems, e.g. vehicle, ship, plane - Y02E 60/10: Energy storage using batteries - G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - H01M: PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY - H02J: ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY - Y02E: REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION ## Abstract A learning apparatus is provided, which comprises: a characteristic acquisition unit for acquiring one of a voltage and a capacity of a storage battery and a characteristic corresponding to a capacity change and a voltage change of the storage battery under a plurality of conditions during at least one of charging and discharging of the storage battery; and a learning unit for learning a relationship between one of the voltage and the capacity of the storage battery and the characteristic under each of the plurality of conditions to generate an estimation model for estimating at least one of a maximum capacity and a remaining capacity of the storage battery from one of the voltage and the capacity of the storage battery and the characteristic. ## Claims ### Claim 1 A learning apparatus comprising: a characteristic acquisition unit for acquiring one of a voltage and a capacity of a storage battery and a characteristic corresponding to a capacity change and a voltage change of the storage battery under a plurality of conditions during at least one of charging and discharging of the storage battery; and a learning unit for learning a relationship between one of the voltage and the capacity of the storage battery and the characteristic under each condition of the plurality of conditions to generate an estimation model for estimating at least one of a maximum capacity and a remaining capacity of the storage battery from one of the voltage and the capacity of the storage battery and the characteristic; wherein the learning unit is configured to use a modular network Self Organizing Map, wherein a module is a neural network, to generate the estimation model. ### Claim 2 (depends on 1) The learning apparatus according to claim 1, wherein the learning unit is configured to generate the estimation model, which interpolates the relationship under a different condition from the relationship under at least one of the plurality of conditions. ### Claim 3 (depends on 1) The learning apparatus according to claim 1, wherein the characteristic is a differential capacity characteristic calculated by differentiating the capacity of the storage battery by the voltage. ### Claim 4 (depends on 3) The learning apparatus according to claim 3, wherein the learning unit is configured to associate the maximum capacity of the storage battery with each condition individually. ### Claim 5 (depends on 4) The learning apparatus according to claim 4, wherein the learning unit is configured to calculate the maximum capacity of the storage battery to associate with each condition based on an integration characteristic calculated by integrating the differential capacity characteristic by the voltage of the storage battery. ### Claim 6 (depends on 1) The learning apparatus according to claim 1, wherein the characteristic is a differential voltage characteristic calculated by differentiating the voltage of the storage battery by the capacity. ### Claim 7 An estimation apparatus comprising: a model acquisition unit for acquiring an estimation model for estimating at least one of a maximum capacity and a remaining capacity of a storage battery from a relationship between one of a voltage and a capacity of the storage battery and a characteristic corresponding to a capacity change and a voltage change of the storage battery under a plurality of conditions during at least one of charging and discharging of the storage battery; a data acquisition unit for acquiring a time series data regarding the voltage and a current of the storage battery; a calculation unit for calculating the characteristic based on the time series data regarding the voltage and the current of the storage battery; and an estimation unit for using the estimation model to estimate at least one of the maximum capacity and the remaining capacity of the storage battery from one of the voltage and the capacity of the storage battery and the calculated characteristic; wherein the estimation model is learned using a modular network Self Organizing Map, wherein a module is a neural network. ### Claim 8 (depends on 7) The estimation apparatus according to claim 7, wherein the estimation model is for interpolating the relationship under a different condition from the relationship under at least one of the plurality of conditions. ### Claim 9 (depends on 7) The estimation apparatus according to claim 7, wherein the characteristic is a differential capacity characteristic calculated by differentiating the capacity of the storage battery by the voltage. ### Claim 10 (depends on 9) The estimation apparatus according to claim 9, wherein the estimation unit is configured to estimate the maximum capacity of the storage battery based on an integration characteristic calculated by integrating the differential capacity characteristic by the voltage of the storage battery. ### Claim 11 (depends on 7) The estimation apparatus according to claim 7, wherein the characteristic is a differential voltage characteristic calculated by differentiating the voltage of the storage battery by the capacity. ### Claim 12 (depends on 9) The estimation apparatus according to claim 9, wherein the estimation unit is configured to acquire a model differential voltage characteristic calculated by differentiating the voltage of the storage battery by the capacity from the estimation model to estimate the remaining capacity of the storage battery with reference to the model differential voltage characteristic. ### Claim 13 (depends on 12) The estimation apparatus according to claim 12, wherein the estimation unit is configured to perform fitting on a partial differential voltage characteristic calculated by differentiating the voltage of the storage battery by the capacity based on the time series data of the voltage and the current of the storage battery with reference to the model differential voltage characteristic to estimate the remaining capacity of the storage battery. ### Claim 14 (depends on 7) The estimation apparatus according to claim 7, the estimation unit is configured to judge whether it can estimate at least one of the maximum capacity and the remaining capacity of the storage battery based on the time series data of the voltage and the current of the storage battery. ### Claim 15 A learning method for a learning apparatus to learn comprising: acquiring one of a voltage and a capacity of a storage battery and a characteristic corresponding to a capacity change and a voltage change of the storage battery under a plurality of conditions during at least one of charging and discharging of the storage battery; and learning a relationship between one of the voltage and the capacity of the storage battery and the characteristic under each condition of the plurality of conditions to generate an estimation model for estimating at least one of a maximum capacity or a remaining capacity of the storage battery from one of the voltage and the capacity of the storage battery and the characteristic; wherein the estimation model is learned using a modular network Self Organizing Map, wherein a module is a neural network. ### Claim 16 An estimation method in which an estimation apparatus estimates comprising: acquiring an estimation model for estimating at least one of a maximum capacity and a remaining capacity of a storage battery from a relationship between one of a voltage and a capacity of the storage battery and a characteristic corresponding to a capacity change and a voltage change of the storage battery under a plurality of conditions during at least one of charging and discharging of the storage battery; acquiring a time series data of the voltage and a current of the storage battery; calculating the characteristic based on the time series data of the voltage and the current of the storage battery; and using the estimation model to estimate at least one of the maximum capacity and the remaining capacity of the storage battery from one of the voltage and the capacity of the storage battery and the calculated characteristic; wherein the estimation model is learned using a modular network Self Organizing Map, wherein a module is a neural network. ### Claim 17 A non-transitory recording medium having recorded thereon a learning program to be implemented by a computer for causing the computer to function as: a characteristic acquisition unit for acquiring one of a voltage and a capacity of a storage battery and a characteristic corresponding to a capacity change and a voltage change of the storage battery under a plurality of conditions during at least one of charging and discharging of the storage battery; a learning unit for learning a relationship between one of the voltage and the capacity of the storage battery and the characteristic under each of the plurality of conditions to generate an estimation model for estimating at least one of a maximum capacity and a remaining capacity of the storage battery from one of the voltage and the capacity of the storage battery and the characteristic; wherein the estimation model is learned using a modular network Self Organizing Map, wherein a module is a neural network. ### Claim 18 A non-transitory recording medium having recorded thereon an estimation program to be implemented by a computer for causing the computer to function as: a model acquisition unit for acquiring an estimation model for estimating at least one of a maximum capacity and a remaining capacity of a storage battery from a relationship between one of a voltage and a capacity of the storage battery and a characteristic corresponding to a capacity change and a voltage change of the storage battery under a plurality of conditions during at least one of charging and discharging of the storage battery; a data acquisition unit for acquiring a time series data regarding the voltage and a current of the storage battery; a calculation unit for calculating the characteristic based on the time series data regarding the voltage and the current of the storage battery; and an estimation unit for using the estimation model to estimate at least one of the maximum capacity and the remaining capacity of the storage battery from one of the voltage and the capacity of the storage battery and the calculated characteristic; wherein the estimation model is learned using a modular network Self Organizing Map, wherein a module is a neural network.