CN116224074A - Soft package lithium ion battery state of charge estimation method, device and storage medium - Google Patents
Soft package lithium ion battery state of charge estimation method, device and storage medium Download PDFInfo
- Publication number
- CN116224074A CN116224074A CN202310184114.5A CN202310184114A CN116224074A CN 116224074 A CN116224074 A CN 116224074A CN 202310184114 A CN202310184114 A CN 202310184114A CN 116224074 A CN116224074 A CN 116224074A
- Authority
- CN
- China
- Prior art keywords
- charge
- battery
- state
- stress
- deep learning
- 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
Images
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]
- G01R31/382—Arrangements for monitoring battery or accumulator variables, e.g. SoC
- G01R31/3842—Arrangements for monitoring battery or accumulator variables, e.g. SoC combining voltage and current measurements
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
Abstract
The invention relates to a soft package lithium ion battery state of charge estimation method, a device and a storage medium based on dynamic stress and deep learning, wherein the method comprises the following steps: establishing a battery charge-discharge data set containing dynamic stress signals, wherein the battery charge-discharge data set comprises current, terminal voltage, dynamic stress data and a reference state of charge; offline training is based on dynamic stress and deep learning soft package lithium ion battery state of charge estimation model: according to the input form required by the deep learning network, recombining the time sequence characteristics of current, terminal voltage and dynamic stress in a battery charge and discharge data set into a three-dimensional tensor sample set, constructing a deep learning network structure of a battery state-of-charge estimation model, and offline training the battery state-of-charge estimation model based on a reference state-of-charge and the sample set; on-line battery state of charge estimation. Compared with the prior art, the method has the advantages of strong adaptability to different working conditions, good estimation performance when the data length is limited, strong robustness to measurement noise interference and the like.
Description
Technical Field
The invention relates to the field of energy storage batteries of electric automobiles, in particular to a soft package lithium ion battery state of charge estimation method, device and storage medium based on dynamic stress and deep learning.
Background
At present, low-carbon prospects promote the development of new energy automobiles, and one important route is a pure electric automobile. The power battery is responsible for energy storage and power output. The lithium ion battery has the advantages of high energy/power density, long service life, environmental friendliness and the like, and becomes a research hotspot for whole vehicle manufacturers, universities and scientific research institutions. The battery state of charge is defined as the ratio of the remaining power to the maximum available capacity, and is an important state quantity for determining the remaining driving range, the charge/discharge power, and the safe operating range, for indicating the remaining power of the battery.
Conventional state of charge estimation methods include ampere-hour integration, model-based methods, and data-driven methods. The ampere-hour integration method is defined based on the state of charge, is an open-loop estimation and is greatly affected by initial errors and measurement noise. The method based on the model estimates the state of charge of the battery by establishing a battery model and constructing an adaptive filter, and the estimation accuracy, the adaptability and the anti-interference capability on measurement noise are greatly influenced by the model and an observer. The data driven method implements state of charge estimation based on a relationship between battery measurable information and internal states.
The general idea of the data driving method is to measure the current, voltage, temperature, etc. of the battery when the battery is charged and discharged, directly take these data as input, build a data driving model and train to obtain proper model parameters, and estimate the state of charge of the battery by using the trained model. Therefore, there are two main issues to be considered in improving the estimation performance of the data driving method: firstly, how to select and build a data driving model with strong adaptability, and secondly, how to acquire high-quality battery data.
Data-driven models currently in common use for state-of-charge estimation are largely divided into traditional machine learning models and neural network models. Traditional machine learning methods include gaussian process regression, support vector machines, and the like; neural network models include convolutional neural networks, recurrent neural networks, and the like. Compared with the traditional machine learning technology, the deep learning technology can process a relatively larger sliding window length, and the deep learning model for processing the time sequence comprises a long-period memory neural network and a gating cyclic neural network.
Battery sensing technology and multidimensional data are key to data-driven battery state estimation, and a conventional state-of-charge estimation method generally adopts electric signals, thermal signals and the like to realize state-of-charge estimation. State of charge estimation based on data-driven methods currently in common use typically takes as model input measurements of battery current, voltage, and temperature within a sliding window. With the development of advanced battery sensing technology, it is hopeful to use more sensing signals to realize battery state estimation, such as battery stress signals. During charge and discharge of the battery, intercalation and deintercalation of lithium ions from the electrode active material may cause structural changes in the electrode material, thereby causing changes in the volume of the battery, resulting in changes in stress.
Combining advanced sensing techniques and machine learning techniques promises to achieve finer battery state estimation. The problems of the current method are mainly that: firstly, the current data driving model does not consider the time sequence characteristics of battery measurement data and can not effectively utilize historical information; secondly, the stress information of the battery cannot be effectively utilized in the current state of charge estimation, and accurate, effective and reliable battery SOC estimation cannot be realized.
Disclosure of Invention
The invention aims to provide a soft package lithium ion battery state of charge estimation method, a device and a storage medium based on dynamic stress and deep learning, which consider the time sequence characteristics of battery measurement data, effectively utilize the stress information of a battery and improve the reliability of battery state of charge estimation.
The aim of the invention can be achieved by the following technical scheme:
a soft package lithium ion battery state of charge estimation method based on dynamic stress and deep learning comprises the following steps:
step 1) establishing a battery charge-discharge data set containing dynamic stress signals, wherein the battery charge-discharge data set comprises current, terminal voltage, dynamic stress data and a reference state of charge;
step 2) off-line training of a soft package lithium ion battery state of charge estimation model based on dynamic stress and deep learning;
step 21), recombining time series characteristics of current, terminal voltage and dynamic stress in a battery charge and discharge data set into a three-dimensional tensor sample set according to an input form required by a deep learning network;
step 22), constructing a deep learning network structure of a battery state-of-charge estimation model, and offline training the battery state-of-charge estimation model based on a reference state-of-charge and a sample set;
step 3) online battery state of charge estimation:
step 31) acquiring battery current, terminal voltage and dynamic stress signals on line in the actual charging and discharging process of the battery;
step 32) carrying out on-line estimation on the state of charge of the power battery according to the trained battery state of charge estimation model.
Said step 1) comprises the steps of:
step 11), constructing a soft package battery stress measurement experimental device by adopting a constraint fixture and a strain sensor;
step 12) designing a battery charge and discharge experiment according to the recommended environmental temperature and charge and discharge current of the battery, testing according to set working conditions, recording the current, terminal voltage and dynamic stress of the battery in the testing process, calculating the reference state of charge of the battery, and constructing a battery charge and discharge data set.
In the step 11), plane stress measurement is performed in a laboratory environment through a constraint fixture and a stress sensor, and the stress sensor is arranged between two soft package batteries in vehicle-mounted application to measure plane stress.
In the step 12), the dynamic stress S recorded in the test process d Is the stress related to the running state of the battery, which is the total stress S t Subtracting the initial static stress S s The following values:
S d ＝S t -S s
wherein the total stress is an instantaneous stress measured directly on the battery surface by the pressure sensor during charge/discharge, and the static stress is a stress obtained after the battery reaches an equilibrium state for a sufficient rest time;
the battery reference state of charge is calculated by definition:
wherein SOC is k And I k The battery state of charge and current at time k, respectively, Δt being the sampling time, C m Is the battery capacity.
In the step 12), the set working conditions include: long-term charge-discharge working condition, short-term charge-discharge working condition, pulse charge-discharge working condition and dynamic driving working condition.
In the step 21), the charge and discharge data of the battery at the time k is x k ＝[I k ,V k ,F k ]Wherein I k Is the current at time k, V k Is the terminal voltage at time k, F k Is the dynamic stress at time k;
a sliding window with the length of n is adopted to obtain the input X for training the deep learning network k ＝[x k-n+1 ,…,x k-1 ,x k ]The data length is recorded as N data The sample size obtained is:
N samples ＝N data -n+1
reorganizing the data into a three-dimensional tensor form: [ N ] samples ,n,N features ]Wherein N is features Representing the number of features.
In the step 22), the deep learning network includes an input layer, two LSTM hidden layers, a fully connected layer, and an output layer, where the LSTM network activation process is as follows:
wherein f t Forgetting door, i t Is an input door o t Is an output door which is provided with a plurality of output doors,
is a candidate state for the update process, W x Is the weight, b x Is the bias, σ (·) is the sigmoid activation function, and tanh (·) is the hyperbolic tangent activation function.
In the step 32), the method for estimating the state of charge of the battery is as follows:
SOC * ＝LSTM(x * )
wherein LSTM is a trained battery state of charge estimation model, x * The input vector is composed of the battery current, terminal voltage and dynamic stress obtained on line.
The utility model provides a fill stake plug and play function testing arrangement, includes memory, treater and the program that stores in the memory, the method is realized to the treater when executing the program.
A storage medium having stored thereon a program which when executed performs a method as described above.
Compared with the prior art, the invention has the following beneficial effects:
according to the battery state of charge estimation method disclosed by the invention, dynamic stress signals are additionally considered outside current and voltage, so that the volume change caused by lithium ion intercalation and deintercalation in the battery charging and discharging process can be reflected, and the on-line estimation of the battery state of charge is realized by combining a deep learning network, so that the model has strong adaptability to different working conditions, good estimation performance under the condition of limited data length, and strong robustness to battery measurement noise interference.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of the relationship between cell stress and state of charge;
FIG. 3 is a graph showing the state of charge estimation of a soft-pack lithium ion battery based on dynamic stress and deep learning;
fig. 4 is a diagram of a state of charge estimation error estimated based on the method of the present invention.
Detailed Description
The invention will now be described in detail with reference to the drawings and specific examples. The present embodiment is implemented on the premise of the technical scheme of the present invention, and a detailed implementation manner and a specific operation process are given, but the protection scope of the present invention is not limited to the following examples.
The embodiment provides a soft package lithium ion battery state of charge estimation method based on dynamic stress and deep learning, as shown in fig. 1, comprising the following steps:
step 1) a battery charge and discharge data set containing dynamic stress signals is established, wherein the battery charge and discharge data set comprises current, terminal voltage, dynamic stress data and reference charge state.
And 11) constructing a soft package battery stress measurement experimental device by adopting a constraint fixture and a strain sensor.
In addition to measuring voltage, current and temperature during operation of the battery, it is necessary to measure the planar stress of the battery. Plane stress measurement is carried out through constraint clamps and stress sensors in a laboratory environment, and plane stress is measured by arranging the stress sensors between two soft package batteries in vehicle-mounted application.
The battery used in the embodiment is a soft-package lithium ion battery, and the anode material and the cathode material are respectively lithium manganate and graphite. The nominal capacity of the battery is 8Ah, the nominal voltage is 3.7V, the charge cut-off voltage is 4.2V, and the discharge cut-off voltage is 2.8V.
Step 12) designing a battery charge and discharge experiment according to the recommended environmental temperature and charge and discharge current of the battery, testing according to set working conditions, recording the current, terminal voltage and dynamic stress of the battery in the testing process, calculating the reference state of charge of the battery, and constructing a battery charge and discharge data set.
Lithium ions are intercalated and deintercalated between the two electrodes during the charge and discharge of the battery, so that the volume of the battery is changed, and stress is generated. Dynamic stress S recorded during test d Is the stress related to the running state (such as current, charge state and the like) of the battery, and is the total stress S t Subtracting the initial static stress S s The following values:
S d ＝S t -S s
where the total stress is the instantaneous stress measured directly on the cell surface by the pressure sensor during charge/discharge and the static stress is the stress obtained after a sufficient rest time (typically 2 hours) to bring the cell to equilibrium.
In this step, the set working conditions should cover the actual working scenario of the battery as far as possible, and in this embodiment, the method includes: long-term charge-discharge working condition, short-term charge-discharge working condition, pulse charge-discharge working condition, dynamic driving working condition and the like, so as to improve the adaptability of the estimation model.
Four typical conditions are designed in this embodiment to simulate the actual charge and discharge of the battery.
Working condition one: constant current working condition simulates long-term charge and discharge of the battery. Constant current charge and discharge at a current magnification of 1C were designated as Cy1 and Cy2.
Working condition II: short-term working conditions simulate short-term charge and discharge of the battery. The cells were charged to a set state of charge at different current rates (e.g., 0.25C, 0.5C, 1.5C, and 2C) and then discharged to a 0% state of charge at a 1C current, denoted Cy3-Cy6, respectively. The set states of charge were 20%, 40%, 60% and 80%, respectively.
