CN116736131A - Battery health state estimation method, terminal device and storage medium - Google Patents
Battery health state estimation method, terminal device and storage medium Download PDFInfo
- Publication number
- CN116736131A CN116736131A CN202310638695.5A CN202310638695A CN116736131A CN 116736131 A CN116736131 A CN 116736131A CN 202310638695 A CN202310638695 A CN 202310638695A CN 116736131 A CN116736131 A CN 116736131A
- Authority
- CN
- China
- Prior art keywords
- charging
- health
- battery
- capacity
- state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000036541 health Effects 0.000 title claims abstract description 167
- 238000000034 method Methods 0.000 title claims abstract description 82
- 238000007600 charging Methods 0.000 claims abstract description 203
- 238000012549 training Methods 0.000 claims description 58
- 230000011218 segmentation Effects 0.000 claims description 23
- 238000004590 computer program Methods 0.000 claims description 21
- 238000010277 constant-current charging Methods 0.000 claims description 14
- 238000010281 constant-current constant-voltage charging Methods 0.000 claims description 6
- 238000013507 mapping Methods 0.000 abstract description 14
- 230000003862 health status Effects 0.000 description 15
- 230000008569 process Effects 0.000 description 12
- 238000012360 testing method Methods 0.000 description 9
- 239000011159 matrix material Substances 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000010354 integration Effects 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 238000005314 correlation function Methods 0.000 description 3
- 230000007423 decrease Effects 0.000 description 3
- PXHVJJICTQNCMI-UHFFFAOYSA-N Nickel Chemical compound [Ni] PXHVJJICTQNCMI-UHFFFAOYSA-N 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000007599 discharging Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 description 1
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 1
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 1
- 239000002253 acid Substances 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- OJIJEKBXJYRIBZ-UHFFFAOYSA-N cadmium nickel Chemical compound [Ni].[Cd] OJIJEKBXJYRIBZ-UHFFFAOYSA-N 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010280 constant potential charging Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012983 electrochemical energy storage Methods 0.000 description 1
- 238000003487 electrochemical reaction Methods 0.000 description 1
- 229910052739 hydrogen Inorganic materials 0.000 description 1
- 239000001257 hydrogen Substances 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 229910052744 lithium Inorganic materials 0.000 description 1
- 229910001416 lithium ion Inorganic materials 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 229910052759 nickel Inorganic materials 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 229920000642 polymer Polymers 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
- Secondary Cells (AREA)
Abstract
The application discloses a battery state of health estimation method, terminal equipment and a storage medium, wherein the method determines a target voltage interval through battery charging condition data, and establishes a mapping relation between the charging capacity of the target voltage interval and the battery state of health through a model.
Description
Technical Field
The application belongs to the technical field of batteries, and particularly relates to a battery health state estimation method, terminal equipment and storage equipment.
Background
Compared with a primary battery, the rechargeable battery has the advantages of being more environment-friendly, repeatedly charged and discharged and the like, and is widely applied to the fields of digital products, new energy automobiles, electrochemical energy storage and the like. However, the overall performance of the rechargeable battery is gradually deteriorated due to the electrochemical reaction during the use process, such as a decrease in battery capacity, an increase in internal resistance, and a decrease in safety. The state of health (SOH) of a battery characterizes the ability of a current battery to store electrical energy relative to a new battery, typically by selecting the current capacity of the battery.
Accurate estimation of battery state of health is of great practical importance for efficient and safe use of rechargeable batteries. In the process of establishing a battery state of health estimation method, it is generally necessary to extract characteristic parameters capable of accurately feeding back the state of health of the battery.
At present, the widely studied methods for extracting health characteristic parameters mainly comprise: incremental capacity method, differential voltage curve method, and pulse discharge method. The incremental capacity curve and the differential voltage curve method both need to measure open-circuit voltage first, the test time is long, the curve peak value is influenced by the precision of test equipment, effective data are difficult to obtain in practical application, the curve data size is large, and the processing is difficult. In the application of the pulse discharge method, the discharge phase is usually discontinuous, so that the discharge curve is incomplete, which is unfavorable for extracting the characteristic parameters.
In summary, it is difficult to accurately extract the characteristic parameters by the above methods, and thus it is difficult to accurately estimate the state of health of the battery.
Disclosure of Invention
The embodiment of the application provides a battery state of health estimation method, terminal equipment and a storage medium, which are used for solving the technical problem that the battery state of health is difficult to accurately estimate in the prior art.
In a first aspect, an embodiment of the present application provides a method for training a battery state of health estimation model, including:
acquiring charging condition data of a battery when the battery is charged in a constant-current constant-voltage charging mode;
determining a target voltage interval according to the charging condition data; wherein the capacity correlation coefficient of the target voltage interval is greater than a correlation threshold; the capacity correlation coefficient is used for representing the correlation between the charging capacity corresponding to the charging voltage interval and the actual total capacity of the battery;
and constructing an input sample according to the charging capacity corresponding to the target voltage interval and the health state label corresponding to the charging capacity, and training a pre-constructed health state estimation model through the input sample.
The target voltage interval is determined through the battery charging condition data, and the mapping relation between the charging capacity of the target voltage interval and the battery health state is constructed through the model, and the health state of the battery can be accurately estimated through the model because the correlation between the charging capacity of the target voltage interval and the actual total capacity of the battery is large.
In a possible implementation manner of the first aspect, the determining the target voltage interval according to the charging condition data includes:
determining a constant-current charging voltage interval according to the charging condition data, and selecting a segmentation voltage threshold value;
dividing the constant current charging voltage interval into a plurality of charging voltage intervals according to the segmentation voltage threshold;
calculating capacity correlation coefficients of a plurality of charging voltage intervals;
and determining a charging voltage interval with a capacity correlation coefficient larger than the correlation threshold as the target voltage interval.
In a possible implementation manner of the first aspect, after the calculating the capacity correlation coefficients of the plurality of charging voltage intervals, the method further includes:
and if the charging voltage interval with the capacity correlation coefficient larger than the correlation threshold value does not exist, the segmentation voltage threshold value is reselected until the target voltage interval is determined.
In a possible implementation manner of the first aspect, in a case that a plurality of segment voltage thresholds are selected, the method further includes:
if no charging voltage interval with the capacity correlation coefficient larger than the correlation threshold exists, selecting one segmentation voltage threshold from a plurality of segmentation voltage thresholds to adjust, and keeping the other segmentation voltage thresholds unchanged until the target voltage interval is determined according to the adjusted segmentation voltage threshold.
In one possible implementation manner of the first aspect, the health status tag includes an actual health status tag; the charging condition data comprise current data of each charging moment;
the constructing an input sample according to the charging capacity corresponding to the target voltage interval and the health state label corresponding to the charging capacity includes:
determining the actual total capacity of the battery according to the current data corresponding to each charging moment;
determining the actual health state of the battery according to the actual total capacity and the nominal rated capacity of the battery, and setting the actual health state as the actual health state label;
and constructing the input sample according to the charging capacity corresponding to the target voltage interval and the actual health state label.
In one possible implementation of the first aspect, the health status tag includes an actual total capacity tag; the charging condition data comprise current data of each charging moment;
the constructing an input sample according to the charging capacity corresponding to the target voltage interval and the health state label corresponding to the charging capacity includes:
determining the actual total capacity of the battery according to the current data corresponding to each charging moment and setting the actual total capacity as the actual total capacity label;
and constructing the input sample according to the charging capacity corresponding to the target voltage interval and the actual total capacity label.
In a possible implementation manner of the first aspect, the training a pre-built health state estimation model through the input samples includes:
initializing parameters of each network layer in the health state estimation model;
in each training round, inputting the input sample into the health state estimation model for forward propagation to obtain a health state estimation result of the input sample;
determining an error according to the health state estimation result and the health state label corresponding to the input sample;
and when the error is not in the preset error range, inputting the error into the health state estimation model for back propagation so as to correct the parameters of the network layer until the error is in the preset error range or the training iteration times are reached.
In a second aspect, an embodiment of the present application provides a method for estimating a state of health of a battery, including:
acquiring the charging capacity of a battery to be estimated corresponding to any charging period;
inputting the charging capacity into a health state estimation model to obtain a battery health state estimation result; the health state estimation model is a model obtained by training by the battery health state estimation model training method provided by the first aspect.
In a third aspect, an embodiment of the present application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the computer program when executed by the processor implements the method for training a battery state of health estimation model as provided in the first aspect, or where the computer program when executed by the processor implements the method for estimating a battery state of health as provided in the second aspect.
In a fourth aspect, an embodiment of the present application provides a computer readable storage medium storing a computer program, which when executed by a processor implements the battery state of health estimation model training method as provided in the first aspect, or which when executed by a processor implements the battery state of health estimation method as provided in the second aspect.
It will be appreciated that the advantages of the second to fourth aspects may be found in the relevant description of the first aspect and are not repeated here.
Drawings
FIG. 1 is a flowchart of a method for training a battery state of health estimation model according to an embodiment of the present application;
fig. 2 is a schematic diagram of an implementation flow of step S12 in the battery state of health estimation model training method according to the embodiment of the present application;
fig. 3 is a schematic implementation flow chart of step S13 in the battery state of health estimation model training method according to the embodiment of the present application;
FIG. 4 is a graph comparing the battery state of health estimation results obtained by the battery state of health estimation model training method according to the embodiment of the present application with actual results;
FIG. 5 is a schematic diagram of model errors corresponding to different charge cycles obtained by the battery state of health estimation model training method according to an embodiment of the present application;
fig. 6 is a flowchart of a battery state of health estimation method according to an embodiment of the present application;
FIG. 7 is a block diagram of a battery state of health estimation model training apparatus according to an embodiment of the present application;
fig. 8 is a block diagram of a terminal device according to an embodiment of the present application.
Detailed Description
In order to make the technical problems, technical schemes and beneficial effects to be solved more clear, the application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
Most of the existing battery state-of-health estimation methods adopt a capacity-increasing method, a differential voltage curve method and a pulse discharge method when extracting health characteristic parameters. However, these methods have difficulty in accurately extracting the characteristic parameters, and thus in accurately estimating the battery state of health.
In order to solve the technical problem that a battery state of health estimation result is inaccurate, the embodiment of the application provides a battery state of health estimation model training method and a battery state of health estimation method. The execution subject of the method is a terminal device, including but not limited to: server, computer, smart phone and tablet computer.
The method provided by the embodiment of the application is described in detail below with reference to the accompanying drawings.
Referring to fig. 1, fig. 1 is a training method for a battery state of health estimation model according to a first embodiment of the present application, including steps S11 to S13:
s11, acquiring charging condition data when the battery is charged in a constant-current constant-voltage charging mode.
The constant-current constant-voltage charging mode is to charge with a constant current to an upper limit cut-off voltage, and then charge with the constant voltage until the current is reduced to a small multiplying power current (e.g. 0.05C) of the constant-current charging current.
By way of example, the battery may include, but is not limited to, a lithium ion battery, a nickel cadmium battery, a nickel hydrogen battery, a lithium polymer battery, a lead acid battery, or the like.
Exemplary, the charging condition data includes, but is not limited to, charging current and voltage corresponding to each charging time.
In one embodiment, the terminal device may be connected to a battery charging tester, and read the charging condition data output by the battery charging tester. Specifically, the battery charging tester is a common test device for rechargeable batteries, has a constant-current constant-voltage charging and discharging function, can realize automatic life cycle, and can record relevant test data such as current, voltage, temperature, charge quantity and the like in real time. Therefore, the battery charging tester can obtain charging condition data corresponding to a plurality of cycle charging tests of the battery.
In another embodiment, the terminal device may be connected to the charging device and the digital multimeter, and control the charging device to perform constant current and constant voltage charging tests on the battery for multiple times, and obtain charging condition data corresponding to each test by reading measurement data of the digital multimeter.
In other embodiments, the terminal device may also read the charging condition data in the historical charging process of the battery from the local memory or the cloud database.
In the above embodiment, after acquiring the charging condition data such as the charging current and the voltage value at each charging time, the terminal device may generate the charging voltage curve and the charging current curve according to the charging current and the voltage value at each charging time.
S12, determining a target voltage interval according to the charging working condition data.
Wherein the capacity correlation coefficient of the target voltage interval is greater than a correlation threshold; the capacity correlation coefficient is used for representing the correlation between the charging capacity corresponding to the charging voltage interval and the actual total capacity of the battery.
The terminal device can calculate the correlation between the charging capacity corresponding to the charging voltage interval and the actual total capacity of the battery through the correlation function. By way of example, the correlation function may employ a corrcoef function or other correlation function.
The actual total capacity of the battery can be determined by the battery working condition data, for example, according to charging current data of each charging time, an ampere-hour integration method is adopted to obtain the actual total capacity of the battery.
In one embodiment, the terminal device may extract the voltage value at each charging time after obtaining the charging condition data; the terminal equipment can randomly select any two different voltage values to pair to obtain a charging voltage interval, for example, the terminal equipment obtains voltage values U1 and U2, generates charging voltage intervals [ U1 and U2] according to the U1 and U2, obtains the charging capacity of the charging voltage interval by adopting an ampere-hour integration method according to the charging current of each charging moment in a charging stage corresponding to the charging voltage interval, and further calculates the capacity correlation coefficient of the charging voltage interval to find a target voltage interval with the capacity correlation coefficient larger than the correlation threshold.
In other embodiments, after acquiring the charging condition data, the terminal device may determine a complete charging voltage interval (or a complete charging voltage curve) corresponding to the complete charging process, select a segment voltage threshold, divide the complete charging voltage interval into a plurality of charging voltage intervals according to the segment voltage threshold, and calculate a capacity correlation coefficient of each charging voltage interval, thereby determining the target voltage interval. For example, after the terminal device obtains the complete charging voltage interval [ U3, U4], selecting the segment voltage threshold U0 from the [ U3, U4], dividing the [ U3, U4] into the charging voltage interval [ U3, U0] and the [ U0, U4] according to the U0, and calculating the capacity correlation coefficient corresponding to the [ U3, U0] and the [ U0, U4], to find the target voltage interval with the capacity correlation coefficient larger than the correlation threshold.
Wherein, the correlation threshold value can be set according to actual situations or experience. The correlation threshold may be set to 0.7, for example.
