CN113391225B - Lithium battery state-of-charge estimation method considering capacity degradation - Google Patents
Lithium battery state-of-charge estimation method considering capacity degradation Download PDFInfo
- Publication number
- CN113391225B CN113391225B CN202110545488.6A CN202110545488A CN113391225B CN 113391225 B CN113391225 B CN 113391225B CN 202110545488 A CN202110545488 A CN 202110545488A CN 113391225 B CN113391225 B CN 113391225B
- Authority
- CN
- China
- Prior art keywords
- lithium battery
- state
- input
- charge
- battery
- 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.)
- Active
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]
- G01R31/385—Arrangements for measuring battery or accumulator variables
- G01R31/387—Determining ampere-hour charge capacity or SoC
-
- 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
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
Abstract
The invention discloses a lithium battery state-of-charge estimation method considering capacity degradation, which comprises the steps of firstly, acquiring a lithium battery experimental data set, and calculating battery state-of-charge and health state data to obtain a complete data set; then constructing a hybrid neural network with two input ends, one output end, one or more circulating network layers and one or more full connection layers; and after the established hybrid neural network is trained by the complete data set, the hybrid neural network is used for estimating the state of charge of the battery to be estimated. The method is suitable for the state of charge estimation of the lithium battery considering capacity degradation, namely different health states, and has high estimation precision and high practical application value.
Description
Technical Field
The invention relates to the technical field of lithium batteries and artificial neural networks, provides a lithium battery state of charge estimation method, and particularly relates to a lithium battery state of charge estimation method considering capacity degradation.
Background
Accurate lithium battery state of charge estimation is of great importance to a lithium battery management system, and the method can help the battery management system to predict the remaining driving mileage and avoid overcharge or overdischarge of the battery and the like, so that the safety and reliability of the lithium battery are guaranteed. However, accurately estimating the state of charge of a lithium battery at the time of capacity degradation remains quite challenging.
Common lithium battery state-of-charge estimation methods mainly include an ampere-hour integration method, an open-circuit voltage method, a model-based method, a data driving method and the like. The ampere-hour integration method is simple, but has a large accumulated error; the open circuit voltage method requires the battery to be kept still for a long time to obtain a stable voltage value, and after the capacity of the battery is degraded, the change curve of the state of charge of the battery along with the open circuit voltage also changes, so that the open circuit voltage method cannot ensure the accuracy of the estimated value of the state of charge when the capacity degradation is considered; the model-based method usually simulates the dynamic characteristics of the interior of the battery by constructing an equivalent circuit model, an electrochemical model and the like of the battery, and estimates the state quantity in the battery model based on various filter algorithms such as extended kalman filter, particle filter and various observers such as a slip film observer, an extended state observer and the like, thereby realizing the estimation of the state of charge of the battery. The accuracy of the state of charge estimation result of the method highly depends on the accuracy of the constructed battery model, deep research and test on the battery are required for constructing the accurate battery model, and an unavoidable contradiction exists between the complexity of the model and the accuracy of the model, so that the applicability and the generalization capability of the model-based method are limited. Different from a model-based method, the data driving method does not care about the internal dynamic characteristics of the battery, but considers the battery as a black box, and directly learns the corresponding relation between some measured values such as voltage, current and the like and the charge state of the battery from experimental data, and the commonly used data driving method specifically comprises a support vector machine, a neural network and the like. Although the estimation results of the data-driven type method do not perform well in consideration of the dynamic discharge current or the dynamic ambient temperature, currently, there is less research on the estimation of the state of charge of the battery in consideration of the capacity degradation.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention provides a lithium battery state of charge estimation method considering capacity degradation. The method is suitable for estimating the state of charge of the lithium battery in different degradation states, and has great practical application value.
Step 1: acquiring a lithium battery experiment data set, wherein the data set comprises the rated capacity of a lithium battery and the voltage, current and time data of the lithium battery in the experiment process;
step 2: calculating the state of charge and the state of health of the lithium battery based on the lithium battery experimental data set to obtain a complete data set;
and 3, step 3: building a neural network comprising two input ends and one output end;
and 4, step 4: and training the neural network by using the complete data set, and using the trained neural network for estimating the state of charge of the lithium battery.