And (3) working condition III: and the pulse working condition simulates the power characteristic of the battery. The operating mode includes two sub-operating modes: the battery is charged to a set charge state by using different current multiplying powers (0.25C, 0.5C, 1.5C and 2C) under the first sub-working condition, and then the battery is kept stand for 2 hours and is marked as Cy7-Cy10; the second working condition is that the first working condition is charged to a set charge state by a constant current of 1C, then 60 seconds of discharging-charging pulses with different current multiplying powers (0.25C, 0.5C, 1.5C and 2C) are loaded, and the second working condition is kept stand for 1 hour between two adjacent charge state points and is respectively marked as Cy11-Cy14.
And (4) working condition four: and simulating actual running of the electric automobile under the dynamic working condition. Two common driving conditions were used, new European Driving Cycle (NEDC) and urban road cycle (UDDS), denoted Cy15 and Cy16, respectively.
The measured cell strain is the strain value after subtracting the initial pressure. During the experiment, the fixture and test cell were placed in an incubator to ensure environmental consistency, with the ambient temperature set at 25 ℃.
Calculating a battery reference state of charge by current integration according to definition, the reference state of charge being output of offline trainingAnd (5) a label. The state of charge of the battery is defined as the ratio of the remaining usable electric quantity of the battery to the current capacity of the battery, and the SOC at the previous moment can be used k-1 Adding the current time-varying electric quantity I k Δt (discharge time I) k Negative, at charging I k Positive) and capacity C m In the present embodiment, the sampling time Δt=1s, the battery capacity C m =8ah. The battery reference state of charge is then calculated as:
as shown in fig. 2, the relationship between battery stress and state of charge is less affected by battery operating current. The relationship between cell stress/strain and state of charge is quantitatively evaluated by Pearson and Spearman correlation coefficients, which indicate a linear correlation of dynamic stress and state of charge, and Spearman correlation coefficients, which indicate a monotonic relationship of dynamic stress and state of charge. The two correlation coefficients between the dynamic stress and the state of charge of the battery exceed 0.97 and 0.985 respectively, which indicate the strong correlation between the dynamic stress signal and the state of charge of the battery, and are suitable for being used as the input of a state of charge estimation model.
Step 2) offline training a soft package lithium ion battery state of charge estimation model based on dynamic stress and deep learning.
Step 21) recombining the time series characteristics of current, terminal voltage and dynamic stress in the battery charge and discharge data set into a three-dimensional tensor sample set according to the input form required by the deep learning network.
The charge and discharge data of the battery at the moment k is x k ＝[I k ,V k ,F k ]Wherein I k Is the current at time k, V k Is the terminal voltage at time k, F k Is the dynamic stress at time k.
Using a sliding window of length n, i.e., using battery sense data at times k-n+ 1 through k at time k, an input X for training the deep learning network is obtained k ＝[x k-n+1 ,…,x k-1 ,x k ]The data length is recorded as N data The sample size obtained is:
N samples ＝N data -n+ 1
reorganizing the data into a three-dimensional tensor form: [ N ] samples ,n,N features ]Wherein N is features The number of features is denoted as 3 in this embodiment, which is current, terminal voltage and dynamic stress, respectively.
The data dividing matrix for training the state of charge estimation model and for online testing in this embodiment is shown in table 1:
table 1 data set partitioning matrix
Step 22), constructing a deep learning network structure of a battery state-of-charge estimation model, and offline training the battery state-of-charge estimation model based on the reference state-of-charge and the sample set.
In this embodiment, the deep learning network includes a sequence input layer, two LSTM hidden layers with 0.05Dropout, a fully connected layer and a regression layer. Wherein each layer of LSTM selects 100 hidden units. The sequence input layer is responsible for inputting sequence data into the constructed network, the fully connected layer multiplies the output of the LSTM by a weight matrix and adds bias, and the regression layer is used to perform regression tasks.
The hidden layer is composed of a plurality of repeated LSTM cells, each LSTM cell transferring two states to the next cell, namely cell state (c t ) And hidden state (h) t )。x t And y t Respectively corresponding model input and output sequences. The cell state contains the information learned from a previous time step, the hidden state also being referred to as the output state. At each time step, the LSTM layer adds or discards valid information to or from invalid information from previous cell states, including forgetting, updating, and outputting processes of cell and hidden states. These operations are controlled by three different gates, including input gate i t Forgetting door f t And an output gate o t . Forgetting process determines that should be madeWhich information is discarded or retained with a forget gate. The level of process control unit state update is updated and information is added to the unit state using the input gates. Finally, the output gate and cell state are used to determine the next hidden state.
Representing candidate states of the update procedure, W x Representing weights, b x Is bias, σ (·) is sigmoid activation function, tanh (·) is hyperbolic tangent activation function, key activation operations are as follows:
training data used by the LSTM model are a reference state of charge and a three-dimensional tensor sample set. Wherein the label at time k is SOC k Input is X k 。
LSTM model training uses an Adam optimizer, the initial learning rate is set to 0.001, and the fall-off factor is 0.2 for every 100 epoch learning rate segments; gradient attenuation factor beta 1 Is 0.9, square gradient decay factor beta 2 0.999. Because the training process of the deep learning model has certain randomness, each model is trained three times, and the final state of charge estimation result takes the average value of three times.
Step 3) on-line battery state of charge estimation
Step 31) acquiring battery current, terminal voltage and dynamic stress signals on line in the actual charging and discharging process of the battery.
When the vehicle-mounted application is carried out, the battery management system collects the current, terminal voltage and dynamic stress of the battery in real time and takes the current, terminal voltage and dynamic stress as model input. It should be noted that here, the cell current, terminal voltage and dynamic stress data measured on line are also recombined into a three-dimensional tensor form according to step 21).
Step 32) carrying out on-line estimation of the state of charge of the power battery according to the trained battery state of charge estimation model:
SOC * ＝LSTM(x * )
wherein LSTM is a trained battery state of charge estimation model, x * The input vector is composed of the battery current, terminal voltage and dynamic stress obtained on line.
As shown in fig. 3, the state of charge estimation result of Cy14 shows that the state of charge estimation method of the battery based on mechanical signals and deep learning can well estimate the state of charge of the power battery in three scenes. As shown in FIG. 4, the overall RMSE and MAE for each cycle state of charge estimate was 1.88% and 1.35%, respectively, with the maximum of RMSE below 5% and the maximum of MAE below 4% for these cycles, indicating good estimation accuracy. As shown in table 1, the fourth working condition in the scenario 1 and the scenario 2 does not participate in training, but small RMSE and MAE are obtained during online estimation, which indicates that the method has adaptability to different working conditions.
In addition, due to the addition of the stress data, compared with the situation that only current and voltage information is applied, the estimation accuracy of the state of charge of the battery is improved by 0.24%, the sensitivity to the sliding window length is low, and good estimation performance can be obtained under the condition that the data length is limited.
In step 3), after 0.5% of measurement noise is added to the three measurement values, the state of charge estimation error is only increased by 5.09% compared with the noiseless data, so that the good robustness of the invention is embodied.
In summary, an embodiment of the present invention is feasible, and the error between the estimation result and the actual state of charge data is smaller, so that the present invention has adaptability to different working conditions, supports the estimation of small data length, and has good robustness to measurement noise.
The above functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The foregoing describes in detail preferred embodiments of the present invention. It should be understood that numerous modifications and variations can be made in accordance with the concepts of the invention by one of ordinary skill in the art without undue burden. Therefore, all technical solutions which can be obtained by logic analysis, reasoning or limited experiments based on the prior art by a person skilled in the art according to the inventive concept shall be within the scope of protection defined by the claims.
Claims (10)
1. The soft package lithium ion battery state of charge estimation method based on dynamic stress and deep learning is characterized by comprising the following steps:
step 1) establishing a battery charge-discharge data set containing dynamic stress signals, wherein the battery charge-discharge data set comprises current, terminal voltage, dynamic stress data and a reference state of charge;
step 2) off-line training of a soft package lithium ion battery state of charge estimation model based on dynamic stress and deep learning;
step 21), recombining time series characteristics of current, terminal voltage and dynamic stress in a battery charge and discharge data set into a three-dimensional tensor sample set according to an input form required by a deep learning network;
step 22), constructing a deep learning network structure of a battery state-of-charge estimation model, and offline training the battery state-of-charge estimation model based on a reference state-of-charge and a sample set;
step 3) online battery state of charge estimation:
step 31) acquiring battery current, terminal voltage and dynamic stress signals on line in the actual charging and discharging process of the battery;
step 32) carrying out on-line estimation on the state of charge of the power battery according to the trained battery state of charge estimation model.
2. The method for estimating the state of charge of the soft package lithium ion battery based on dynamic stress and deep learning according to claim 1, wherein the step 1) comprises the following steps:
step 11), constructing a soft package battery stress measurement experimental device by adopting a constraint fixture and a strain sensor;
step 12) designing a battery charge and discharge experiment according to the recommended environmental temperature and charge and discharge current of the battery, testing according to set working conditions, recording the current, terminal voltage and dynamic stress of the battery in the testing process, calculating the reference state of charge of the battery, and constructing a battery charge and discharge data set.
3. The method for estimating the state of charge of the soft-pack lithium ion battery based on dynamic stress and deep learning according to claim 2, wherein in the step 11), plane stress measurement is performed by a constraint fixture and a stress sensor in a laboratory environment, and the stress sensor is arranged between two soft-pack batteries in vehicle-mounted application to measure plane stress.
4. The method for estimating the state of charge of a soft package lithium ion battery based on dynamic stress and deep learning as claimed in claim 2, wherein in said step 12), the dynamic stress S recorded during the test is d Is the stress related to the running state of the battery, which is the total stress S t Subtracting the initial static stress S s The following values:
S d ＝S t -S s
wherein the total stress is an instantaneous stress measured directly on the battery surface by the pressure sensor during charge/discharge, and the static stress is a stress obtained after the battery reaches an equilibrium state for a sufficient rest time;
the battery reference state of charge is calculated by definition:
wherein SOC is k And I k The battery state of charge and current at time k, respectively, Δt being the sampling time, C m Is the battery capacity.
5. The method for estimating the state of charge of the soft-pack lithium ion battery based on dynamic stress and deep learning according to claim 2, wherein in the step 12), the set working conditions include: long-term charge-discharge working condition, short-term charge-discharge working condition, pulse charge-discharge working condition and dynamic driving working condition.
6. The method for estimating a state of charge of a soft package lithium ion battery based on dynamic stress and deep learning as defined in claim 1, wherein in the step 21), the battery charge and discharge data at the k moment is x k ＝[I k ，V k ，F k ]Wherein I k Is the current at time k, V k Is the terminal voltage at time k, F k Is the dynamic stress at time k;
a sliding window with the length of n is adopted to obtain the input X for training the deep learning network k ＝[x k-n+1 ，...，x k-1 ，x k ]The data length is recorded as N data The sample size obtained is:
N samples ＝N daia -n+1
reorganizing the data into a three-dimensional tensor form: [ N ] samples ，n，N features ]Wherein N is features Representing the number of features.
7. The method for estimating the state of charge of the soft package lithium ion battery based on dynamic stress and deep learning according to claim 1, wherein in the step 22), the deep learning network comprises an input layer, two LSTM hidden layers, a full connection layer and an output layer, and the LSTM network activation process is as follows:
wherein f t Forgetting door, i t Is an input door o t Is an output door which is provided with a plurality of output doors,
is a candidate state for the update process, W x Is the weight, b x Is the bias, σ (·) is the sigmoid activation function, and tanh (·) is the hyperbolic tangent activation function.
8. The method for estimating the state of charge of the soft package lithium ion battery based on dynamic stress and deep learning according to claim 1, wherein in the step 32), the method for estimating the state of charge of the battery is as follows:
SOC * ＝LSTM(x * )
wherein LSTM is a trained battery state of charge estimation model, x * The input vector is composed of the battery current, terminal voltage and dynamic stress obtained on line.
9. A plug and play function test device for a charging pile, comprising a memory, a processor and a program stored in the memory, wherein the processor implements the method of any one of claims 1-8 when executing the program.
10. A storage medium having a program stored thereon, wherein the program, when executed, implements the method of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310184114.5A CN116224074A (en) | 2023-02-28 | 2023-02-28 | Soft package lithium ion battery state of charge estimation method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310184114.5A CN116224074A (en) | 2023-02-28 | 2023-02-28 | Soft package lithium ion battery state of charge estimation method, device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116224074A true CN116224074A (en) | 2023-06-06 |
Family
ID=86584007
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310184114.5A Pending CN116224074A (en) | 2023-02-28 | 2023-02-28 | Soft package lithium ion battery state of charge estimation method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116224074A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118033432A (en) * | 2024-04-12 | 2024-05-14 | 中国科学院大连化学物理研究所 | Battery state of charge estimation method and device and computer equipment |
| CN118409225A (en) * | 2024-07-03 | 2024-07-30 | 钛深科技(深圳)有限公司 | SOC calculation method, detection assembly, system and device for rechargeable battery |
| CN118409225B (en) * | 2024-07-03 | 2024-11-08 | 钛深科技(深圳)有限公司 | SOC calculation method, detection assembly, system and device for rechargeable battery |
-
2023
- 2023-02-28 CN CN202310184114.5A patent/CN116224074A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118033432A (en) * | 2024-04-12 | 2024-05-14 | 中国科学院大连化学物理研究所 | Battery state of charge estimation method and device and computer equipment |
| CN118409225A (en) * | 2024-07-03 | 2024-07-30 | 钛深科技(深圳)有限公司 | SOC calculation method, detection assembly, system and device for rechargeable battery |
| CN118409225B (en) * | 2024-07-03 | 2024-11-08 | 钛深科技(深圳)有限公司 | SOC calculation method, detection assembly, system and device for rechargeable battery |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Takyi-Aninakwa et al. | An optimized relevant long short-term memory-squared gain extended Kalman filter for the state of charge estimation of lithium-ion batteries | |
| Song et al. | Model-based health diagnosis for lithium-ion battery pack in space applications | |
| Che et al. | Semi-supervised self-learning-based lifetime prediction for batteries | |
| CN112557907A (en) | SOC estimation method of electric vehicle lithium ion battery based on GRU-RNN | |
| CN111832220A (en) | Lithium ion battery health state estimation method based on codec model | |
| CN113740736A (en) | Electric vehicle lithium battery SOH estimation method based on deep network self-adaptation | |
| CN109459699A (en) | A kind of lithium-ion-power cell SOC method of real-time | |
| Wei et al. | Lyapunov-based state of charge diagnosis and health prognosis for lithium-ion batteries | |
| CN112630659A (en) | Lithium battery SOC estimation method based on improved BP-EKF algorithm | |
| CN112782594B (en) | Method for estimating SOC (state of charge) of lithium battery by data-driven algorithm considering internal resistance | |
| Jiang et al. | An aging-aware soc estimation method for lithium-ion batteries using xgboost algorithm | |
| Zhao et al. | State-of-health estimation with anomalous aging indicator detection of lithium-ion batteries using regression generative adversarial network | |
| CN113093014B (en) | Online collaborative estimation method and system for SOH and SOC based on impedance parameters | |
| Li et al. | Multi-state joint estimation for a lithium-ion hybrid capacitor over a wide temperature range | |
| CN115629314B (en) | Battery parameter and state joint estimation method and system based on improved Jaya | |
| CN115308608A (en) | All-vanadium redox flow battery voltage prediction method, device and medium | |
| CN116224074A (en) | Soft package lithium ion battery state of charge estimation method, device and storage medium | |
| CN115407205A (en) | Multi-time scale SOC and SOH collaborative estimation method considering temperature influence | |
| Lyu et al. | State-of-charge estimation of lithium-ion batteries based on deep neural network | |
| CN118011211B (en) | Lithium battery state of charge and energy state estimation method | |
| CN117129898A (en) | Impedance spectrum data and R-GPR-based health state estimation method and system | |
| CN116125314A (en) | Lithium battery state of charge estimation method based on electrochemical model and particle filtering | |
| CN118156549A (en) | All-vanadium redox flow battery SOC prediction method based on artificial neural network | |
| Wang et al. | A novel weight coefficient calculation method for the real‐time state monitoring of the lithium‐ion battery packs under the complex current variation working conditions | |
| Zheng et al. | Remaining useful life indirect prediction of lithium-ion batteries using CNN-BiGRU fusion model and TPE optimization. |
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 |