S13, constructing an input sample according to the charging capacity corresponding to the target voltage interval and the health state label corresponding to the charging capacity, and training a pre-constructed health state estimation model through the input sample.
After the terminal device obtains the target voltage intervals corresponding to the multiple cyclic charging tests through steps S11 to S12, one or more target voltage intervals may be obtained in each cyclic charging test, and a sample may be formed according to each target voltage interval and the corresponding health status label.
The health state estimation model may be a neural network or other machine learning model, which is not limited by the present application.
In one embodiment, the terminal device may obtain the mapping relationship between the charging capacity corresponding to the charging voltage interval and the battery state of health SOH estimation value by training the state of health estimation model. In this embodiment, the health status tag may be an actual battery health status tag.
Since the battery state of health can be obtained by estimating the ratio between the current total capacity of the battery and the initial nominal rated capacity of the battery. Thus, the total capacity estimate of the battery may also be used to characterize the state of health of the battery.
In another embodiment, the terminal device may obtain the mapping relationship between the charging capacity corresponding to the charging voltage interval and the estimated value of the total capacity of the battery by training the health state estimation model. In this embodiment, the health status tag may be an actual battery total capacity tag.
According to the embodiment of the application, the target voltage interval is determined through the battery charging condition data, and the mapping relation between the charging capacity of the target voltage interval and the battery health state is trained through the model, and as the correlation between the charging capacity of the target voltage interval and the actual total capacity of the battery is larger, the mapping relation between the charging capacity of the corresponding voltage interval obtained through model training and the battery health state is more accurate, so that the health state of the battery can be accurately estimated through the model.
Referring to fig. 2, fig. 2 is a schematic diagram of an implementation flow of S12 in the battery state of health estimation model training method according to the embodiment of the present application. In this embodiment, the S12 may include S121 to S125:
s121, determining a constant-current charging voltage interval according to the charging condition data, and selecting a segmentation voltage threshold.
In the cycle charging process of the rechargeable battery, along with the decline of the battery capacity, the time required for the voltage to reach the upper limit cut-off voltage in the constant-current charging stage can be gradually shortened, and the voltage abrupt change when the rechargeable battery is nearly fully charged can be gradually advanced, so that the capacity and the time corresponding to the corresponding voltage interval are reduced. Therefore, the charging condition data of the constant-current charging stage is selected for feature extraction, and the health features of the battery health state can be better reflected.
S122, dividing the constant-current charging voltage interval into a plurality of charging voltage intervals according to the segmentation voltage threshold.
In the constant current charging stage, the charging curve may be incomplete (that is, the charging condition data of the partial charging stage cannot be obtained) due to the incomplete last discharging of the battery, so that the determination of the total capacity of the battery is directly affected, and the extraction of the health feature (in the case that the total capacity of the battery is taken as the health feature) is further affected.
The segment voltage threshold value can be set according to the type and capacity of the battery. For example, for a ternary battery, the segmentation voltage threshold may be set to 4.0V, two charging voltage intervals of [3.5V,4.0V ] and [4.0V,4.20V ] are extracted, and the capacity correlation coefficients of the two charging voltage intervals are further determined to determine the target voltage interval.
S123, calculating capacity correlation coefficients of a plurality of charging voltage intervals.
In this embodiment, correlation between the charge capacity of each of the charge voltage intervals and the current actual total capacity of the battery may be calculated in MATLAB using a corrcoef function, so as to obtain a capacity correlation coefficient of each of the charge voltage intervals.
And S124, determining a charging voltage interval with the capacity correlation coefficient larger than the correlation threshold as the target voltage interval.
In this embodiment, the charging voltage interval with the capacity correlation coefficient greater than the correlation threshold value indicates that the charging capacity corresponding to the charging voltage interval has a strong correlation with the current actual total capacity of the battery, so that the charging capacity corresponding to the charging voltage interval is used as a health feature, and the state of health of the battery is estimated according to the health feature, thereby obtaining a more accurate estimation result.
And S125, if no charging voltage interval with the capacity correlation coefficient larger than the correlation threshold exists, reselecting the segmentation voltage threshold until the target voltage interval is determined.
For example, when the charging voltage intervals U1 and U2 are divided according to the initially set segment voltage threshold V1, if the charging capacities corresponding to U1 and U2 are not greater than the correlation threshold, the value of V1 is adjusted until one or more target voltage intervals are determined.
In this embodiment, the constant current charging voltage interval is divided into a plurality of charging voltage intervals, a target voltage interval with a capacity correlation coefficient greater than the correlation threshold is selected, and the charging capacity corresponding to the target voltage interval is used as a health feature, so that the estimation result obtained in the subsequent estimation of the battery health state according to the battery health feature is more accurate.
In other embodiments, in the case that a plurality of segment voltage thresholds are selected, after S123, the method may further include:
if no charging voltage interval with the capacity correlation coefficient larger than the correlation threshold exists, selecting one segmentation voltage threshold from a plurality of segmentation voltage thresholds to adjust, and keeping the other segmentation voltage thresholds unchanged until the target voltage interval is determined according to the adjusted segmentation voltage threshold.
For example, when the charging voltage intervals U1, U2 and U3 are divided according to the initially set segment voltage thresholds V1 and V2, if the charging capacities corresponding to U1, U2 and U3 are not greater than the correlation threshold, the value of V2 is kept unchanged, and the value of the segment voltage threshold V1 is adjusted for multiple times until one or more target voltage intervals are determined.
In one embodiment, the health status tag in S13 includes an actual health status tag, and correspondingly, S13 may include:
determining the actual total capacity of the battery according to the current data corresponding to each charging moment;
determining the actual health state of the battery according to the actual total capacity and the nominal rated capacity of the battery, and setting the actual health state as the actual health state label;
and constructing the input sample according to the charging capacity corresponding to the target voltage interval and the actual health state label.
In the implementation, the terminal equipment can obtain current data corresponding to each charging time, calculate the actual total capacity by adopting an ampere-hour integration method, and further calculate the ratio of the actual total capacity to the nominal rated capacity of the battery to obtain the actual health state of the battery as a health state label.
In this embodiment, a mapping relationship between the charging capacity corresponding to the charging voltage interval and the health state estimation result is established through the model, so that the health state estimation result can be directly obtained through the model.
In another embodiment, the health status tag in S13 above includes an actual total capacity tag; correspondingly, the step S13 may include:
determining the actual total capacity of the battery according to the current data corresponding to each charging moment and setting the actual total capacity as the actual total capacity label;
and constructing the input sample according to the charging capacity corresponding to the target voltage interval and the actual total capacity label.
In this embodiment, a mapping relationship between the charging capacity corresponding to the charging voltage interval and the total battery capacity estimation result is established through the model, and since the battery health state can be represented by the battery capacity, the total battery capacity estimation result output through the model can obtain the battery health state estimation result.
In one embodiment, the step S13 may include:
initializing parameters of each network layer in the health state estimation model;
in each training round, inputting the input sample into the health state estimation model for forward propagation to obtain a health state estimation result of the input sample;
determining an error according to the health state estimation result and the health state label corresponding to the input sample;
and when the error is not in the preset error range, inputting the error into the health state estimation model for back propagation so as to correct the parameters of the network layer until the error is in the preset error range or the training iteration times are reached.
In this embodiment, the health state estimation model may construct a mapping relationship between the charging capacity corresponding to the charging voltage interval and the battery health state estimation result.
Taking the neural network as an example of the health state estimation model, the health state estimation model can feed back the mapping relation to the multi-layer feedforward neural network of the network, and the neural network can directly input the charging capacity of a certain charging voltage interval to obtain a health state estimation result by training the mapping relation and storing the mapping relation in the network.
The health state estimation model comprises two processes in the training process, namely, input layer information is transmitted backwards and output layer errors are transmitted forwards. When the error is transmitted backward, the input sample is transmitted to the input layer through the hidden layer to carry out backward transmission, the mapping relation is established in the output layer after the input sample is processed, the error is calculated by the output layer, if the mean square error of the error does not meet the expectations, the input sample is transmitted to the input layer through the hidden layer to carry out the backward transmission, and the error is used as a correction weight to correct the mapping relation.
Referring to fig. 3, fig. 3 illustrates a schematic implementation flow chart of the above S13, where in this example, the above S13 may include the following steps:
s131, carrying out normalization processing on the input samples.
S132, initializing network parameters, and setting an input matrix and an output matrix of the model.
Wherein the initial value of the network parameter is set to a random non-zero value; the input matrix and the output matrix are respectively a matrix form corresponding to the input sample and a matrix form corresponding to the output result, for example, in this embodiment, the input sample is a matrix of 1*1, and the output result is a matrix of 1*1.
S133, in the ith training round, inputting an input sample into the model for forward propagation.
Wherein i is an integer greater than or equal to 1. The forward propagation process includes calculating input and output values for the input layer, calculating input and output values for each neuron in each hidden layer from the output values for the input layer, and calculating input and output values for the output layer from the output values for the hidden layer.
S134, calculating errors of the output layer and errors of the hidden layer.
The error of the output layer is determined according to the output value of the output layer and the health state label of the input sample. The error of the hidden layer is determined according to the error of the output layer and the parameter weight of the hidden layer.
S135, correcting parameters of the output layer and the hidden layer according to the errors of the output layer and the errors of the hidden layer.
S136, judging whether the error of the output layer is within a preset range, if so, ending the training process; otherwise, let training round number i=i+1, repeat the above steps S133 to S136 until the error of the output layer is within the preset error range.
In an example, the accuracy of the health state estimation model obtained by training is further verified, in the example, charging condition data corresponding to the 1 st charging cycle number and the 2 nd charging cycle number … th 170 th charging cycle number are selected, the charging capacity of a target voltage interval is selected as a health feature for each charging cycle number, the health feature of each charging cycle number is input into the health state estimation model obtained by training in the steps S11 to S13, health state estimation results corresponding to each charging cycle number are output respectively, and the health state estimation results corresponding to each charging cycle number are compared with actual battery health state results, so that a graph of the estimated battery health state results and the actual battery health state results shown in fig. 4 is obtained. As can be seen from fig. 4, the two curves substantially coincide, which illustrates that the accuracy of the health state estimation model of this embodiment is higher. Referring to fig. 5, fig. 5 is a graph illustrating model errors corresponding to different charge cycles, and it can be seen from fig. 5 that the model average error tends to zero.
Referring to fig. 6, fig. 6 is a flowchart of a battery state of health estimation method according to an embodiment of the present application, where the embodiment includes S61 to S62:
s61, acquiring the charging capacity of the battery to be estimated corresponding to any charging period.
For example, the terminal device may select charging periods t1 to t2, where charging periods t1 to t2 correspond to charging voltage intervals [ U1, U2]. If the charging current of the battery to be estimated is I during the charging period t 1-t 2, the terminal equipment can calculate and obtain the charging capacity corresponding to the charging period t1, t2 by an ampere-hour integration method.
S62, inputting the charging capacity into a health state estimation model to obtain a battery health state estimation result; the health state estimation model is a model obtained by training by adopting the health state estimation model training method provided by any one embodiment.
According to the embodiment of the application, the charging capacity of any charging period is used as the health characteristic and is input into the health state estimation model, so that the health state estimation result of the battery can be directly obtained.
Referring to fig. 7, fig. 7 is a block diagram of a health state estimation model training apparatus according to an embodiment of the present application. The apparatus of this embodiment comprises:
the acquiring module 71 is configured to acquire charging condition data when the battery is charged in a constant-current constant-voltage charging manner.
A determining module 72, configured to determine a target voltage interval according to the charging condition data; wherein the capacity correlation coefficient of the target voltage interval is greater than a correlation threshold; the capacity correlation coefficient is used for representing the correlation between the charging capacity corresponding to the charging voltage interval and the actual total capacity of the battery.
The training module 73 is configured to construct an input sample according to the charging capacity corresponding to the target voltage interval and the health status label corresponding to the charging capacity, and train a pre-constructed health status estimation model according to the input sample.
In one possible implementation, the determining module 72 may include a selecting unit, a dividing unit, a calculating unit, and a determining unit.
The selecting unit is used for determining a constant-current charging voltage interval according to the charging condition data and selecting a segmented voltage threshold.
The dividing unit is used for dividing the constant current charging voltage interval into a plurality of charging voltage intervals according to the segmentation voltage threshold value.
The calculation unit is used for calculating capacity correlation coefficients of a plurality of charging voltage intervals.
The determination unit is configured to determine a charging voltage section in which a capacity correlation coefficient is greater than the correlation threshold as the target voltage section.
In a possible implementation manner, the determining unit is further configured to reselect the segment voltage threshold until the target voltage interval is determined when there is no charging voltage interval in which the capacity correlation coefficient is greater than the correlation threshold.
In a possible implementation manner, the selecting unit is further configured to select a plurality of segment voltage thresholds, and the determining unit is further configured to select, when there is no charging voltage interval with a capacity correlation coefficient greater than the correlation threshold, one segment voltage threshold from the plurality of segment voltage thresholds to adjust, and keep other segment voltage thresholds unchanged until the target voltage interval is determined according to the adjusted segment voltage threshold.
In one possible implementation, the health status tag includes an actual health status tag; the charging condition data comprise current data of each charging moment; correspondingly, the training module 73 may include a sample construction unit and a training unit.
The sample construction unit is used for determining the actual total capacity of the battery according to the current data corresponding to each charging moment; determining the actual health state of the battery according to the actual total capacity and the nominal rated capacity of the battery, and setting the actual health state as the actual health state label; and constructing the input sample according to the charging capacity corresponding to the target voltage interval and the actual health state label.
The training unit is used for training a pre-constructed health state estimation model through the input samples.
In one possible implementation, the health status tag includes an actual total capacity tag; the charging condition data comprise current data of each charging moment; correspondingly, the sample construction unit is further used for determining the actual total capacity of the battery according to the current data corresponding to each charging moment and setting the actual total capacity as the actual total capacity label; and constructing the input sample according to the charging capacity corresponding to the target voltage interval and the actual total capacity label.
In one possible implementation manner, the training unit is specifically configured to:
initializing parameters of each network layer in the health state estimation model;
in each training round, inputting the input sample into the health state estimation model for forward propagation to obtain a health state estimation result of the input sample;
determining an error according to the health state estimation result and the health state label corresponding to the input sample;
and when the error is not in the preset error range, inputting the error into the health state estimation model for back propagation so as to correct the parameters of the network layer until the error is in the preset error range or the training iteration times are reached.
It should be noted that, because the content of information interaction and execution process between the above devices is based on the same concept as the method embodiment of the present application, specific functions and technical effects thereof may be referred to in the method embodiment section, and will not be described herein.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, the specific names of the functional units and modules are only for distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.
Referring to fig. 8, fig. 8 is a block diagram of a terminal device according to an embodiment of the present application. In this embodiment, the terminal device 8 includes a memory 81, a processor 80, and a computer program 82 stored in the memory 81 and executable on the processor 80, where the processor 80 implements the battery state of health estimation model training method or the battery state of health estimation method according to any of the embodiments described above when executing the computer program 82.
The terminal device 8 may be a computing device such as a desktop computer, a notebook computer, a palm top computer, and a cloud server, and may include, but is not limited to, a processor 80 and a memory 81. It will be appreciated by those skilled in the art that fig. 8 is merely an example of the terminal device 8 and is not limiting of the terminal device 8, and may include more or fewer components than shown, or may combine certain components, or different components, such as may also include input-output devices, network access devices, etc.
The processor 80 may be a central processing unit (Central Processing Unit, CPU), the processor 80 may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 81 may in some embodiments be an internal storage unit of the terminal device 8, such as a hard disk or a memory of the terminal device 8. The memory 81 may in other embodiments also be an external storage device of the terminal device 8, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the terminal device 8. Further, the memory 81 may also include both an internal storage unit and an external storage device of the terminal device 8. The memory 81 is used for storing an operating system, application programs, boot loader (BootLoader), data, other programs etc., such as program codes of the computer program etc. The memory 81 may also be used to temporarily store data that has been output or is to be output.
Accordingly, an embodiment of the present application further provides a computer readable storage medium storing a computer program, which when executed by a processor implements the battery state of health estimation model training method or the battery state of health estimation method of any of the embodiments described above.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present application may implement all or part of the flow of the method of the above embodiments, and may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when the computer program is executed by a processor, the computer program may implement the steps of each of the method embodiments described above. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include at least: any entity or device capable of carrying computer program code to a photographing device/terminal apparatus, recording medium, computer Memory, read-Only Memory (ROM), random access Memory (RAM, randomAccess Memory), electrical carrier signal, telecommunications signal, and software distribution medium.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application.
Claims (10)
1. A battery state of health estimation model training method, comprising:
acquiring charging condition data of a battery when the battery is charged in a constant-current constant-voltage charging mode;
determining a target voltage interval according to the charging condition data; wherein the capacity correlation coefficient of the target voltage interval is greater than a correlation threshold; the capacity correlation coefficient is used for representing the correlation between the charging capacity corresponding to the charging voltage interval and the actual total capacity of the battery;
and constructing an input sample according to the charging capacity corresponding to the target voltage interval and the health state label corresponding to the charging capacity, and training a pre-constructed health state estimation model through the input sample.
2. The battery state of health estimation model training method of claim 1, wherein said determining a target voltage interval from said charging regime data comprises:
determining a constant-current charging voltage interval according to the charging condition data, and selecting a segmentation voltage threshold value;
dividing the constant current charging voltage interval into a plurality of charging voltage intervals according to the segmentation voltage threshold;
calculating capacity correlation coefficients of a plurality of charging voltage intervals;
and determining a charging voltage interval with a capacity correlation coefficient larger than the correlation threshold as the target voltage interval.
3. The battery state of health estimation model training method according to claim 2, further comprising, after said calculating capacity correlation coefficients for a plurality of said charging voltage intervals:
and if the charging voltage interval with the capacity correlation coefficient larger than the correlation threshold value does not exist, the segmentation voltage threshold value is reselected until the target voltage interval is determined.
4. The battery state of health estimation model training method of claim 2, wherein in the event that a plurality of said segment voltage thresholds are selected, said method further comprises:
if no charging voltage interval with the capacity correlation coefficient larger than the correlation threshold exists, selecting one segmentation voltage threshold from a plurality of segmentation voltage thresholds to adjust, and keeping the other segmentation voltage thresholds unchanged until the target voltage interval is determined according to the adjusted segmentation voltage threshold.
5. The battery state of health estimation model training method of claim 1, wherein said state of health label comprises an actual state of health label; the charging condition data comprise current data of each charging moment;
the constructing an input sample according to the charging capacity corresponding to the target voltage interval and the health state label corresponding to the charging capacity includes:
determining the actual total capacity of the battery according to the current data corresponding to each charging moment;
determining the actual health state of the battery according to the actual total capacity and the nominal rated capacity of the battery, and setting the actual health state as the actual health state label;
and constructing the input sample according to the charging capacity corresponding to the target voltage interval and the actual health state label.
6. The battery state of health estimation model training method of claim 1, wherein said state of health label comprises an actual total capacity label; the charging condition data comprise current data of each charging moment;
the constructing an input sample according to the charging capacity corresponding to the target voltage interval and the health state label corresponding to the charging capacity includes:
determining the actual total capacity of the battery according to the current data corresponding to each charging moment and setting the actual total capacity as the actual total capacity label;
and constructing the input sample according to the charging capacity corresponding to the target voltage interval and the actual total capacity label.
7. The battery state of health estimation model training method of claim 1, wherein said training a pre-built state of health estimation model from said input samples comprises:
initializing parameters of each network layer in the health state estimation model;
in each training round, inputting the input sample into the health state estimation model for forward propagation to obtain a health state estimation result of the input sample;
determining an error according to the health state estimation result and the health state label corresponding to the input sample;
and when the error is not in the preset error range, inputting the error into the health state estimation model for back propagation so as to correct the parameters of the network layer until the error is in the preset error range or the training iteration times are reached.
8. A battery state of health estimation method, comprising:
acquiring the charging capacity of a battery to be estimated corresponding to any charging period;
inputting the charging capacity into a health state estimation model to obtain a battery health state estimation result; the health state estimation model is a model trained by the health state estimation model training method according to any one of claims 1 to 7.
9. Terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the computer program when executed by the processor implements the battery state of health estimation model training method according to any of claims 1 to 7 or the computer program when executed by the processor implements the battery state of health estimation method according to claim 8.
10. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the battery state of health estimation model training method according to any one of claims 1 to 7, or wherein the computer program when executed by a processor implements the battery state of health estimation method according to claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310638695.5A CN116736131A (en) | 2023-05-31 | 2023-05-31 | Battery health state estimation method, terminal device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310638695.5A CN116736131A (en) | 2023-05-31 | 2023-05-31 | Battery health state estimation method, terminal device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116736131A true CN116736131A (en) | 2023-09-12 |
Family
ID=87902139
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310638695.5A Pending CN116736131A (en) | 2023-05-31 | 2023-05-31 | Battery health state estimation method, terminal device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116736131A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117150275A (en) * | 2023-11-01 | 2023-12-01 | 宁德时代新能源科技股份有限公司 | Machine learning model construction method, battery health degree prediction method and device |
-
2023
- 2023-05-31 CN CN202310638695.5A patent/CN116736131A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117150275A (en) * | 2023-11-01 | 2023-12-01 | 宁德时代新能源科技股份有限公司 | Machine learning model construction method, battery health degree prediction method and device |
| CN117150275B (en) * | 2023-11-01 | 2024-04-09 | 宁德时代新能源科技股份有限公司 | Machine learning model construction method, battery health degree prediction method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108957337B (en) | Method and device for determining state of health of battery, storage medium and electronic equipment | |
| US10921383B2 (en) | Battery diagnostic system for estimating capacity degradation of batteries | |
| US11156668B2 (en) | Method for iteratively identifying parameters of equivalent circuit model of battery | |
| CN113219343A (en) | Lithium battery health state prediction method, system, equipment and medium based on elastic network | |
| CN114371409B (en) | Training method of battery state prediction model, battery state prediction method and device | |
| US20210173012A1 (en) | Method and system for estimation of open circuit voltage of a battery cell | |
| CN112986842B (en) | Method, device and equipment for estimating state of charge of battery | |
| CN109991545B (en) | Battery pack electric quantity detection method and device and terminal equipment | |
| CN114705990B (en) | Method and system for estimating state of charge of battery cluster, electronic device and storage medium | |
| CN110888065B (en) | Battery pack state of charge correction method and device | |
| CN116736131A (en) | Battery health state estimation method, terminal device and storage medium | |
| CN114609530A (en) | Method, device, equipment and medium for correcting battery state of charge | |
| CN116466236A (en) | Battery remaining life prediction method, device, equipment and readable storage medium | |
| CN116298993A (en) | Method and device for identifying abnormal internal resistance of battery cell and terminal equipment | |
| CN104076284A (en) | Battery state of charge (SOC) tracking method and device | |
| CN114744723A (en) | Method and device for adjusting charging request current and electronic equipment | |
| CN115113049A (en) | Method for determining initial value of battery SOC and related device | |
| CN112748348B (en) | Battery low-temperature performance distribution level detection method and system and storage medium | |
| CN117129874A (en) | Parameter identification method and device for battery equivalent circuit model | |
| WO2023229921A1 (en) | Apparatus and method for battery soc estimation | |
| US20220413057A1 (en) | Battery diagnosing apparatus and method | |
| CN113900028B (en) | Battery health state estimation method and system considering initial charge state and charge-discharge path | |
| CN115993554A (en) | Method and device for evaluating health degree of underground coal mine power box battery | |
| CN116256636A (en) | Method and device for identifying parameters of battery equivalent circuit model and readable storage medium | |
| CN117471328B (en) | Method, system and terminal equipment for determining capacity of lead-acid battery |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |