CN111983457A - Battery pack SOH estimation method based on LSTM neural network - Google Patents

Battery pack SOH estimation method based on LSTM neural network Download PDF

Info

Publication number
CN111983457A
CN111983457A CN202010678378.2A CN202010678378A CN111983457A CN 111983457 A CN111983457 A CN 111983457A CN 202010678378 A CN202010678378 A CN 202010678378A CN 111983457 A CN111983457 A CN 111983457A
Authority
CN
China
Prior art keywords
neural network
lstm neural
battery pack
battery
output
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
Application number
CN202010678378.2A
Other languages
Chinese (zh)
Inventor
何志刚
沈晓宇
盘朝奉
李尧太
魏涛
陶袁雪
梁军
王丽梅
薛安荣
蔡涛
陈伟鹤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu University
Original Assignee
Jiangsu University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangsu University filed Critical Jiangsu University
Priority to CN202010678378.2A priority Critical patent/CN111983457A/en
Publication of CN111983457A publication Critical patent/CN111983457A/en
Pending legal-status Critical Current

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]
    • G01R31/392—Determining battery ageing or deterioration, e.g. state of health
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/048—Activation functions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Abstract

The invention provides a battery pack SOH estimation method based on an LSTM neural network, which comprises the steps of firstly, collecting battery pack charging data of an electric vehicle which actually runs for more than one year, carrying out correlation analysis, and extracting input characteristics of the LSTM neural network; calculating the maximum available capacity of the battery pack by using an SOC-electric quantity gain method, and taking the maximum available capacity as the output characteristic of the LSTM neural network; constructing an LSTM neural network model and determining an output value of the LSTM neural network; and finally, estimating the SOH of the battery pack by using the trained and verified LSTM neural network model. The input characteristic and the output characteristic of the LSTM neural network are both obtained by battery pack charging data of the electric vehicle which actually runs for more than one year, the method is suitable for estimating the SOH of the battery pack when the electric vehicle runs in a full state and in a full climate, and the problem that an estimation method based on laboratory verification is difficult to adapt to a complex actual working environment is solved.

Description

Battery pack SOH estimation method based on LSTM neural network
Technical Field
The invention belongs to the field of health state estimation of new energy electric vehicle battery packs, and particularly relates to a battery pack SOH estimation method based on an LSTM neural network.
Background
The lithium ion power battery is applied to the electric automobile due to the advantages of long cycle life, high energy density, good safety and the like. In order to meet the energy and power requirements of the electric automobile during actual operation, a large number of single batteries are connected in series and in parallel to the battery pack. From the viewpoint of safety and power performance of the electric vehicle, online estimation of State of Health (SOH) of the battery pack is important. The state of health of the battery pack directly reflects the aging condition of the battery pack and the current state of the battery pack (including the capacity and the power output capability). The correct estimation of the health state of the battery pack not only can avoid unsafe behaviors of the battery in time, but also can provide guarantee for maintenance and replacement of the battery pack.
Currently, the main methods for estimating the SOH of the battery pack are:
(1) table look-up method: testing the corresponding relation between the cycle number of the battery pack and the SOH through experiments, and estimating the SOH of the battery pack by off-line table look-up; however, the test period of experimental data is long, and the error of the estimation result is large.
(2) Internal resistance method: estimating the SOH of the battery pack by taking the internal resistance as an evaluation index of the SOH of the battery pack, but the internal resistance of the battery pack cannot be directly obtained, and a battery pack model needs to be established to perform parameter identification on the internal resistance of the battery pack; the accuracy of the established battery pack model and the identification precision of the model parameters greatly influence the SOH of the battery pack.
(3) The capacity increment method comprises the following steps: and selecting a characteristic variable from the IC curve, and taking the characteristic as an input data set to construct an SOH estimation model to estimate the SOH of the battery pack. Since ica (incremental Capacity analysis) generally requires constant charge and discharge data, this method is mostly applied to laboratory estimation of SOH of battery packs.
The estimation method needs to be completed in a laboratory, and in the actual driving process of the electric automobile, driving conditions, weather and driving behaviors of a driver have certain influences on the SOH of the battery pack, and the influences cannot be obtained through experimental simulation, so that the SOH of the battery pack cannot be estimated.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a battery pack SOH estimation method based on an LSTM neural network, which solves the problem that the estimation method based on laboratory verification is difficult to adapt to a complex actual working environment.
The present invention achieves the above-described object by the following technical means.
A battery pack SOH estimation method based on an LSTM neural network comprises the following steps:
step (1), collecting battery pack charging data of an electric vehicle which actually runs for more than one year, carrying out correlation analysis, and extracting input characteristics of an LSTM neural network;
step (2), calculating the maximum available capacity of the battery pack by using an SOC-electric quantity gain method, and using the maximum available capacity as the output characteristic of the LSTM neural network;
step (3), constructing an LSTM neural network model, and determining an output value of the LSTM neural network;
a forgetting gate sigmoid layer forgets part of information in the t-1 moment memory unit, then updates the information in the t moment memory unit, and an output gate determines an output value of the LSTM neural network;
and (4) estimating the SOH of the battery pack by using the trained and verified LSTM neural network model.
According to a further technical scheme, the input characteristics comprise current, voltage, temperature and SOC of the battery pack during charging.
In a further technical scheme, a calculation formula of the maximum available capacity of the battery pack is as follows:
Figure BDA0002584966220000021
wherein, CnormThe maximum available capacity of the battery pack, delta Q is the variation of the electric quantity of the battery pack, delta SOC is the variation of the SOC of the battery pack, and SOC (t)0) For the initial moment of discharge, the state of charge of the battery, SOC (t)end) For the state of charge of the battery at the end of discharge, i (t) is the battery current at time t, and η is the coulombic efficiency.
In a further technical scheme, the forgetting gate sigmoid layer forgets part of information in the memory unit at the time t-1, and the formula is as follows:
ft＝σ(Wfx·xt+Wfh·ht-1+bf)
wherein f istIs the output vector of the forgetting gate, sigma is sigmoid function,
Figure BDA0002584966220000022
is the input vector of the LSTM neural network at time t, M represents the input dimension,
Figure BDA0002584966220000023
is the output state of the LSTM neural network at time t-1,
Figure BDA0002584966220000024
and
Figure BDA0002584966220000025
are all a matrix of weights, and are,
Figure BDA0002584966220000026
to forget the biasing of the gate state, N represents the output dimension and R is the data matrix.
According to a further technical scheme, the information in the memory unit at the time t is updated according to the formula:
Ct＝ftCt-1+it tanh(Wzxxt+Wzhht-1+bz)
wherein itIs the output vector of the input gate, and it＝σ(Wix·xt+Wih·ht-1+bi)，Ct-1And CtThe memory cell states at time t-1 and time t respectively,
Figure BDA0002584966220000027
are all a matrix of weights, and are,
Figure BDA0002584966220000028
in order to input the offset of the gate state,
Figure BDA0002584966220000029
is the bias of the candidate gate state.
In a further technical scheme, the output gate determines the output value of the LSTM neural network, and the formula is as follows:
ht＝Ottanh(ct)
wherein, OtIs the output vector of the output gate, and Ot＝σ(Woxxt+Wohht-1+bo)，
Figure BDA0002584966220000031
For the output state of the LSTM neural network at time t,
Figure BDA0002584966220000032
are all a matrix of weights, and are,
Figure BDA0002584966220000033
is the biasing of the output gate state.
In a further technical scheme, the LSTM neural network model is trained by using input features and output features.
In a further technical scheme, the formula for estimating the SOH of the battery pack is as follows:
Figure BDA0002584966220000034
wherein C islstmPredicted battery capacity, C, for LSTM neural network0Is the initial capacity of the battery.
The invention has the beneficial effects that:
(1) the method adopts the LSTM neural network to estimate the SOH of the battery pack, and the input characteristic and the output characteristic of the LSTM neural network are obtained by the battery pack charging data of the electric vehicle which actually runs for more than one year, so that the estimation method is suitable for the SOH of the battery pack when the electric vehicle runs in the full state and the full climate, is not limited by the running environment, and overcomes the problem that the estimation method based on laboratory verification is difficult to adapt to the complex actual working environment.
(2) The LSTM neural network model not only forgets part of information in the t-1 moment memory unit, but also updates the information in the t moment memory unit, and can be improved according to data accumulated at any time, so that the method for estimating the SOH of the battery pack based on the LSTM neural network has stronger timeliness and applicability.
Drawings
FIG. 1 is a schematic diagram of the LSTM neural network model according to the present invention.
Detailed Description
The invention will be further described with reference to the following figures and specific examples, but the scope of the invention is not limited thereto.
A battery pack SOH estimation method based on an LSTM neural network specifically comprises the following steps:
and S1, collecting battery pack charging data of the electric vehicle which actually runs for more than one year (including four seasons of spring, summer, autumn and winter), carrying out correlation analysis on the collected data, and extracting input characteristics of the LSTM neural network, including current, voltage, temperature and SOC (residual electric quantity) of the battery pack during charging.
S2, calculating the maximum available capacity of the current battery pack through an SOC-electric quantity gain method
Because the charging data of the battery pack of the electric automobile is sampled every 20s and all current data in 20s cannot be obtained, the maximum available capacity of the battery pack is calculated by adopting a discretization SOC-electric quantity gain method and is used as the output characteristic of the LSTM neural network; the formula is as follows:
Figure BDA0002584966220000041
in the formula, CnormThe maximum available capacity of the battery pack, delta Q is the variation of the electric quantity of the battery pack, delta SOC is the variation of the SOC of the battery pack, and SOC (t)0) For the initial moment of discharge, the state of charge of the battery, SOC (t)end) For the state of charge of the battery at the end of discharge, i (t) is the battery current at time t, and η is the coulombic efficiency (in general, η ≈ 1).
S3, constructing an LSTM neural network model, and determining the output value of the LSTM neural network
As shown in FIG. 1, the LSTM neural network model includes an input gate, a forgetting gate, and an output gate, where the cell input given the current time is xtThe hidden state at the previous moment is ht-1The last moment memory cell state is Ct-1. Firstly, completing the memory cell C at the last moment by a forgetting gate sigmoid layert-1The partial information in the above step (A) is forgotten, and then the current time memory cell C is completed together with the input gate sigmoid layer and the candidate gate tanh layertThe state of the output gate is updated, and finally, the output gate sigmoid layer and the current moment memory cell C are usedtFinishing the update of the output value at the moment on the LSTM neural network, namely determining the output state h at the current momentt。
S31, firstly, a forgetting gate sigmoid layer forgets partial information in the t-1 time memory unit, and the formula is as follows:
ft＝σ(Wfx·xt+Wfh·ht-1+bf)
in the formula (f)tIs the output vector of the forgetting gate, sigma is sigmoid function,
Figure BDA0002584966220000042
is the input vector of the LSTM neural network at time t, M represents the input dimension,
Figure BDA0002584966220000043
is the output state of the LSTM neural network at time t-1,
Figure BDA0002584966220000044
and
Figure BDA0002584966220000045
are all a matrix of weights, and are,
Figure BDA0002584966220000046
to forget the biasing of the gate state, N represents the output dimension and R is the data matrix.
S32, the information in the storage unit at time t is updated, and the formula is as follows:
it＝σ(Wix·xt+Wih·ht-1+bi)
Ct＝ftCt-1+it tanh(Wzxxt+Wzhht-1+bz)
in the formula itFor the output vector of the input gate, tanh is a tanh function, Ct-1And CtThe memory cell states at time t-1 and time t respectively,
Figure BDA0002584966220000047
are all a matrix of weights, and are,
Figure BDA0002584966220000048
in order to input the offset of the gate state,
Figure BDA0002584966220000049
is the bias of the candidate gate state.
And S33, determining the output value of the LSTM neural network by the output gate, wherein the formula is as follows:
Ot＝σ(Woxxt+Wohht-1+bo)
ht＝Ottanh(ct)
in the formula, OtIs the output vector of the output gate,
Figure BDA0002584966220000051
for the output state of the LSTM neural network at time t,
Figure BDA0002584966220000052
Figure BDA0002584966220000053
are all a matrix of weights, and are,
Figure BDA0002584966220000054
is the biasing of the output gate state.
S4, estimating SOH of battery pack based on LSTM neural network model
S41, training LSTM neural network model
The LSTM neural network model is trained by using a certain electric vehicle data set (data provided by a manufacturer for more than one year) in a certain region, wherein the LSTM algorithm is a machine learning algorithm with supervised learning, and the LSTM neural network model is trained by using the maximum available capacity of a battery pack calculated by using a SOC-electric quantity gain method as an output characteristic and using voltage, current, temperature and SOC as input characteristics.
S42, testing four electric vehicle data sets (data provided by manufacturers for more than one year) in the same region, taking the RMSE (root mean square error) of a predicted value (LSTM neural network model output value) and a true value (namely the maximum available capacity of the battery pack) of the battery pack as an evaluation index, wherein the RMSE (root mean square error) is used for measuring the deviation between the predicted value and the true value of the battery capacity in the whole life cycle of the battery pack, and the RMSE of the four electric vehicles is shown in the table 1, except that the RMSE of the third vehicle is slightly larger than 1, the RMSE of the other vehicles is smaller than 1, so that the LSTM neural network model is effective.
TABLE 1 prediction error for four electric vehicles
Verifying the number of cars First vehicle The second vehicle Third vehicle Fourth vehicle
RMSE 0.715 0.472 1.044 0.503
S43, estimating the SOH of the battery pack by using the LSTM neural network model, and adopting the following formula:
Figure BDA0002584966220000055
wherein, ClstmPredicted battery capacity, C, for LSTM neural network0The initial capacity of the battery (provided by the manufacturer).
The present invention is not limited to the above-described embodiments, and any obvious improvements, substitutions or modifications can be made by those skilled in the art without departing from the spirit of the present invention.

Claims (8)

1. A battery pack SOH estimation method based on an LSTM neural network is characterized by comprising the following steps:
step (1), collecting battery pack charging data of an electric vehicle which actually runs for more than one year, carrying out correlation analysis, and extracting input characteristics of an LSTM neural network;
step (2), calculating the maximum available capacity of the battery pack by using an SOC-electric quantity gain method, and using the maximum available capacity as the output characteristic of the LSTM neural network;
step (3), constructing an LSTM neural network model, and determining an output value of the LSTM neural network;
a forgetting gate sigmoid layer forgets part of information in the t-1 moment memory unit, then updates the information in the t moment memory unit, and an output gate determines an output value of the LSTM neural network;
and (4) estimating the SOH of the battery pack by using the trained and verified LSTM neural network model.
2. The LSTM neural network-based battery SOH estimation method of claim 1, wherein the input characteristics include current, voltage, temperature and SOC of the battery while charging.
3. The LSTM neural network-based battery SOH estimation method of claim 1, wherein the calculation formula of the maximum available capacity of the battery is:
Figure FDA0002584966210000011
wherein, CnormThe maximum available capacity of the battery pack, delta Q is the variation of the electric quantity of the battery pack, delta SOC is the variation of the SOC of the battery pack, and SOC (t)0) For the initial moment of discharge, the state of charge of the battery, SOC (t)end) To the end time of dischargeThe state of charge of the battery, I (t), is the battery current at time t, and eta is the coulombic efficiency.
4. The battery pack SOH estimation method based on the LSTM neural network as claimed in claim 1, wherein the forgetting gate sigmoid layer forgets part of information in the memory unit at the time t-1, and the formula is as follows:
ft＝σ(Wfxxt+Wfh·ht-1+bf)
wherein f istIs the output vector of the forgetting gate, sigma is sigmoid function,
Figure FDA0002584966210000012
is the input vector of the LSTM neural network at time t, M represents the input dimension,
Figure FDA0002584966210000013
is the output state of the LSTM neural network at time t-1,
Figure FDA0002584966210000014
and
Figure FDA0002584966210000015
are all a matrix of weights, and are,
Figure FDA0002584966210000016
to forget the biasing of the gate state, N represents the output dimension and R is the data matrix.
5. The method of claim 4, wherein the updating of the information in the memory unit at time t is performed according to the formula:
Ct＝ftCt-1+ittanh(Wzxxt+Wzhht-1+bz)
wherein itIs the output vector of the input gate, and it＝σ(Wix·xt+Wih·ht-1+bi)，Ct-1And CtThe memory cell states at time t-1 and time t respectively,
Figure FDA0002584966210000021
are all a matrix of weights, and are,
Figure FDA0002584966210000022
in order to input the offset of the gate state,
Figure FDA0002584966210000023
is the bias of the candidate gate state.
6. The LSTM neural network-based battery SOH estimation method of claim 5, wherein the LSTM neural network output value is determined by an output gate, and the formula is:
ht＝Ottanh(ct)
wherein, OtIs the output vector of the output gate, and Ot＝σ(Woxxt+Wohht-1+bo)，
Figure FDA0002584966210000024
For the output state of the LSTM neural network at time t,
Figure FDA0002584966210000025
are all a matrix of weights, and are,
Figure FDA0002584966210000026
is the biasing of the output gate state.
7. The LSTM neural network-based battery SOH estimation method of claim 1, wherein the LSTM neural network model is trained using input features and output features.
8. The LSTM neural network-based battery SOH estimation method of claim 1, wherein the estimation of the battery SOH is based on the formula:
Figure FDA0002584966210000027
wherein C islstmPredicted battery capacity, C, for LSTM neural network0Is the initial capacity of the battery.
CN202010678378.2A 2020-07-15 2020-07-15 Battery pack SOH estimation method based on LSTM neural network Pending CN111983457A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010678378.2A CN111983457A (en) 2020-07-15 2020-07-15 Battery pack SOH estimation method based on LSTM neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010678378.2A CN111983457A (en) 2020-07-15 2020-07-15 Battery pack SOH estimation method based on LSTM neural network

Publications (1)

Publication Number Publication Date
CN111983457A true CN111983457A (en) 2020-11-24

Family

ID=73437876

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010678378.2A Pending CN111983457A (en) 2020-07-15 2020-07-15 Battery pack SOH estimation method based on LSTM neural network

Country Status (1)

Country Link
CN (1) CN111983457A (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112763929A (en) * 2020-12-31 2021-05-07 华东理工大学 Method and device for predicting health of battery monomer of energy storage power station system
CN113064093A (en) * 2021-03-22 2021-07-02 山东建筑大学 Energy storage battery state of charge and state of health joint estimation method and system
CN113128672A (en) * 2021-04-16 2021-07-16 昆明理工大学 Lithium ion battery pack SOH estimation method based on transfer learning algorithm
CN113219358A (en) * 2021-04-29 2021-08-06 东软睿驰汽车技术（沈阳）有限公司 Battery pack health state calculation method and system and electronic equipment
CN113740735A (en) * 2021-08-30 2021-12-03 昆明理工大学 Method for estimating SOC of lithium ion battery
CN113917337A (en) * 2021-10-13 2022-01-11 国网福建省电力有限公司 Battery health state estimation method based on charging data and LSTM neural network
CN114167301A (en) * 2021-11-30 2022-03-11 同济大学 Power battery evaluation method based on real vehicle data of electric vehicle
CN114791993A (en) * 2022-05-16 2022-07-26 江苏大学 SOH prediction method and system for power battery pack

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150066406A1 (en) * 2013-08-27 2015-03-05 The Regents Of The University Of Michigan On-board state of health monitoring of batteries using incremental capacity analysis
CN110598300A (en) * 2019-09-05 2019-12-20 东软睿驰汽车技术（沈阳）有限公司 Battery SOH prediction method and device
CN110824364A (en) * 2019-10-24 2020-02-21 重庆邮电大学 Lithium battery SOH estimation and RUL prediction method based on AST-LSTM neural network
US20200081070A1 (en) * 2017-11-20 2020-03-12 The Trustees Of Columbia University In The City Of New York Neural-network state-of-charge and state of health estimation
CN111308375A (en) * 2020-02-04 2020-06-19 浙江大学 LSTM-FFNN-based electric forklift lithium ion battery health state prediction method

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150066406A1 (en) * 2013-08-27 2015-03-05 The Regents Of The University Of Michigan On-board state of health monitoring of batteries using incremental capacity analysis
US20200081070A1 (en) * 2017-11-20 2020-03-12 The Trustees Of Columbia University In The City Of New York Neural-network state-of-charge and state of health estimation
CN110598300A (en) * 2019-09-05 2019-12-20 东软睿驰汽车技术（沈阳）有限公司 Battery SOH prediction method and device
CN110824364A (en) * 2019-10-24 2020-02-21 重庆邮电大学 Lithium battery SOH estimation and RUL prediction method based on AST-LSTM neural network
CN111308375A (en) * 2020-02-04 2020-06-19 浙江大学 LSTM-FFNN-based electric forklift lithium ion battery health state prediction method

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112763929A (en) * 2020-12-31 2021-05-07 华东理工大学 Method and device for predicting health of battery monomer of energy storage power station system
CN112763929B (en) * 2020-12-31 2024-03-08 华东理工大学 Method and device for predicting health of battery monomer of energy storage power station system
CN113064093A (en) * 2021-03-22 2021-07-02 山东建筑大学 Energy storage battery state of charge and state of health joint estimation method and system
CN113064093B (en) * 2021-03-22 2024-03-08 百思科新能源技术(青岛)有限公司 Method and system for jointly estimating state of charge and state of health of energy storage battery
CN113128672A (en) * 2021-04-16 2021-07-16 昆明理工大学 Lithium ion battery pack SOH estimation method based on transfer learning algorithm
CN113219358A (en) * 2021-04-29 2021-08-06 东软睿驰汽车技术（沈阳）有限公司 Battery pack health state calculation method and system and electronic equipment
CN113740735A (en) * 2021-08-30 2021-12-03 昆明理工大学 Method for estimating SOC of lithium ion battery
CN113917337A (en) * 2021-10-13 2022-01-11 国网福建省电力有限公司 Battery health state estimation method based on charging data and LSTM neural network
CN114167301A (en) * 2021-11-30 2022-03-11 同济大学 Power battery evaluation method based on real vehicle data of electric vehicle
CN114791993A (en) * 2022-05-16 2022-07-26 江苏大学 SOH prediction method and system for power battery pack

Similar Documents

Publication Publication Date Title
CN111983457A (en) Battery pack SOH estimation method based on LSTM neural network
Lin et al. State of charge estimation with the adaptive unscented Kalman filter based on an accurate equivalent circuit model
Shu et al. Stage of charge estimation of lithium-ion battery packs based on improved cubature Kalman filter with long short-term memory model
CN111581904B (en) Lithium battery SOC and SOH collaborative estimation method considering cycle number influence
Tian et al. State of charge estimation of lithium-ion batteries using an optimal adaptive gain nonlinear observer
Ouyang et al. Improved parameters identification and state of charge estimation for lithium-ion battery with real-time optimal forgetting factor
US11965935B2 (en) Method and apparatus for operating a system for providing predicted states of health of electrical energy stores for a device using machine learning methods
CN104714188A (en) Method and system for estimating measured noise variance matrix matched battery state of charge (SOC)
KR102297343B1 (en) Battery Output Voltage Response and State-of-Charge Forecasting Method using Hybrid VARMA and LSTM
Li et al. Discharge capacity estimation for Li-ion batteries based on particle filter under multi-operating conditions
US20220099743A1 (en) Method and Apparatus for Operating a System for Providing Predicted States of Health of Electrical Energy Stores for a Device Using Machine Learning Methods
AU2021100373A4 (en) A Battery Pack SOH Estimation Method Based on LSTM Neural Network
CN112269133B (en) SOC estimation method based on pre-charging circuit model parameter identification
CN111458646A (en) Lithium battery SOC estimation method based on PSO-RBF neural network
CN113805075A (en) BCRLS-UKF-based lithium battery state of charge estimation method
Dong et al. Low complexity state-of-charge estimation for lithium-ion battery pack considering cell inconsistency
CN113406503A (en) Lithium battery SOH online estimation method based on deep neural network
KR102246633B1 (en) Battery Output Voltage Response and State-of-Charge Forecasting Optimization Method using Hybrid VARMA and LSTM
CN114861545A (en) Lithium battery SOP online estimation method based on RNN neural network and multi-parameter constraint
CN114779089A (en) Method for calculating battery state of charge based on energy storage lithium battery equivalent circuit model
CN115201686A (en) Lithium ion battery health state assessment method under incomplete charging and discharging data
US20220011370A1 (en) Apparatus for mining battery characteristic data having sensitivity to solid phase diffusion coefficient of battery electrode and method thereof
CN113495214A (en) Super capacitor charge state estimation method based on temperature change model
Sugashini et al. Hybrid Deep Learning Algorithm for the State of Charge Prediction of the Lithium-Ion Battery for Electric Vehicles
CN116466250A (en) Dynamic working condition model error characteristic-based power battery health state estimation method

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20201124