CN117054892B - Evaluation method, device and management method for battery state of energy storage power station - Google Patents
Evaluation method, device and management method for battery state of energy storage power station Download PDFInfo
- Publication number
- CN117054892B CN117054892B CN202311309810.0A CN202311309810A CN117054892B CN 117054892 B CN117054892 B CN 117054892B CN 202311309810 A CN202311309810 A CN 202311309810A CN 117054892 B CN117054892 B CN 117054892B
- Authority
- CN
- China
- Prior art keywords
- battery
- model
- soh
- target
- energy storage
- 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
- 238000004146 energy storage Methods 0.000 title claims abstract description 80
- 238000011156 evaluation Methods 0.000 title claims abstract description 52
- 238000007726 management method Methods 0.000 title claims abstract description 8
- 238000000034 method Methods 0.000 claims abstract description 68
- 238000012549 training Methods 0.000 claims abstract description 48
- 238000013210 evaluation model Methods 0.000 claims abstract description 44
- 230000036541 health Effects 0.000 claims abstract description 28
- 238000013526 transfer learning Methods 0.000 claims abstract description 22
- 238000013528 artificial neural network Methods 0.000 claims abstract description 10
- 230000008569 process Effects 0.000 claims description 40
- 238000007599 discharging Methods 0.000 claims description 38
- 210000004027 cell Anatomy 0.000 claims description 33
- 239000000178 monomer Substances 0.000 claims description 21
- 238000012360 testing method Methods 0.000 claims description 13
- 238000005520 cutting process Methods 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 11
- 238000009499 grossing Methods 0.000 claims description 11
- 238000010276 construction Methods 0.000 claims description 10
- 238000013508 migration Methods 0.000 claims description 10
- 230000005012 migration Effects 0.000 claims description 10
- 238000007600 charging Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 6
- 230000015654 memory Effects 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 4
- 230000002035 prolonged effect Effects 0.000 claims description 4
- 238000010277 constant-current charging Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 claims 1
- 239000000758 substrate Substances 0.000 claims 1
- 238000009826 distribution Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 7
- 238000012983 electrochemical energy storage Methods 0.000 description 6
- 230000002457 bidirectional effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 230000006403 short-term memory Effects 0.000 description 3
- 230000032683 aging Effects 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- GELKBWJHTRAYNV-UHFFFAOYSA-K lithium iron phosphate Chemical compound [Li+].[Fe+2].[O-]P([O-])([O-])=O GELKBWJHTRAYNV-UHFFFAOYSA-K 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 206010000117 Abnormal behaviour Diseases 0.000 description 1
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 1
- BNOODXBBXFZASF-UHFFFAOYSA-N [Na].[S] Chemical compound [Na].[S] BNOODXBBXFZASF-UHFFFAOYSA-N 0.000 description 1
- 239000002253 acid Substances 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010280 constant potential charging Methods 0.000 description 1
- 238000010281 constant-current constant-voltage charging Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 229910001416 lithium ion Inorganic materials 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 238000013102 re-test Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
-
- G—PHYSICS
- 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]
- G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G06N3/096—Transfer learning
Abstract
The invention provides an evaluation method, an evaluation device and a management method for SOH (state of health) of an energy storage power station battery, and relates to the technical field of batteries. The method comprises the following steps: constructing an SOH evaluation model based on a neural network; acquiring source domain data, training an SOH evaluation model by the source domain data, and obtaining a source domain model; acquiring a preset number of target domain data, retraining a source domain model by the preset number of target domain data, and obtaining a transfer learning model for SOH evaluation; and estimating the health state of the target battery of the energy storage power station based on the SOH estimation transfer learning model. The method at least solves the problems that in the related art, the comprehensive battery pack operation data cannot be obtained to train an evaluation model, and the data distribution is different when an evaluation object is changed from a battery cell to a battery pack, so that SOH evaluation accuracy is low and the evaluation model fails.
Description
Technical Field
The invention relates to the technical field of batteries, in particular to an evaluation method, an evaluation device and a management method for SOH (state of health) of a battery of an energy storage power station.
Background
Energy storage power station cells typically have a large capacity and high energy density, and can be charged at low loads and discharged at high loads to balance supply and demand differences in the power system. The energy storage power station battery can be used for balancing load fluctuation of a power grid, providing standby power, storing renewable energy sources and the like. Common types of energy storage power station batteries include lead acid batteries, lithium ion batteries, sodium sulfur batteries, and the like.
The use strategy of the battery can be optimized, the service life of the battery can be prolonged, the fault risk can be reduced, and a basis is provided for the design and management of an energy storage system by accurately evaluating and online monitoring the health state of the battery of the energy storage power station; scheduling, capacity planning and fault diagnosis can be optimized, so that the cycle times of the battery are prolonged, and stable operation of the energy storage system is ensured; potential fault characteristics and abnormal behaviors can be identified, and early warning information is provided so as to take corresponding maintenance and replacement measures, and the influence of battery faults on an energy storage system is avoided.
Currently, the method for evaluating the state of health of an energy storage battery includes: the method comprises the steps of obtaining actual operation data in the actual operation process of an energy storage battery, inputting the actual operation data into a pre-trained long-short-period memory network LSTM for estimating the SOH (State of Health) of the energy storage battery, and obtaining an SOH estimated value of the energy storage battery.
In the actual deployment of the evaluation method, when the evaluation object is required to be converted into a battery cell, a battery module, a battery cluster, a battery stack and a battery compartment of the electrochemical energy storage power station, the problem that the evaluation accuracy is reduced or the model fails and the energy storage power station cannot accurately evaluate SOH and RUL (Remaining Useful Life, life remaining prediction) in the initial stage of the operation stage is caused by the difference in data distribution, and the performance of the evaluation model can be fully exerted only by retraining the operation data of the complete energy storage power station. However, the time and labor costs required to obtain the operational data required for retraining are significant, severely hampering the application of this method in energy storage power stations.
Disclosure of Invention
The technical problem to be solved by the invention is to provide an evaluation method, an evaluation device and a management method for SOH of the battery state of the energy storage power station, which aim to accurately evaluate SOH of the battery cells, the battery modules, the battery clusters, the battery stacks and the battery cabins of the power station by combining a small amount of operation data retraining evaluation models of the energy storage power station on the basis of the remote training of SOH evaluation models of laboratory data.
In a first aspect, the present invention provides a method for evaluating the SOH of a battery in an energy storage power station, including: constructing an SOH evaluation model based on a neural network; acquiring source domain data, training an SOH evaluation model by the source domain data, and obtaining a source domain model; acquiring a preset number of target domain data, retraining a source domain model by the preset number of target domain data, and obtaining a transfer learning model for SOH evaluation; and estimating the health state of the target battery of the energy storage power station based on the SOH estimation transfer learning model.
Preferably, the target battery includes any one of a battery cell, a battery module, a battery cluster, a battery stack, and a battery compartment.
The step of acquiring source domain data and training an SOH evaluation model by the source domain data specifically comprises the following steps: acquiring capacity increment data sets of a plurality of battery monomers in a laboratory in a discharging process respectively to obtain source domain data, wherein the capacity increment satisfies the following formula:
，
and->The discharge capacity at two adjacent moments is respectively, the unit of the discharge capacity is ampere hour, and the unit of the discharge capacity is +.>And->The terminal voltages of the batteries at two adjacent moments are respectively in units of volts; and (5) inputting the source domain data as a model, and training and testing an SOH evaluation model.
Preferably, after the acquiring of the incremental capacity data sets of the plurality of battery cells during discharge, respectively, and before the obtaining of the source domain data, the evaluation method further comprises: smoothing capacity increment data sets of a plurality of battery monomers in a laboratory in a discharging process respectively; cutting the smoothed capacity increment data set into two first short characteristic sample sets according to the voltage of the following formula:
wherein,to the discharge cut-off voltage of the laboratory cell,band the charge cut-off voltage of the battery cell of the laboratory.
Preferably, the acquiring the capacity increment data set of the plurality of battery cells in the laboratory during the discharging process specifically includes: and acquiring a capacity increment data set of a plurality of battery monomers in a laboratory, wherein the capacity increment data set respectively meets the following preset conditions in the discharging process, the preset conditions comprise that the voltage amplitude is larger than a first threshold value, and the unit change amplitude of the voltage amplitude is larger than a second threshold value.
Preferably, the acquiring the preset number of target domain data and retraining the source domain model with the preset number of target domain data specifically includes: acquiring a capacity increment data set of a target battery of an energy storage power station in a discharging process respectively to obtain target domain data of a preset number; and inputting a preset number of target domain data as a source domain model, and retraining and retesting the source domain model.
Preferably, after the acquiring the capacity increment data set of the target battery of the energy storage power station during discharging respectively and before obtaining the preset number of target domain data, the evaluation method further comprises: smoothing capacity increment data sets of the target battery of the energy storage power station in the discharging process respectively; cutting the smoothed capacity increment data set into two second short characteristic sample sets according to the voltage of the following formula:
wherein,Ato discharge cut-off voltage of the energy storage power station target battery,Band (5) charging cut-off voltage of the target battery of the energy storage power station.
Preferably, the constructing the SOH evaluation model based on the neural network specifically includes: constructing an SOH evaluation model based on the long-short-term memory network LSTM and the gating circulating unit GRU; determining the number of layers of the SOH evaluation model, the number of neurons in each layer and an activation function; initializing parameters of the SOH evaluation model.
In a second aspect, the invention further provides an evaluation device for the SOH of the battery state of health of the energy storage power station, which comprises a construction module, a pre-training module, a retraining module and an evaluation module.
And the construction module is used for constructing an SOH evaluation model based on the neural network. The pre-training module is connected with the construction module and used for acquiring source domain data, training the SOH evaluation model with the source domain data and obtaining a source domain model. The retraining module is connected with the pre-training module and used for acquiring the preset number of target domain data, retraining the source domain model with the preset number of target domain data and obtaining the migration learning model of SOH evaluation. And the evaluation module is connected with the retraining module and is used for evaluating the health state of the target battery of the energy storage power station based on the SOH evaluation transfer learning model.
Preferably, the target battery includes any one of a battery cell, a battery module, a battery cluster, a battery stack, and a battery compartment.
The pre-training module is specifically configured to obtain capacity increment data sets of a plurality of battery monomers in a laboratory in a discharging process respectively, so as to obtain source domain data, where the capacity increment satisfies the following formula:
and->The discharge capacity at two adjacent moments, respectively +.>And->Respectively the terminal voltages of batteries at two adjacent moments, and the source domainThe data is input as a model, and the SOH assessment model is trained and tested.
In a third aspect, the present invention further provides a method for managing a battery of an energy storage power station, including: obtaining the state of health of the target battery of the energy storage power station according to the method for evaluating the state of health SOH of the battery of the energy storage power station in the first aspect; optimizing a usage policy of the target battery or predicting a failure of the target battery according to the state of health.
According to the assessment method, the assessment device and the management method for the SOH of the battery state of the energy storage power station, the pre-training of the assessment model is finished in the source domain data, the pre-trained assessment model is migrated to a small number of data sets of the target domain data to be retrained, and finally the migration learning model capable of assessing the SOH assessment of the target battery such as the battery pack, the battery cluster and the like is obtained, and even if an assessment object or a scene is changed, the migration learning model of the SOH assessment is still effective, so that an assessment result with high accuracy can be obtained. The problem that a traditional machine learning model cannot be directly used after deployment and field data retraining is solved by breaking through distribution limitation among data through transfer learning, the aim of evaluating the SOH of the battery pack by using a single battery SOH evaluation model is fulfilled, and the possibility of evaluating the running state of the whole station is fulfilled by using an evaluation model formed by a small number of samples.
Drawings
Fig. 1 is a flow chart of a method for evaluating SOH of battery state of health of an energy storage power station according to embodiment 1 of the present invention;
FIG. 2 is a schematic diagram of a SOH evaluation transfer learning model according to embodiment 1 of the present invention;
FIG. 3 is a schematic diagram of a migration path of a SOH-estimated migration learning model according to embodiment 1 of the present invention;
fig. 4 is a schematic structural diagram of an evaluation device for SOH of battery state of health of an energy storage power station according to embodiment 2 of the present invention.
Detailed Description
In order to make the technical scheme of the present invention better understood by those skilled in the art, the following detailed description of the embodiments of the present invention will be given with reference to the accompanying drawings.
It is to be understood that the specific embodiments and figures described herein are merely illustrative of the invention, and are not limiting of the invention.
It is to be understood that the various embodiments of the invention and the features of the embodiments may be combined with each other without conflict.
It is to be understood that only the portions relevant to the present invention are shown in the drawings for convenience of description, and the portions irrelevant to the present invention are not shown in the drawings.
It should be understood that each unit and module in the embodiments of the present invention may correspond to only one physical structure, may be formed by a plurality of physical structures, or may be integrated into one physical structure.
It will be appreciated that, without conflict, the functions and steps noted in the flowcharts and block diagrams of the present invention may occur out of the order noted in the figures.
It is to be understood that the flowcharts and block diagrams of the present invention illustrate the architecture, functionality, and operation of possible implementations of systems, apparatuses, devices, methods according to various embodiments of the present invention. Where each block in the flowchart or block diagrams may represent a unit, module, segment, code, or the like, which comprises executable instructions for implementing the specified functions. Moreover, each block or combination of blocks in the block diagrams and flowchart illustrations can be implemented by hardware-based systems that perform the specified functions, or by combinations of hardware and computer instructions.
It should be understood that the units and modules related in the embodiments of the present invention may be implemented by software, or may be implemented by hardware, for example, the units and modules may be located in a processor.
Example 1
As shown in fig. 1, the embodiment provides a method for evaluating the state of health SOH of a battery of an energy storage power station, which includes:
and step 101, constructing an SOH evaluation model based on the neural network.
Specifically, constructing the SOH evaluation model based on the neural network includes: constructing an SOH evaluation model based on the long-short-term memory network LSTM and the gating circulating unit GRU; determining the number of layers of the SOH evaluation model, the number of neurons in each layer and an activation function; initializing parameters of the SOH evaluation model.
In this embodiment, LSTM and GRU networks and their variants are selected as the main network for modeling, and as shown in fig. 2, the main network is composed of 2 hidden layers, the number of neurons of the first hidden layer (i.e. hidden layer 1) is designed to be 32, and the number of neurons of the second hidden layer (i.e. hidden layer 2) is designed to be 48. When the transfer learning model taking GRU and the like as the main network is used for evaluating the SOH of the battery pack, the transfer learning model meets the requirement of the operation detection of the battery in the energy storage power station on timeliness due to short training and prediction time, and is suitable for deployment in a scene of a large-scale electrochemical energy storage power station. The main network may also be LSTM (Long Short-Term Memory network), bi-LSTM (Bidirectional Long Short-Term Memory network, bidirectional Long-Term Memory network), CNN-LSTM (Convolutional Neural Network-Long Short-Term Memory network, convolutional neural network-Long-Term Memory network), GRU (Gated Recurrent Unit, gated loop unit), bi-GRU (Bidirectional Gated Recurrent Unit, bidirectional gated loop unit).
Step 102, acquiring source domain data, and training an SOH evaluation model with the source domain data to obtain a source domain model.
Specifically, acquiring source domain data and training an SOH evaluation model with the source domain data, including: acquiring capacity increment data sets of a plurality of battery monomers in a laboratory in a discharging process respectively to obtain source domain data, wherein the capacity increment satisfies the following formula:
，
and->Respectively adjacent twoThe discharge capacity at each moment in time, the unit of discharge capacity is ampere hour->And->The terminal voltages of the batteries at two adjacent moments are respectively in units of volts; and (5) inputting the source domain data as a model, and training and testing an SOH evaluation model.
In this embodiment, as shown in fig. 3, the source domain data is a set of battery monomer data generated in a laboratory, the data includes SOH, voltage, capacity, multiplying power, and the like, the capacity increment curve in the discharging process is selected as the input of the SOH evaluation model in this embodiment, the source domain data can be divided into a training set and a testing set, the number ratio of the training set and the testing set is preferably 9:1, and the greater the number ratio, the better the quality of the model evaluation effect can be ensured.
Optionally, acquiring capacity increment data sets of a plurality of battery cells in a laboratory during discharging respectively, which specifically includes: and acquiring a capacity increment data set of a plurality of battery monomers in a laboratory, wherein the capacity increment data set respectively meets the following preset conditions in the discharging process, the preset conditions comprise that the voltage amplitude is larger than a first threshold value, and the unit change amplitude of the voltage amplitude is larger than a second threshold value.
In this embodiment, in the discharging process of the energy storage battery, the information directly recorded by the test device has the characteristics of being dense and smooth, but the reflected information is relatively less. The capacity of the laboratory cell as a sample varies only slightly over a large voltage interval, and only over a range above a first threshold (e.g., 3V). Similarly, the voltage curves only show a clear difference near the end of the discharge, and sample data for different cycles, different SOHs, are less different during the plateau occupying most of the time. The adoption of directly recorded information as the characteristics is unfavorable for the rapid and accurate completion of training and evaluation tasks in transfer learning. Therefore, the SOH evaluation model is input by adopting a capacity increment curve in the discharging process of the battery, and the capacity increment is obtained by calculating the ratio of the capacity increment and the voltage of a plurality of battery monomer samples in a laboratory at the same time, wherein the calculation method of the capacity increment is as follows:
。
and because the lithium iron phosphate battery voltage curve has long and gentle plateau period, and errors are caused by the occurrence of conditions in the process of extracting the characteristics of the IC (Incremental Capacity, capacity increment), the second threshold (such as 0.3 mV) related to voltage change is set, the proportion of data in the plateau period is shortened, the obstacle of characteristic extraction is solved, and the model training efficiency is improved.
Optionally, after the acquiring the incremental capacity data sets of the plurality of battery cells during discharge, respectively, and before the obtaining the source domain data, the evaluating method further comprises: smoothing capacity increment data sets of a plurality of battery monomers in a laboratory in a discharging process respectively; cutting the smoothed capacity increment data set into two first short characteristic sample sets according to the voltage of the following formula:
wherein,to the discharge cut-off voltage of the laboratory cell,band the charge cut-off voltage of the battery cell of the laboratory.
In this embodiment, the smoothing process includes Savitzky-Golay (savitz-Golay Lei Fangfa), and the purpose of the smoothing process is to remove noise of data, so that the accuracy of the evaluation model can be effectively improved. The effect of cutting the data set into two short feature sample sets based on voltage: the capacity increment dQ/dV is obtained by the capacity and the voltage, so that the voltage is used for distinguishing, the sequence length is divided, and the operation is convenient. And the voltage is used for replacing the SOC (State of Charge) as a cutting basis, so that the problem of repeated calibration of the SOC caused by capacity degradation of the battery is avoided. Cut into short sectionsThe feature sample set can shorten the training and evaluation time of the model and improve the evaluation efficiency. The voltage ranges of the two first short feature sample sets of the laboratory battery cell of the present embodiment are respectivelyAnd->。
And step 103, acquiring a preset number of target domain data, and retraining a source domain model by the preset number of target domain data to obtain a transfer learning model for SOH evaluation.
In this embodiment, the preset number may be understood as a small number, that is, the present embodiment may retrain with a small amount of target domain data. The target battery comprises any one of a battery cell, a battery module, a battery cluster, a battery stack and a battery compartment. The target domain data is a data set of a battery cell, a battery module, a battery cluster, a battery stack and a battery compartment of the energy storage power station, and as shown in fig. 3, the data includes SOH, voltage, capacity, multiplying power and the like. When the SOH evaluation transfer learning model is required to be used for evaluating the health state of the battery cluster, the source domain model is transferred to a small number of data sets of the battery cluster for retraining, so that the SOH evaluation transfer learning model applicable to the battery cluster is obtained. Similarly, if the source domain model is migrated to a small number of data sets of the battery module for retraining, a migration learning model applicable to SOH evaluation of the battery module is obtained. The premise of the transfer learning can be expressed as: source domainIs a migrated object, generally the data volume of the source domain is larger, and the labels are rich; target domainIs an object for giving knowledge and realizing annotation. Learning a predictive function on a target domain using source domain data>Make->The migration of knowledge from the source domain to the target domain is completed with minimal error in the target domain,xis a sample of the sample and,yis a label which is a label of the type,Nsin order to be a source domain space,Ntis the target domain space.
The migration objective function may be expressed as:
is a function->The variable takes its value when the minimum value is reached,
for functions in the dataset->As a result of the above-mentioned desire,
as a loss function.
Model test Source Domain and test target Domain dataset acquisition approach examples: referring to the running state of the electrochemical energy storage power station when participating in the peak shaving task, the discharge current of the energy storage battery changes frequently and rapidly. The battery in the energy storage power station is low in charge-discharge multiplying power adopted when working, and the charging and discharging are rarely performed at the multiplying power of more than 1C when the multiplying power is between 0.5 and 1C. Under the requirement of simplifying the experimental flow, the main process step of the experiment is designed as a constant current charging and discharging process of 0.8C, and charging and discharging currents with different magnitudes are set according to the difference of the self capacities of the samples. In order to ensure the relative stability of the input power and the output power, the constant-current-constant-voltage charging mode commonly adopted in the power battery is not adopted in the operation process of the electrochemical energy storage power station, and the constant-voltage charging process is avoided. In addition, in the pre-experiment, after the battery pack experimental sample reaches the cut-off voltage after charge and discharge, the open circuit voltage needs to be restored to be stable longer than the battery cell experimental sample, so that the standing time in the experimental flow of the battery pack is prolonged by 10 minutes, and the possible influence of polarization effect on measurement is eliminated.
Specifically, acquiring a preset number of target domain data, and retraining a source domain model with the preset number of target domain data, which specifically includes: acquiring a capacity increment data set of a target battery of an energy storage power station in a discharging process respectively to obtain target domain data of a preset number; and inputting a preset number of target domain data as a source domain model, and retraining and retesting the source domain model.
In this embodiment, after the training of the traditional machine learning model on the battery monomer data set is completed, the traditional machine learning model is directly used for retraining all the aging operation data of the battery pack of the energy storage power station, which is required to be obtained, after the training is completed, on the battery monomer data set, so that all the aging operation data are difficult to obtain in the initial stage of the energy storage power station. As shown in fig. 2, after the source domain model is obtained, the first layer hidden layer information of the source domain model is frozen, the second layer hidden layer is kept activated, target domain data randomly extracted from the energy storage power station is imported to perform training optimization again on the second layer hidden layer parameters, test evaluation is performed on the SOH of the battery pack, and finally the migration learning model for evaluating the SOH of the battery pack is obtained. The target domain data can be divided into a training set and a testing set, and the quantity ratio of the training set to the testing set is preferably 9:1.
Optionally, after the acquiring the capacity increment data set of the target battery of the energy storage power station during the discharging process respectively and before the obtaining of the preset number of target domain data, the evaluation method further includes: smoothing capacity increment data sets of the target battery of the energy storage power station in the discharging process respectively; cutting the smoothed capacity increment data set into two second short characteristic sample sets according to the voltage of the following formula:
wherein,Ato discharge cut-off voltage of the energy storage power station target battery,Band (5) charging cut-off voltage of the target battery of the energy storage power station.
In this embodiment, the short feature sample set is cut, so that training and evaluation time of the shift learning model for SOH evaluation can be shortened, and evaluation efficiency can be improved. The voltage ranges of the two second short feature sample sets of the battery pack of the energy storage power station in this embodiment are respectivelyAnd->。
And 104, estimating the health state of the target battery of the energy storage power station based on the SOH estimated transfer learning model.
According to the method for evaluating the SOH of the battery state of health of the energy storage power station, from the application point of the energy storage power station, the method for evaluating the SOH of the high-capacity lithium iron phosphate energy storage system which is migrated from the battery monomer to the battery module, the battery cluster, the battery stack and the battery compartment is designed. Because the traditional machine learning model cannot be directly put into use after being established, and the cold start problem of retraining is needed through a large amount of data, the embodiment optimizes the evaluation model by using a small amount of sample data and transfer learning, selects an applicable model main network, adopts a short characteristic sample data set, and finally realizes the rapid and accurate evaluation of the battery cells, the battery modules, the battery clusters, the battery stacks and the battery cabins SOH of the energy storage power station through the small-scale short characteristic sample data set on the basis of a source domain model of the battery cells generated in a laboratory, thereby supporting the accurate evaluation of the battery SOH of the electrochemical energy storage power station for cold start. Furthermore, when the transfer learning model taking GRU and the like as the main network is used for evaluating the SOH of the battery pack, the transfer learning model meets the requirement of the operation detection of the battery in the energy storage power station on timeliness due to short training and prediction time, and is suitable for deployment in a scene of a large-scale electrochemical energy storage power station. In addition, the smoothing process is used for removing noise of the data, so that the accuracy of the evaluation model can be effectively improved. The effect of cutting the data set into two short feature sample sets based on voltage: the operation is more convenient, and the problem of repeated calibration of the SOC caused by capacity degradation of the battery is avoided by taking voltage instead of the SOC (State of Charge) as a cutting basis. Cutting into short feature sample sets can shorten training and evaluation time of the model and improve evaluation efficiency.
Example 2
As shown in fig. 4, the embodiment provides an evaluation device for the battery state of health SOH of an energy storage power station, which includes a construction module 41, a pre-training module 42, a retraining module 43 and an evaluation module 44.
A construction module 41 for constructing an SOH assessment model based on the neural network.
The pre-training module 42 is connected with the construction module 41, and is configured to obtain source domain data, and train the SOH evaluation model with the source domain data to obtain a source domain model.
The retraining module 43 is connected to the pre-training module 42, and is configured to obtain a preset number of target domain data, retrain the source domain model with the preset number of target domain data, and obtain a shift learning model for SOH evaluation.
The evaluation module 44 is connected with the retraining module 43 and is used for evaluating the health state of the target battery of the energy storage power station based on the SOH evaluation transfer learning model.
Optionally, the target battery includes any one of a battery cell, a battery module, a battery cluster, a battery stack, and a battery compartment.
The pre-training module is specifically configured to obtain capacity increment data sets of a plurality of battery monomers in a laboratory in a discharging process respectively, so as to obtain source domain data, where the capacity increment satisfies the following formula:
and->The discharge capacity at two adjacent moments, respectively +.>And->And the terminal voltages of the batteries at two adjacent moments are respectively used for inputting source domain data as a model, and training and testing an SOH evaluation model.
Optionally, the pre-training module is further configured to smooth the capacity increment data sets of the plurality of battery cells in the laboratory during discharging, and cut the smoothed capacity increment data sets into two first short feature sample sets according to a voltage of the following formula:
wherein,and b is the discharge cut-off voltage of the laboratory battery cell, and b is the charge cut-off voltage of the laboratory battery cell.
Optionally, the pre-training module is configured to obtain a capacity increment data set of each of the plurality of battery cells in the laboratory that meets the following preset conditions during discharging. The preset condition comprises that the voltage amplitude is larger than a first threshold value, and the unit change amplitude of the voltage amplitude is larger than a second threshold value.
Optionally, the retraining module is specifically configured to obtain a capacity increment data set of the target battery of the energy storage power station in a discharging process respectively, obtain a preset number of target domain data, and input the preset number of target domain data as a source domain model, retrain and retest the source domain model.
Optionally, the retraining module is configured to smooth the capacity increment data sets of the target battery of the energy storage power station during the discharging process, and cut the smoothed capacity increment data sets into two second short feature sample sets according to the voltage of the following formula:
wherein,Ato discharge cut-off voltage of the energy storage power station target battery,Band (5) charging cut-off voltage of the target battery of the energy storage power station.
Optionally, the construction module is specifically configured to construct an SOH estimation model based on the long-short-term memory network LSTM and the gate control loop unit GRU, determine the number of layers, the number of neurons in each layer, and an activation function of the SOH estimation model, and initialize parameters of the SOH estimation model.
Example 3
The embodiment provides a method for managing an energy storage power station battery, which comprises the following steps:
step 301, obtaining the state of health of the target battery of the energy storage power station according to the method for evaluating the state of health SOH of the battery of the energy storage power station described in embodiment 1.
Step 302, optimizing a usage policy of the target battery or predicting a failure of the target battery according to the state of health.
It is to be understood that the above embodiments are merely illustrative of the application of the principles of the present invention, but not in limitation thereof. Various modifications and improvements may be made by those skilled in the art without departing from the spirit and substance of the invention, and are also considered to be within the scope of the invention.
Claims (11)
1. The method for evaluating the SOH of the battery state of the energy storage power station is characterized by comprising the following steps of:
constructing an SOH evaluation model based on a neural network;
acquiring source domain data, training an SOH evaluation model by the source domain data, and obtaining a source domain model;
acquiring a preset number of target domain data, retraining a source domain model by the preset number of target domain data, and obtaining a transfer learning model for SOH evaluation;
the state of health of the target battery of the energy storage power station is estimated based on a transfer learning model of SOH estimation,
the obtaining the preset number of target domain data and retraining the source domain model with the preset number of target domain data specifically includes:
acquiring a capacity increment data set of a target battery of an energy storage power station in a discharging process respectively to obtain target domain data of a preset number;
inputting a preset number of target domain data as a source domain model, retraining and retesting the source domain model, wherein the target battery comprises any one of a battery cell, a battery module, a battery cluster, a battery stack and a battery cabin,
the method for acquiring the source domain data specifically comprises the following steps:
acquiring capacity increment data sets of a plurality of battery monomers in a laboratory in a discharging process respectively;
smoothing capacity increment data sets of a plurality of battery monomers in a laboratory in a discharging process respectively;
cutting the capacity increment data set after smoothing into two first short characteristic sample sets according to the voltage of the following formula to obtain source domain data:
wherein,ato the discharge cut-off voltage of the laboratory cell,bfor the charge cut-off voltage of laboratory battery monomer, acquire the capacity increment dataset of a plurality of battery monomers of laboratory respectively in discharge process, specifically include:
acquiring capacity increment data sets of a plurality of battery monomers in a laboratory, wherein the capacity increment data sets respectively meet the following preset conditions in the discharging process,
the preset condition comprises that the voltage amplitude is larger than a first threshold value, and the unit change amplitude of the voltage amplitude is larger than a second threshold value, wherein the first threshold value is 3 volts, and the second threshold value is 0.3 millivolts.
2. The method of claim 1, wherein the step of determining the position of the substrate comprises,
the capacity increment satisfies the following formula:
，
and->The discharge capacity at two adjacent moments is respectively, the unit of the discharge capacity is ampere hour, and the unit of the discharge capacity is +.>And->The terminal voltages of the batteries at two adjacent moments are respectively in the unit of volt,
the training of the SOH evaluation model by the source domain data specifically comprises the following steps:
and (5) inputting the source domain data as a model, and training and testing an SOH evaluation model.
3. The method of claim 2, further comprising, after the acquiring the incremental data sets of capacity of the energy storage plant target battery during discharge, respectively, and before obtaining the predetermined number of target domain data:
smoothing capacity increment data sets of the target battery of the energy storage power station in the discharging process respectively;
cutting the smoothed capacity increment data set into two second short characteristic sample sets according to the voltage of the following formula:
wherein,Ato discharge cut-off voltage of the energy storage power station target battery,Band (5) charging cut-off voltage of the target battery of the energy storage power station.
4. The method according to claim 1, wherein the constructing the SOH assessment model based on the neural network specifically comprises:
constructing an SOH evaluation model based on the long-short-term memory network LSTM and the gating circulating unit GRU;
determining the number of layers of the SOH evaluation model, the number of neurons in each layer and an activation function;
initializing parameters of the SOH evaluation model.
5. The method of claim 1, further comprising, prior to said acquiring the incremental capacity dataset of the laboratory plurality of cells during discharge, respectively:
the process step of the laboratory is set as a constant current charging and discharging process of 0.8C, and corresponding charging and discharging currents are set according to the difference of the capacities of the process step and the laboratory battery cell.
6. The method of claim 1, further comprising, prior to the acquiring the incremental data set of capacity of the energy storage plant target battery during discharge, respectively:
if the target battery is a battery pack, the standing time in the charging and discharging process of the battery pack is prolonged by 10 minutes.
7. The evaluation device for the SOH of the battery state of the energy storage power station is characterized by comprising a construction module, a pre-training module, a retraining module and an evaluation module,
a construction module for constructing an SOH evaluation model based on the neural network,
the pre-training module is connected with the construction module and used for acquiring source domain data, training an SOH evaluation model with the source domain data to obtain a source domain model,
the retraining module is connected with the pre-training module and used for acquiring the target domain data of the preset quantity, retraining the source domain model of the target domain data of the preset quantity to obtain a migration learning model of SOH evaluation,
the evaluation module is connected with the retraining module and is used for evaluating the health state of the target battery of the energy storage power station based on the SOH evaluation transfer learning model,
the retraining module is specifically used for acquiring capacity increment data sets of target batteries of the energy storage power station in the discharging process respectively to obtain preset number of target domain data, inputting the preset number of target domain data as a source domain model, retraining and retesting the source domain model, wherein the target batteries comprise any one of battery monomers, battery modules, battery clusters, battery stacks and battery cabins,
the pre-training module is specifically used for acquiring capacity increment data sets of a plurality of battery monomers in a laboratory in the discharging process respectively to obtain source domain data,
the pre-training module is further used for performing smoothing processing on capacity increment data sets of a plurality of battery cells in a laboratory in a discharging process respectively, and cutting the smoothed capacity increment data sets into two first short characteristic sample sets according to the voltage of the following formula:
wherein a is the discharge cut-off voltage of the laboratory battery cell, b is the charge cut-off voltage of the laboratory battery cell,
the pre-training module is further used for acquiring capacity increment data sets of a plurality of battery monomers in a laboratory, wherein the capacity increment data sets meet the following preset conditions in the discharging process, the preset conditions comprise that the voltage amplitude is larger than a first threshold value, and the unit change amplitude of the voltage amplitude is larger than a second threshold value, wherein the first threshold value is 3 volts, and the second threshold value is 0.3 millivolts.
8. The apparatus of claim 7, wherein the device comprises a plurality of sensors,
the capacity increment satisfies the following formula:，
and->The discharge capacity at two adjacent moments is respectively, the unit of the discharge capacity is ampere hour, and the unit of the discharge capacity is +.>And->The terminal voltages of the batteries at two adjacent moments are respectively in the unit of volt,
the pre-training module is specifically used for inputting source domain data as a model, training and testing an SOH evaluation model.
9. The apparatus of claim 7, wherein the pre-training module is further configured to set a laboratory step to a constant current charge-discharge process of 0.8C, and to set a corresponding charge-discharge current according to a difference between the laboratory step and a capacity of a laboratory cell itself.
10. The apparatus of claim 7, wherein the retraining module is further configured to extend the rest time in the battery pack charge-discharge flow by 10 minutes when the target battery is a battery pack.
11. The management method of the energy storage power station battery is characterized by comprising the following steps of:
the method for evaluating the state of health SOH of the energy storage power station battery according to any one of claims 1-6, to obtain the state of health of the energy storage power station target battery;
optimizing a usage policy of the target battery or predicting a failure of the target battery according to the state of health.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311309810.0A CN117054892B (en) | 2023-10-11 | 2023-10-11 | Evaluation method, device and management method for battery state of energy storage power station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311309810.0A CN117054892B (en) | 2023-10-11 | 2023-10-11 | Evaluation method, device and management method for battery state of energy storage power station |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117054892A CN117054892A (en) | 2023-11-14 |
| CN117054892B true CN117054892B (en) | 2024-02-27 |
Family
ID=88667766
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311309810.0A Active CN117054892B (en) | 2023-10-11 | 2023-10-11 | Evaluation method, device and management method for battery state of energy storage power station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117054892B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117669272B (en) * | 2024-02-01 | 2024-04-19 | 华北电力大学 | Modeling method, system and equipment for multi-state mixing precision of battery energy storage container |
Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111638465A (en) * | 2020-05-29 | 2020-09-08 | 浙大宁波理工学院 | Lithium battery health state estimation method based on convolutional neural network and transfer learning |
| WO2020191800A1 (en) * | 2019-03-27 | 2020-10-01 | 东北大学 | Method for predicting remaining service life of lithium-ion battery employing wde-optimized lstm network |
| CN112834945A (en) * | 2020-12-31 | 2021-05-25 | 东软睿驰汽车技术（沈阳）有限公司 | Evaluation model establishing method, battery health state evaluation method and related product |
| WO2021208309A1 (en) * | 2020-04-17 | 2021-10-21 | 许继集团有限公司 | Method and system for online evaluation of electrochemical cell of energy storage power station |
| CN113962154A (en) * | 2021-10-22 | 2022-01-21 | 山东大学 | Lithium battery nuclear temperature evaluation method and system based on transfer learning |
| CN114167301A (en) * | 2021-11-30 | 2022-03-11 | 同济大学 | Power battery evaluation method based on real vehicle data of electric vehicle |
| CN114839540A (en) * | 2022-05-17 | 2022-08-02 | 西安交通大学 | Deep learning algorithm-based lithium ion battery residual life online prediction method |
| CN114859249A (en) * | 2022-07-06 | 2022-08-05 | 苏州清研精准汽车科技有限公司 | Method and device for detecting battery pack capacity |
| CN115015760A (en) * | 2022-05-10 | 2022-09-06 | 香港中文大学（深圳） | Lithium battery health state evaluation method based on neural network and migration integrated learning |
| CN115166561A (en) * | 2022-07-12 | 2022-10-11 | 西南石油大学 | Lithium battery life prediction method based on CNN-GRU combined neural network |
| CN115219937A (en) * | 2022-07-20 | 2022-10-21 | 重庆理工大学 | Method for estimating health states of energy storage batteries with different aging paths based on deep learning |
| CN115389946A (en) * | 2022-09-20 | 2022-11-25 | 金陵科技学院 | Lithium battery health state estimation method based on isobaric rise energy and improved GRU |
| WO2023027048A1 (en) * | 2021-08-24 | 2023-03-02 | 恒林日本株式会社 | Battery parameter estimation method, apparatus, device, and medium |
| DE102022211063A1 (en) * | 2021-10-27 | 2023-04-27 | Robert Bosch Engineering And Business Solutions Private Limited | Controller for estimating characteristic parameters of a battery and method therefor |
| WO2023130776A1 (en) * | 2022-01-07 | 2023-07-13 | 国网浙江省电力有限公司电力科学研究院 | Method and system for predicting working condition health status of battery in energy storage power station |
| CN116540134A (en) * | 2023-05-06 | 2023-08-04 | 重庆邮电大学 | GWO-LSTM-based lithium ion battery health state estimation method |
| CN116609673A (en) * | 2023-05-22 | 2023-08-18 | 重庆大学 | Battery remaining life prediction method and device based on charging characteristics and transfer learning |
| CN116679232A (en) * | 2023-06-08 | 2023-09-01 | 重庆大学 | Lithium battery health state estimation method based on hybrid neural network and transfer learning |
| CN116819372A (en) * | 2023-05-30 | 2023-09-29 | 单则明 | Method for calculating health state of energy storage battery by neural network |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102221756B1 (en) * | 2014-07-18 | 2021-03-02 | 삼성전자주식회사 | Method and apparatus for estimating state of battery |
| KR102574257B1 (en) * | 2015-10-30 | 2023-09-01 | 삼성전자주식회사 | Apparatus and method for estimating state of health for battery, apparatus and method for generating model for estimating state of health for battery |
| DE102020215297A1 (en) * | 2020-12-03 | 2022-06-09 | Robert Bosch Gesellschaft mit beschränkter Haftung | Method and device for operating a system for providing predicted aging states of electrical energy stores for a device using machine learning methods |
-
2023
- 2023-10-11 CN CN202311309810.0A patent/CN117054892B/en active Active
Patent Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020191800A1 (en) * | 2019-03-27 | 2020-10-01 | 东北大学 | Method for predicting remaining service life of lithium-ion battery employing wde-optimized lstm network |
| WO2021208309A1 (en) * | 2020-04-17 | 2021-10-21 | 许继集团有限公司 | Method and system for online evaluation of electrochemical cell of energy storage power station |
| CN111638465A (en) * | 2020-05-29 | 2020-09-08 | 浙大宁波理工学院 | Lithium battery health state estimation method based on convolutional neural network and transfer learning |
| CN112834945A (en) * | 2020-12-31 | 2021-05-25 | 东软睿驰汽车技术（沈阳）有限公司 | Evaluation model establishing method, battery health state evaluation method and related product |
| WO2023027048A1 (en) * | 2021-08-24 | 2023-03-02 | 恒林日本株式会社 | Battery parameter estimation method, apparatus, device, and medium |
| CN113962154A (en) * | 2021-10-22 | 2022-01-21 | 山东大学 | Lithium battery nuclear temperature evaluation method and system based on transfer learning |
| DE102022211063A1 (en) * | 2021-10-27 | 2023-04-27 | Robert Bosch Engineering And Business Solutions Private Limited | Controller for estimating characteristic parameters of a battery and method therefor |
| CN114167301A (en) * | 2021-11-30 | 2022-03-11 | 同济大学 | Power battery evaluation method based on real vehicle data of electric vehicle |
| WO2023130776A1 (en) * | 2022-01-07 | 2023-07-13 | 国网浙江省电力有限公司电力科学研究院 | Method and system for predicting working condition health status of battery in energy storage power station |
| CN115015760A (en) * | 2022-05-10 | 2022-09-06 | 香港中文大学（深圳） | Lithium battery health state evaluation method based on neural network and migration integrated learning |
| CN114839540A (en) * | 2022-05-17 | 2022-08-02 | 西安交通大学 | Deep learning algorithm-based lithium ion battery residual life online prediction method |
| CN114859249A (en) * | 2022-07-06 | 2022-08-05 | 苏州清研精准汽车科技有限公司 | Method and device for detecting battery pack capacity |
| CN115166561A (en) * | 2022-07-12 | 2022-10-11 | 西南石油大学 | Lithium battery life prediction method based on CNN-GRU combined neural network |
| CN115219937A (en) * | 2022-07-20 | 2022-10-21 | 重庆理工大学 | Method for estimating health states of energy storage batteries with different aging paths based on deep learning |
| CN115389946A (en) * | 2022-09-20 | 2022-11-25 | 金陵科技学院 | Lithium battery health state estimation method based on isobaric rise energy and improved GRU |
| CN116540134A (en) * | 2023-05-06 | 2023-08-04 | 重庆邮电大学 | GWO-LSTM-based lithium ion battery health state estimation method |
| CN116609673A (en) * | 2023-05-22 | 2023-08-18 | 重庆大学 | Battery remaining life prediction method and device based on charging characteristics and transfer learning |
| CN116819372A (en) * | 2023-05-30 | 2023-09-29 | 单则明 | Method for calculating health state of energy storage battery by neural network |
| CN116679232A (en) * | 2023-06-08 | 2023-09-01 | 重庆大学 | Lithium battery health state estimation method based on hybrid neural network and transfer learning |
Non-Patent Citations (2)
| Title |
|---|
| Multiobjective Optimization of Data-Driven Model for Lithium-Ion Battery SOH Estimation With Short-Term Feature;Lei Cai等;《IEEE TRANSAC TIONS ON POWER ELECTRONICS》;第35卷(第11期);全文 * |
| 基于迁移学习的锂电池SOH估算实验设计;郑文斌等;《实验技术与管理》;第39卷(第8期);文第57-58页以及图3 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117054892A (en) | 2023-11-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8306781B2 (en) | Professional diagnosis method of battery performance analysis | |
| CN109725266B (en) | Method and device for calculating SOH (state of health) of battery | |
| CN110161425B (en) | Method for predicting remaining service life based on lithium battery degradation stage division | |
| EP2581754B1 (en) | Battery adaptive learning management system | |
| US20170115355A1 (en) | Maximum capacity estimator for battery state of health and state of charge determinations | |
| CN110398697B (en) | Lithium ion health state estimation method based on charging process | |
| CN111190111B (en) | Electrochemical energy storage battery state of charge estimation method, device and system | |
| CN112051511A (en) | Power battery state of health estimation method and system based on multichannel technology | |
| CN114371409B (en) | Training method of battery state prediction model, battery state prediction method and device | |
| CN113052464B (en) | Method and system for evaluating reliability of battery energy storage system | |
| CN117054892B (en) | Evaluation method, device and management method for battery state of energy storage power station | |
| 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 | |
| JPWO2014119328A1 (en) | Battery state estimation device | |
| Wei et al. | Lyapunov-based state of charge diagnosis and health prognosis for lithium-ion batteries | |
| CN107037375B (en) | Method and device for measuring direct current internal resistance of battery | |
| Jiang et al. | An aging-aware soc estimation method for lithium-ion batteries using xgboost algorithm | |
| Che et al. | Battery health prognostic with sensor-free differential temperature voltammetry reconstruction and capacity estimation based on multi-domain adaptation | |
| CN111537888A (en) | Data-driven echelon battery SOC prediction method | |
| CN115480180A (en) | New energy battery health diagnosis and analysis method | |
| De Sousa et al. | Comparison of different approaches for lead acid battery state of health estimation based on artificial neural networks algorithms | |
| Savargaonkar et al. | A novel neural network with gaussian process feedback for modeling the state-of-charge of battery cells | |
| Luiz et al. | Representing the accumulator ageing in an automotive lead-acid battery model | |
| Chen et al. | State of health estimation of lithium-ion batteries based on equivalent circuit model and data-driven method | |
| CN116819364A (en) | Battery health determination method, application method, electronic device and storage medium | |
| Alvarez-Monteserin et al. | An online fade capacity estimation of lithium-ion battery using a new health indicator based only on a short period of the charging voltage profile |
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 |