Optionally, the step 3 further comprises the following steps: the input end 1 of the neural network takes the voltage and the current of the battery as input and the input format is [ V, Cur ], wherein V and Cur are the voltage value and the current value of the battery respectively; the input end 2 of the neural network takes the state of health of the battery as input and the input format is [ SOH ], wherein SOH is the state of health of the battery; the neural network further comprises four circulating network layers and 2 full-connection layers, namely a circulating network layer 1, a circulating network layer 2, a circulating network layer 3, a circulating network layer 4, a full-connection layer 1 and a full-connection layer 2, wherein the circulating network layer 1 takes an input end 1 as an input, the circulating network layer 2 takes an output of the circulating network layer 1 as an input, the full-connection layer 1 takes an output of the circulating network layer 2 and an input end 2 as inputs, the full-connection layer 2 takes an output of the full-connection layer 1 as an input, the circulating network layer 3 takes an output of the full-connection layer 2 as an input, the circulating network layer 4 takes an output of the circulating network layer 3 as an input, and the output of the circulating network layer 4 is an output end of the neural network and outputs a lithium battery charge state value.
The invention has the beneficial effects that: the invention adopts the neural network with double input ends and combined with the circulating network layer and the full connection layer, and makes full use of the voltage data, the current data and the health state data of the lithium battery, thereby being capable of carrying out accurate state of charge estimation on the lithium battery with capacity degradation, namely under different health states.
Drawings
Fig. 1 is a flowchart illustrating steps of a lithium battery state of charge estimation method considering capacity degradation according to an embodiment of the present invention.
Fig. 2 is a diagram of a neural network structure constructed according to an embodiment of the present invention.
Fig. 3 shows the estimation result of the state of charge of the battery according to the embodiment of the invention.
Detailed Description
The invention is further described with reference to the accompanying drawings and specific embodiments.
As shown in fig. 1, a state of charge estimation method of a lithium battery considering capacity degradation includes the steps of:
s1, acquiring a lithium battery experiment data set, wherein the data set comprises the rated capacity of a lithium battery and voltage, current and time data in the experiment process. In the embodiment, five lithium battery standard charge and discharge cycle experimental data which are provided by CALCE center of Maryland university and are numbered as CS2_33, CS2_34, CS2_35, CS2_36 and CS2_37 and have rated capacity of 1.1Ah are taken as data sources, wherein the experimental data of the CS2_33, CS2_34, CS2_35 and CS2_36 batteries are taken as lithium battery experimental data sets, and the CS2_37 battery is taken as a battery to be estimated in the state of charge. In this experiment, five cells were all subjected to the same standard charging process at room temperature with a constant current rate of 0.5C until the voltage reached 4.2V, and then the charging voltage was maintained at 4.2V until the charging current dropped below 0.05A. Meanwhile, the discharge cutoff voltages of the four batteries are all 2.7V, wherein the discharge current multiplying powers of CS2_33 and CS2_34 are 0.5C, and the discharge current multiplying powers of CS2_35, CS2_36 and CS2_37 are all 1C.
And S2, calculating the state of charge and the health state of the lithium battery based on the lithium battery experimental data set to obtain a complete data set. In this example, the ampere-hour integral method was used to calculate the state of charge of the lithium battery:
therein, SOC0Is the initial state of charge, SOC, of the lithium batterytThe charge state value of the lithium battery at the time t, i is the discharge current of the lithium battery, and C is the current capacity of the lithium battery;
in this embodiment, the lithium battery health status is calculated using the following formula:
wherein C is the current capacity of the lithium battery, C0The lithium battery rated capacity is adopted, and the SOH is the lithium battery health state.
S3, establishing a neural network comprising two input ends, one output end, four circulation network layers and 2 full connection layers, wherein the circulation network layer 1 takes the input end 1 as input, the circulation network layer 2 takes the output of the circulation network layer 1 as input, the full connection layer 1 takes the output of the circulation network layer 2 and the input end 2 as input, the circulation network layer 3 takes the output of the full connection layer 1 as input, the circulation network layer 4 takes the output of the circulation network layer 3 as input, the full connection layer 2 takes the output of the circulation network layer 4 as input, and the output of the full connection layer 2 is the output end of the neural network and outputs the charge state value of the lithium battery. In this embodiment, the built neural network comprises 4 gated cyclic network layers and 2 fully-connected layers, and the specific structure is as shown in fig. 2, wherein layer 1, layer 2, layer 5 and layer 6 are gated cyclic network layers, layer 1, layer 2 and layer 5 are all composed of 80 gated cyclic units, layer 6 is composed of 1 gated cyclic unit, layer 3 and layer 4 are fully-connected layers, and each layer is composed of 80 fully-connected units.
And S4, training the neural network by using the complete data set, and using the trained neural network for estimating the state of charge of the lithium battery. In this embodiment, a complete data set is normalized, and the processed voltage, current, battery state of health, and corresponding battery state of charge data are used as a processed experimental data set, according to the following equation of 7: a scale of 3 divides the complete data set into a training set and a validation set and trains the neural network in conjunction with the adam optimizer.
And S5, estimating the state of charge of the lithium battery by using the trained neural network, wherein in the embodiment, the trained neural network is used for estimating the state of charge of the CS2_37 battery when the capacity is degraded to a state of health (SOH) of 1 and 0.8 respectively, and the estimation result of the state of charge and the experimental result based on an ampere-hour integration method are shown in FIG. 3. In two estimated discharge cycles, the root mean square errors between the estimated value of the state of charge of the battery and the experimental value based on the ampere-hour integration method are respectively 2.38% and 2.44%, and the feasibility and the effectiveness of the method in the state of charge estimation of the capacity-degraded lithium battery are proved.
Claims (1)
1. A lithium battery state-of-charge estimation method considering capacity degradation is characterized by comprising the following steps:
step 1: acquiring a lithium battery experiment data set, wherein the data set comprises the rated capacity of a lithium battery and the voltage, current and time data of the lithium battery in the experiment process;
step 2: calculating the state of charge and the state of health of the lithium battery based on the lithium battery experimental data set to obtain a complete data set;
and step 3: the method comprises the steps of building a neural network comprising two input ends and one output end, wherein the input end 1 of the neural network takes the voltage and the current of a battery as input and has an input format of [ V, Cur ], wherein V and Cur are respectively the voltage value and the current value of the battery, the input end 2 of the neural network takes the health state of the battery as input and has an input format of [ SOH ], wherein SOH is the health state of the battery, the neural network further comprises four circulation network layers and 2 full connection layers, namely a circulation network layer 1, a circulation network layer 2, a circulation network layer 3, a circulation network layer 4, a full connection layer 1 and a full connection layer 2, wherein the circulation network layer 1 takes the input end 1 as input, the circulation network layer 2 takes the output of the circulation network layer 1 as input, the full connection layer 1 takes the output of the circulation network layer 2 and the input end 2 as input, and the full connection layer 2 takes the output of the full connection layer 1 as input, the output of the full connection layer 2 is used as the input of the circulation network layer 3, the output of the circulation network layer 3 is used as the input of the circulation network layer 4, and the output of the circulation network layer 4 is the output end of the neural network and outputs the state of charge value of the lithium battery;
and 4, step 4: and training the neural network by using the complete data set, and using the trained neural network for estimating the state of charge of the lithium battery.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110545488.6A CN113391225B (en) | 2021-05-19 | 2021-05-19 | Lithium battery state-of-charge estimation method considering capacity degradation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110545488.6A CN113391225B (en) | 2021-05-19 | 2021-05-19 | Lithium battery state-of-charge estimation method considering capacity degradation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113391225A CN113391225A (en) | 2021-09-14 |
CN113391225B true CN113391225B (en) | 2022-06-03 |
Family
ID=77618050
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110545488.6A Active CN113391225B (en) | 2021-05-19 | 2021-05-19 | Lithium battery state-of-charge estimation method considering capacity degradation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113391225B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN118316094A (en) * | 2024-06-11 | 2024-07-09 | 国网辽宁省电力有限公司电力科学研究院 | Charging, discharging and storing integrated bidirectional energy intelligent collaboration system and method |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108519556A (en) * | 2018-04-13 | 2018-09-11 | 重庆邮电大学 | A kind of lithium ion battery SOC prediction techniques based on Recognition with Recurrent Neural Network |
CN110703101A (en) * | 2019-09-12 | 2020-01-17 | 北京交通大学 | Lithium ion battery inter-partition cycle capacity decline prediction method |
CN112269134A (en) * | 2020-09-10 | 2021-01-26 | 杭州电子科技大学 | Battery SOC and SOH joint estimation method based on deep learning |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101595956B1 (en) * | 2014-11-12 | 2016-02-22 | 충북대학교 산학협력단 | Apparatus and method for measuring state of charge(soc) for lithium ion battery |
CN105319515B (en) * | 2015-11-18 | 2017-12-19 | 吉林大学 | Charge states of lithium ion battery and health status joint estimate method |
KR20180055192A (en) * | 2016-11-16 | 2018-05-25 | 삼성전자주식회사 | Method and apparatus for estimating state of battery |
CN109031147B (en) * | 2018-08-21 | 2020-12-01 | 湖南兴业绿色电力科技有限公司 | SOC estimation method of lithium iron phosphate battery pack |
CN110082682B (en) * | 2019-03-12 | 2020-04-24 | 浙江大学 | Lithium battery state of charge estimation method |
CN110658459B (en) * | 2019-09-12 | 2021-10-15 | 北京航空航天大学 | Lithium ion battery state of charge estimation method based on bidirectional cyclic neural network |
CN111832220A (en) * | 2020-06-16 | 2020-10-27 | 天津大学 | Lithium ion battery health state estimation method based on codec model |
CN112067998A (en) * | 2020-09-10 | 2020-12-11 | 昆明理工大学 | Lithium ion battery state of charge estimation method based on deep neural network |
CN112557907A (en) * | 2020-12-17 | 2021-03-26 | 杭州六纪科技有限公司 | SOC estimation method of electric vehicle lithium ion battery based on GRU-RNN |
-
2021
- 2021-05-19 CN CN202110545488.6A patent/CN113391225B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108519556A (en) * | 2018-04-13 | 2018-09-11 | 重庆邮电大学 | A kind of lithium ion battery SOC prediction techniques based on Recognition with Recurrent Neural Network |
CN110703101A (en) * | 2019-09-12 | 2020-01-17 | 北京交通大学 | Lithium ion battery inter-partition cycle capacity decline prediction method |
CN112269134A (en) * | 2020-09-10 | 2021-01-26 | 杭州电子科技大学 | Battery SOC and SOH joint estimation method based on deep learning |
Non-Patent Citations (1)
Title |
---|
State of Charge Estimation and State of Health Diagnostic Method Using Multilayer Neural Networks;Jong-Hyun Lee 等;《2021 International Conference on Electronics, Information, and Communication (ICEIC)》;20210310;全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN113391225A (en) | 2021-09-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Zheng et al. | A novel capacity estimation method for lithium-ion batteries using fusion estimation of charging curve sections and discrete Arrhenius aging model | |
Chaoui et al. | State of charge and state of health estimation for lithium batteries using recurrent neural networks | |
Chaoui et al. | Lyapunov-based adaptive state of charge and state of health estimation for lithium-ion batteries | |
Kim | Nonlinear state of charge estimator for hybrid electric vehicle battery | |
CN104535934B (en) | The electrokinetic cell state of charge method of estimation and system of online feedforward compensation | |
CN108445402A (en) | A kind of lithium-ion-power cell state-of-charge method of estimation and system | |
KR101227417B1 (en) | A method for the SOC estimation of Li-ion battery and a system for its implementation | |
CN110658459B (en) | Lithium ion battery state of charge estimation method based on bidirectional cyclic neural network | |
CN112881930B (en) | Energy storage battery health management prediction method and system based on Internet of things | |
CN109459699A (en) | A kind of lithium-ion-power cell SOC method of real-time | |
Qiuting et al. | State of health estimation for lithium-ion battery based on D-UKF | |
Lee et al. | Li-ion battery electrode health diagnostics using machine learning | |
CN110462412A (en) | Device and method for estimating the SOC of battery | |
CN104035035A (en) | Method of determining the residual capacity of a battery | |
CN112946481A (en) | Based on federation H∞Filtering sliding-mode observer lithium ion battery SOC estimation method and battery management system | |
CN113391225B (en) | Lithium battery state-of-charge estimation method considering capacity degradation | |
CN112946480B (en) | Lithium battery circuit model simplification method for improving SOC estimation real-time performance | |
CN106004481A (en) | SOH value estimation method for battery pack of hybrid electric vehicle | |
Chaoui et al. | State of charge estimation of LiFePO 4 batteries with temperature variations using neural networks | |
CN113420444A (en) | Lithium ion battery SOC estimation method based on parameter online identification | |
CN117169723A (en) | Battery evaluation system based on multi-model fusion | |
Cen et al. | Online parameter estimation/tracking for lithium-ion battery rc model | |
Chen et al. | Online estimation of state of power for lithium-ion battery considering the battery aging | |
CN113740735A (en) | Method for estimating SOC of lithium ion battery | |
Chang et al. | An SOC estimation method based on sliding mode observer and the Nernst Equation |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |