US-12618913-B2 - Method and device with battery model optimization

US12618913B2US 12618913 B2US12618913 B2US 12618913B2US-12618913-B2

Abstract

A device with battery model optimization includes: a processor configured to perform optimization on a battery model for determining optimized parameter values of parameters of the battery model, wherein, to perform the optimization, the processor is configured to: select target parameters from among parameters of a battery model; set a current boundary condition for each of the target parameters; determine an optimized parameter value of each of the target parameters based on the set current boundary condition; set a subsequent boundary condition reduced from the current boundary condition based on the determined optimized parameter value; and determine a subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition.

Inventors

Daebong JUNG
Young Hun Sung

Assignees

SAMSUNG ELECTRONICS CO., LTD.

Dates

Publication Date: 20260505
Application Date: 20220308
Priority Date: 20210929

Claims (14)

1 . A device with battery model optimization, the device comprising: a processor configured to: perform optimization on a battery model, for determining optimized parameter values of parameters of the battery model, wherein, to perform the optimization, the processor is configured to: select target parameters from among the parameters of the battery model; set a current boundary condition for each of the target parameters; determine an optimized parameter value of each of the target parameters based on the set current boundary condition; set a subsequent boundary condition reduced from the current boundary condition based on the determined optimized parameter value; and determine a subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition; and estimate a state of a battery using the battery model having the optimized parameter values of the parameters of the battery model, wherein, for the performing optimization, the processor is further configured to perform the optimization for each of state of charge (SOC) intervals by determining the SOC intervals as corresponding to respective degrees of progress in charging or discharging the battery, wherein the processor is configured to determine whether a number of performances of the optimization reaches a preset number of performances, and until the number of performances of the optimization reaches the preset number and until an optimization loss satisfies a defined condition, iteratively perform the setting of the subsequent boundary condition reduced from the current boundary condition and the determining of the subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition, and wherein, for the setting of the subsequent boundary condition, the processor is configured to: determine a target change direction of a diffusion parameter based on a voltage error between a voltage of the battery that is estimated through the battery model and a voltage of the battery that is based on profile data of the battery; and set the subsequent boundary condition based on the determined target change direction.
2 . The device of claim 1 , wherein the processor is configured to perform the optimization for each of predefined temperature intervals.
3 . The device of claim 1 , wherein the processor is configured to select the target parameters based on a value obtained by performing differentiation one or more times on the parameters of the battery model.
4 . The device of claim 1 , wherein for the setting of the subsequent boundary condition, the processor is configured to change the current boundary condition for all the target parameters based on the optimized parameter value retrieved based on the current boundary condition, and the changed boundary condition corresponds to the subsequent boundary condition.
5 . The device of claim 1 , wherein the processor is configured to: select points associated with a diffusion characteristic of the battery from among the parameters of the battery model; and for the selecting of the target parameters, determine the target parameters based on the selected points.
6 . The device of claim 1 , wherein the processor is configured to: determine an estimated state value of the battery model based on the target parameters; determine an optimization loss based on a difference between the estimated state value and an actual state value obtained from profile data of the battery; and adjust at least one of the target parameters such that the optimization loss is reduced.
7 . The device of claim 1 , wherein the parameters of the battery model comprise a diffusion parameter dependent on an SOC level of the battery, and the diffusion parameter comprises a charge parameter associated with charging of the battery and a discharge parameter associated with discharging of the battery.
8 . A processor-implemented method with battery model optimization, the method comprising: selecting target parameters from among parameters of a battery model; performing optimization on the target parameters, wherein the performing of the optimization comprises: setting a current boundary condition for each of the target parameters; determining an optimized parameter value of each of the target parameters based on the set current boundary condition; setting a subsequent boundary condition reduced from the current boundary condition based on the determined optimized parameter value; and determining a subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition; and estimating a state of a battery using the battery model having the optimized parameter values of the parameters of the battery model, wherein the performing optimization further includes performing the optimization for each of state of charge (SOC) intervals by determining the SOC intervals as corresponding to respective degrees of progress in charging or discharging the battery, and wherein the performing optimization further includes determining whether a number of performances of the optimization reaches a preset number of performances, and until the number of performances of the optimization reaches the preset number and until an optimization loss satisfies a defined condition, iteratively performing the setting of the subsequent boundary condition reduced from the current boundary condition and the determining of the subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition, and wherein the setting of the subsequent boundary condition comprises: determining a target change direction of a diffusion parameter based on a voltage error between a voltage of the battery that is estimated through the battery model and a voltage of the battery that is based on profile data of the battery; and setting the subsequent boundary condition based on the determined target change direction.
9 . The method of claim 8 , wherein the performing of the optimization comprises performing the optimization for each of predefined temperature intervals.
10 . A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the method of claim 8 .
11 . A battery power supplier, comprising: a battery configured to supply power to an electronic device; a battery model optimizing device configured to optimize a battery model corresponding to the battery, wherein, for the optimizing of the battery model, the battery model optimizing device is configured to: select target parameters from among parameters of the battery model; set a current boundary condition for each of the target parameters; determine an optimized parameter value of each of the target parameters based on the set current boundary condition; set a subsequent boundary condition reduced from the current boundary condition based on the determined optimized parameter value; and determine a subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition; and a processor configured to estimate a state of a battery using the battery model having the optimized parameter values of the parameters of the battery model, wherein, for the performing optimization, the battery model optimizing device is further configured to perform the optimization for each of state of charge (SOC) intervals by determining the SOC intervals as corresponding to respective degrees of progress in charging or discharging the battery, and wherein the processor is configured to determine whether a number of performances of the optimization reaches a preset number of performances, and until the number of performances of the optimization reaches the preset number and until an optimization loss satisfies a defined condition, iteratively perform the setting of the subsequent boundary condition reduced from the current boundary condition and the determining of the subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition, and wherein, for the setting of the subsequent boundary condition, the battery model optimizing device is configured to: determine a target change direction of a diffusion parameter based on a voltage error between a voltage of the battery that is estimated through the battery model and a voltage of the battery that is based on profile data of the battery; and set the subsequent boundary condition based on the determined target change direction.
12 . A processor-implemented method with battery model optimization, the method comprising: setting a current boundary condition for a target parameter of a battery model; determining an optimized parameter of the target parameter to be within the boundary condition; setting a subsequent boundary condition, with a range reduced from the current boundary condition, based on a difference between a state of the battery estimated using the battery model with the optimized parameter and a predetermined state of the battery; optimizing the battery model by determining a subsequent optimized parameter of the target parameter to be within the subsequent boundary condition, further including optimizing the battery model for each of state of charge (SOC) intervals by determining the SOC intervals as corresponding to respective degrees of progress in charging or discharging the battery; and estimating a state of a battery using the battery model with optimized parameters of the battery model, wherein the optimizing of the battery model further includes determining whether a number of performances of the optimization reaches a preset count of performances, and until the number of performances of the optimization reaches the preset count and until an optimization loss satisfies a defined condition, iteratively performing the setting of the subsequent boundary condition reduced from the current boundary condition and the determining of the subsequent optimized parameter value of the target parameter based on the subsequent boundary condition, and wherein the setting of the subsequent boundary condition comprises: determining a target change direction of a diffusion parameter based on a voltage error between a voltage of the battery that is estimated through the battery model and a voltage of the battery that is based on profile data of the battery; and setting the subsequent boundary condition based on the determined target change direction.
13 . The method of claim 12 , wherein the set current boundary condition comprises a lower limit and an upper limit, and the determining of the optimized parameter comprises determining the optimized parameter to be greater than or equal to the lower limit and less than or equal to the upper limit.
14 . The method of claim 13 , wherein the setting of the subsequent boundary condition comprises, based on whether the state of the battery estimated using the battery model is greater than the predetermined state of the battery, either one of: increasing at least one of the lower limit and the higher limit, and decreasing at least one of the lower limit and the higher limit.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit under 35 USC § 119(a) of Korean Patent Application No. 10-2021-0129038, filed on Sep. 29, 2021 in the Korean Intellectual Property Office, the entire disclosure of which is incorporated herein by reference for all purposes. BACKGROUND 1. Field The following description relates to a method and device with battery model optimization. 2. Description of Related Art For optimal management of a battery, a state of the battery may be estimated using various methods. For example, a state of a battery may be estimated by integrating currents of the battery or by using a battery model (e.g., an electric circuit model or an electrochemical model). The current integration method may calculate a remaining amount of the battery by attaching a current sensor to an end of a battery cell, module, or pack and calculating a sum of charge amounts to be charged or discharged. The electric circuit model may be a circuit model including a resistor and a capacitor that represent a voltage value changing as a battery is charged or discharged, and the electrochemical model may be a model that models internal physical phenomena of the battery, for example, a battery ion concentration, a potential, and the like. SUMMARY This Summary is provided to introduce a selection of concepts in a simplified form that is further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter. In one general aspect, a device with battery model optimization includes: a processor configured to perform optimization on a battery model for determining optimized parameter values of parameters of the battery model, wherein, to perform the optimization, the processor is configured to: select target parameters from among parameters of a battery model; set a current boundary condition for each of the target parameters; determine an optimized parameter value of each of the target parameters based on the set current boundary condition; set a subsequent boundary condition reduced from the current boundary condition based on the determined optimized parameter value; and determine a subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition. The processor may be configured to perform the optimization for each of predefined different state of charge (SOC) intervals corresponding to degrees of progress in charging the battery. The processor may be configured to perform the optimization based on predefined different SOC intervals corresponding to degrees of progress in discharging the battery. For the setting of the subsequent boundary condition, the processor may be configured to: determine a target change direction of a diffusion parameter based on a voltage error between a voltage of the battery that is estimated through the battery model and a voltage of the battery that is based on profile data of the battery; and set the subsequent boundary condition based on the determined target change direction. The processor may be configured to perform the optimization for each of predefined temperature intervals. The processor may be configured to select the target parameters based on a value obtained by performing differentiation one or more times on the parameters of the battery model. The processor may be configured to, until the number of performances of the optimization reaches a set number, iteratively perform the setting of the subsequent boundary condition reduced from the current boundary condition and the determining of the subsequent optimized parameter value of each of the target parameters based on the subsequent boundary condition. For the setting of the subsequent boundary condition, the processor may be configured to change the current boundary condition for all the target parameters based on the optimized parameter value retrieved based on the current boundary condition, and the changed boundary condition may correspond to the subsequent boundary condition. The processor may be configured to: select points associated with a diffusion characteristic of the battery from among the parameters of the battery model; and for the selecting of the target parameters, determine the target parameters based on the selected points. The processor may be configured to: determine an estimated state value of the battery model based on the target parameters; determine an optimization loss based on a difference between the estimated state value and an actual state value obtained from profile data of the battery; and adjust at least one of the target parameters such that the optimization loss is reduced. The parameters of the battery model may include a diffusion parameter dependent on an SOC level of the battery, and the diffusion parameter may include a charge para