CN111814826A - Rapid detection and rating method for residual energy of retired power battery - Google Patents

Rapid detection and rating method for residual energy of retired power battery Download PDF

Info

Publication number
CN111814826A
CN111814826A CN202010512176.0A CN202010512176A CN111814826A CN 111814826 A CN111814826 A CN 111814826A CN 202010512176 A CN202010512176 A CN 202010512176A CN 111814826 A CN111814826 A CN 111814826A
Authority
CN
China
Prior art keywords
mpsobp
power battery
retired power
battery
residual energy
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.)
Granted
Application number
CN202010512176.0A
Other languages
Chinese (zh)
Other versions
CN111814826B (en
Inventor
张梦雅
章尧
王家宁
戴金山
马燕萍
朱涌泉
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.)
Wuhan University of Technology WUT
Original Assignee
Wuhan University of Technology WUT
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 Wuhan University of Technology WUT filed Critical Wuhan University of Technology WUT
Priority to CN202010512176.0A priority Critical patent/CN111814826B/en
Publication of CN111814826A publication Critical patent/CN111814826A/en
Application granted granted Critical
Publication of CN111814826B publication Critical patent/CN111814826B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • 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
    • G06N3/08—Learning methods
    • G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02E60/10—Energy storage using batteries

Abstract

The invention discloses a rapid detection and rating method for the residual energy of a retired power battery, which is characterized in that an MPSOBP model is constructed based on the prediction of the residual energy of the retired power battery, and initial parameters are set according to the evaluation requirement of the residual energy of the battery; acquiring a training sample and a test sample required by MPSOBP model training; training an MPSOBP model by using a training sample and detecting a prediction effect by using a test sample to obtain the MPSOBP model meeting the precision requirement; establishing an MPSOBP-BP composite neural network by using an MPSOBP model, predicting battery capacity measurement data of a retired power battery to be tested, and evaluating the residual energy of the retired power battery; inputting the battery capacity measurement data and the measured battery structure condition data into a combined K-Means clustering algorithm to grade the retired power battery; and outputting the complementary energy evaluation result and the rating result. The invention can rapidly and accurately detect and grade the residual energy of the retired power battery, and the measurement time is within 15 min.

Description

Rapid detection and rating method for residual energy of retired power battery
Technical Field
The invention belongs to the field of battery detection, and particularly relates to a method for rapidly detecting and grading the residual energy of a retired power battery.
Background
The average service life of the power battery is 5-8 years, the performance of the power battery is attenuated along with the increase of charging times, and when the capacity of the power battery is attenuated to be below 80% of the rated capacity, the power battery is not suitable for electric automobiles any more. But the retired battery can still be further utilized in a plurality of fields such as energy storage, distributed photovoltaic power generation, household power consumption, low-speed electric vehicles and the like in a gradient way through links such as detection, maintenance and recombination. Before these retired batteries are utilized in a cascading manner, in order to ensure safe use and optimal performance of the retired batteries, the retired battery module needs to be subjected to complementary energy detection.
According to the standard of 'vehicle power battery recycling complementary energy detection' (GB/T34015-2017) issued by the Chinese national standard committee, the capacity calibration needs to be carried out for 3-5 times after the electric core or the battery module is retired, and the experiment can be finished when the capacity range difference of 3 continuous times is less than 3% of the rated capacity. Therefore, the time that a detection instrument needs to be occupied in the complementary energy detection process of one battery cell or battery module under the greenhouse is 36-60 hours, and the problems that the time cost of complementary energy detection is high, the requirement on the detection instrument is high and the like are undoubtedly caused.
Disclosure of Invention
The invention aims to provide a method for quickly detecting and grading the residual energy of a retired power battery, which can quickly and accurately detect and grade the residual energy of the retired power battery, wherein the measurement time is within 15 min.
The technical scheme adopted by the invention is as follows:
a rapid detection and rating method for the residual energy of a retired power battery comprises the following steps:
step 1, constructing an MPSOBP model based on retired power battery residual energy prediction, and setting initial parameters according to battery residual energy evaluation requirements;
step 2, obtaining a training sample and a test sample required by MPSOBP model training;
step 3, training an MPSOBP model by using the training samples and detecting the prediction effect by using the test samples to obtain the MPSOBP model meeting the precision requirement;
step 4, establishing an MPSOBP-BP composite neural network by using an MPSOBP model, predicting battery capacity measurement data of the retired power battery to be tested, and evaluating the residual energy of the retired power battery;
step 5, inputting the battery capacity measurement data and the measured battery structure condition data into a combined K-Means clustering algorithm to grade the retired power battery;
and 6, outputting the complementary energy evaluation result and the rating result.
The battery capacity measurement data includes charge temperature curve data, charge curve data, discharge temperature curve data, and discharge curve data.
In step 1, the construction of the MPSOBP model is mainly divided into the following steps.
1) Defining a neural network
An MPSOBP model is established according to actual problems, the number of nodes of an input layer, the number of nodes of a hidden layer and the number of nodes of an output layer are Q, R, T respectively, and the connection weight between the input layer and the hidden layer is A ═ A (A ═ A-1,A2,...,AQ) Wherein the connection weight between the ith node of the input layer and the hidden layer is Ai＝(a1,a2,...,aR) The connection weight between the hidden layer and the output layer is B ═ B1,B2,..,.BR) Wherein the connection weight between the ith node of the hidden layer and the output layer is Bi＝(b1,b2,...,bT) The hidden layer threshold is C ═ C1,c2,...,cR) Output layer threshold value of Di＝(d1,d2,...,dT) Training precision, coarse, maximum training frequency, epoch.
2) Incorporating improved PSO algorithms
Firstly, initializing PSO algorithm
Defining an S-dimensional search space, wherein S is QR + RT + R + T, defining the size of a population, the number of iterations as m and n respectively, and randomly assigning W (W) to the position of an individual in the population1,W2,...,Wn) Wherein the location vector W of the ith individuali＝(A,B,C,D)TEach individual position vector represents a group of solutions of BP neural network parameters, and the velocity of each individual is randomly assigned as V ═ V (V)1,V2,...,Vn) Wherein the velocity vector of the i-th individual is Vi＝(Vi1,Vi2,...,ViS)TThe individual extremum is denoted as P ═ P1,P2,..,.Pn) Extreme value of the ith individual is Pi＝(Pi1,Pi2,...,PiS)TGroup extremum Pg＝(Pg1,Pg2,...,PgS)T. A maximum number of iterations epochs is defined.
Determining an evaluation function and calculating the fitness
Selecting a proper individual evaluation function, and setting a neural network training output value V ═ V1,v2,...,vT) Training desired output value E ═ (E)1,e2,...,eT) Then individual W in the population WiIs defined as
Figure BDA0002528720700000021
Where n is the population size.
Thirdly, individual W in the population W obtained by each iterationiAssigning values to the connection weight and the threshold of the BP neural network, inputting training samples to train the neural network, stopping training when the training precision or the maximum training times is reached, and obtaining the output value V ═ of the BP neural network training1,v2,...,vT) And calculating the fitness according to a fitness calculation formula.
Fourthly, calculating individual extreme value and global extreme value
Calculating each individual W according to the input and output dataiFitness is combined with the fitness of the individual obtained by the last calculation to determine an individual extreme value Pi＝(Pi1,Pi2,...,PiS)TAnd a global extremum Pg＝(Pg1,Pg2,...,PgS)TAnd according toAnd updating the individual speed and position by the body extreme value and the global extreme value, wherein the updating model comprises the following steps:
Figure BDA0002528720700000022
Figure BDA0002528720700000023
wherein c is1、c2The value of the acceleration factor is 0-4, r1、r2Is distributed in [0,1 ]]The superscript k denotes the result of the kth iteration, PidIs the d-th dimension, P, of the individual extremum of the i-th variablegdThe d-th dimension of the global optimal solution.
Introducing mutation operator in genetic algorithm
Setting the probability of variation mu1、mu2Randomly selecting m.mu among m individuals1For the selected individuals, S & mu in the position vector is randomly selected2Element, re-initializing it. Inputting the data processed by the mutation operator into a BP neural network model
Sixthly, repeating the steps, and after the maximum iteration number is reached, assigning the value to the MPSOBP model according to the obtained individual position, and training the MPSOBP model.
In step 2, dividing the retired power battery into a training sample and a testing sample, performing charging and discharging for many times, and collecting battery capacity measurement data.
Preferably, the charge and discharge are performed in a cycle of 3 minutes.
In step 3, the training sample adopts complete battery capacity measurement data, the test sample adopts charge and discharge segments in the battery capacity measurement data, and the overall average error is used for precision evaluation to obtain an MPSOBP model when the precision requirement is met.
In step 4, the first layer of the MPSOBP-BP composite neural network is an MPSOBP model, the second layer of the MPSOBP-BP composite neural network is a BP neural network, the retired power battery to be tested is charged and discharged for several times in a period of a certain time, the charging and discharging segments in the battery capacity measuring data are input into the MPSOBP model and output to obtain complete battery capacity measuring data, then the battery capacity measuring data are input into the BP neural network, the hidden layer and the related threshold are simply initialized as well, the hidden layer and the related threshold are adjusted through an MPSO algorithm, and the evaluation result of the retired power battery is output.
Preferably, the retired power battery to be tested is charged and discharged for 2 times in a cycle of 3 minutes.
In step 5, the rating result obtained by the K-Means clustering analysis method is detected through a training battery, if the actual result is detected to indicate that the clustered classification is reasonable, the clustering is finished, otherwise, the clustering is re-clustered through manual intervention.
The invention has the beneficial effects that:
the invention can rapidly and accurately detect and grade the residual energy of the retired power battery, and the measurement time is within 15 min.
Drawings
FIG. 1 is a flow chart of steps in an embodiment of the present invention.
Fig. 2 is a flow chart of an algorithm in an embodiment of the invention.
Detailed Description
The invention is further described below with reference to the figures and examples.
Selecting the retired lithium battery complementary energy detection as a task, and taking charging temperature curve data, charging curve data, discharging temperature curve data and discharging curve data as battery capacity measurement data.
A rapid detection and rating method for the residual energy of a retired power battery comprises the following steps:
step 1, constructing an MPSOBP model based on retired power battery residual energy prediction, and setting initial parameters according to battery residual energy evaluation requirements:
initializing neural network
Establishing a BP neural network model according to analysis, wherein the number of nodes of an input layer, the number of nodes of an implicit layer and the number of nodes of an output layer are Q-8, R-17 and T-8 respectively, and the connection weight between the input layer and the implicit layer is A-A (A)1,A2,...,AQ) Wherein the connection weight between the ith node of the input layer and the hidden layer is Ai＝(a1,a2,...,aR) The connection weight between the hidden layer and the output layer is B ═ B1,B2,...,BR) Wherein the connection weight between the ith node of the hidden layer and the output layer is Bi＝(b1,b2,...,bT) The hidden layer threshold is C ═ C1,c2,...,cR) Output layer threshold value of Di＝(d1,d2,...,dT) Training precision, coarse, maximum training frequency, epoch.
② combining improved PSO algorithm
Defining an S-dimensional search space, wherein S is QR + RT + R + T, defining the size of a population, the number of iterations is m-30, n-100 respectively, and randomly assigning W-W (W) to the positions of individuals in the population1,W2,...,Wn) Wherein the location vector W of the ith individuali＝(A,B,C,D)TEach individual position vector represents a group of solutions of BP neural network parameters, and the velocity of each individual is randomly assigned as V ═ V (V)1,V2,...,Vn) Wherein the velocity vector of the i-th individual is Vi＝(Vi1,Vi2,...,ViS)TThe individual extremum is denoted as P ═ P1,P2,...,Pn) Extreme value of the ith individual is Pi＝(Pi1,Pi2,...,PiS)TGroup extremum Pg＝(Pg1,Pg2,...,PgS)T. A maximum number of iterations epochs is defined.
Selecting a proper individual evaluation function, and setting a neural network training output value V ═ V1) Training desired output value E ═ (E)1) Then individual W in the population WiIs defined as
Figure BDA0002528720700000041
Where n is the population size.
The individuals W in the population W obtained by each iterationiAssigning the connection weight and the threshold value of the BP neural network, inputting a training sample to train the neural network, and achieving the given training precisionOr stopping training when the training times are maximum, and obtaining the output value V ═ of BP neural network training1) And calculating the fitness according to a fitness calculation formula.
Calculating each individual W according to the input and output dataiFitness is combined with the fitness of the individual obtained by the last calculation to determine an individual extreme value Pi＝(Pi1,Pi2,...,PiS)TAnd a global extremum Pg＝(Pg1,Pg2,...,PgS)TAnd updating the individual speed and position according to the individual extreme value and the global extreme value, wherein the updating model is as follows:
Figure BDA0002528720700000042
Figure BDA0002528720700000043
wherein c is1、c2The value of the acceleration factor is 0-4, r1、r2Is distributed in [0,1 ]]The superscript k denotes the result of the kth iteration, PidIs the d-th dimension, P, of the individual extremum of the i-th variablegdThe d-th dimension of the global optimal solution.
Setting the probability of variation m mu1＝0.01、mμ10.02, m · m μ was randomly selected from 30 individuals1For the selected individuals, the S.m μm in the position vector is randomly selected1Element, re-initializing it. And inputting the data processed by the mutation operator into an MPSOBP model.
Step 2, obtaining training samples and test samples required by MPSOBP model training:
selecting a power battery pack which is eliminated and disassembled after an electric vehicle of a certain vehicle is used, testing by using a new power BTS-5V6A battery cell capacity test cabinet, charging a lithium battery at a constant current of 2A under the condition of 20 +/-5 ℃ until the voltage of the lithium battery reaches 4.2V, the terminating current is 0.1A, standing for 1h, then discharging at 1C (namely current 2A) under the condition of 20 +/-5 ℃, standing for 1h again when the terminating voltage is set to be 2.75V, and repeating the cycle for 3 times. In the process, battery charging temperature-time data, charging current-voltage-time data, battery discharging temperature-time data and discharging current-voltage-time data are collected, and the remaining energy of the battery is calculated. According to the measured and calculated actual retired battery residual capacity, the collected data are selected from the data of the lithium battery with the residual capacity of more than 60% to form a total sample pool, and a training set and a testing set are randomly classified.
Respectively carrying out normalization processing on battery charging temperature-time data, charging current-voltage-time data, battery discharging temperature-time data and discharging current-voltage-time data of the lithium battery, then inputting a training set into an MPSOBP model, training the MPSOBP model, and randomly intercepting data for 3 minutes from the data of a test set to predict an overall curve and test the precision. And outputting the trained MPSOBP model after the overall accuracy reaches an expected value.
Step 3, training the MPSOBP model by using the training samples and detecting the prediction effect by using the test samples to obtain the MPSOBP model meeting the precision requirement:
the training sample adopts complete battery capacity measurement data, the test sample adopts charge and discharge segments in the battery capacity measurement data, the overall average error is used for precision evaluation, and an MPSOBP model is obtained when the precision requirement is met. The composite neural network is constructed mainly by generating a new MPSOBP neural network and combining the trained MPSOBP forming neural network. The main way is to associate the input with the output of MPSOBP, and the output is the measured value of the remaining energy of the battery, and the construction method is similar to that in step 1, and will not be described here.
Step 4, establishing an MPSOBP-BP composite neural network by using an MPSOBP model, predicting battery capacity measurement data of the retired power battery to be tested, and evaluating the residual energy of the retired power battery:
in step 4, the first layer of the MPSOBP-BP composite neural network is an MPSOBP model, the second layer of the MPSOBP-BP composite neural network is a BP neural network, the retired power battery to be tested is charged and discharged for several times in a period of a certain time, the charging and discharging segments in the battery capacity measuring data are input into the MPSOBP model and output to obtain complete battery capacity measuring data, then the battery capacity measuring data are input into the BP neural network, the hidden layer and the related threshold are simply initialized as well, the hidden layer and the related threshold are adjusted through an MPSO algorithm, and the evaluation result of the retired power battery is output. Here, the training data is based on the data of the MPSOBP model trained in step 2, and the input of the overall composite neural network is not changed from the MPSOBP model, and the main content is to change the output to the measured battery residual energy. Based on the method, a training set and a test set are reconstructed through data measured in the early stage, training of the whole composite neural network is carried out after normalization, the training is also compared with an expected target, and the trained composite neural network is output after the standard is reached.
Preferably, the retired power battery to be tested is charged and discharged for 2 times in a cycle of 3 minutes.
And 5, inputting the battery capacity measurement data and the measured battery structure condition data into a combined K-Means clustering algorithm to grade the retired power battery:
in step 5, the rating result obtained by the K-Means clustering analysis method is detected through a training battery, if the actual result is detected to indicate that the clustered classification is reasonable, the clustering is finished, otherwise, the clustering is re-clustered through manual intervention.
The process of the combined K-Means cluster analysis method mainly comprises the step of inputting output results of each layer of the composite neural network, namely battery capacity measurement data, combined with battery structure condition data and appearance integrity data, into the K-Means cluster analysis method to carry out comprehensive battery rating on the power battery. In the implementation process of the rating of the K-Means cluster analysis method, the category of rating input data is determined by performing correlation analysis according to the actual condition of each retired power battery; the grading result needs to be compared with the grading result of the traditional method, if the actual result is detected to indicate that the clustered classification is reasonable, the clustering is finished, otherwise, the clustering is carried out again through manual intervention, and the effect of the whole clustering process is optimized.
And 6, outputting the complementary energy evaluation result and the rating result.
It will be understood that modifications and variations can be made by persons skilled in the art in light of the above teachings and all such modifications and variations are intended to be included within the scope of the invention as defined in the appended claims.

Claims (9)

1. A method for rapidly detecting and grading the residual energy of a retired power battery is characterized by comprising the following steps: comprises the steps of (a) carrying out,
step 1, constructing an MPSOBP model based on retired power battery residual energy prediction, and setting initial parameters according to battery residual energy evaluation requirements;
step 2, obtaining a training sample and a test sample required by MPSOBP model training;
step 3, training an MPSOBP model by using the training samples and detecting the prediction effect by using the test samples to obtain the MPSOBP model meeting the precision requirement;
step 4, establishing an MPSOBP-BP composite neural network by using an MPSOBP model, predicting battery capacity measurement data of the retired power battery to be tested, and evaluating the residual energy of the retired power battery;
step 5, inputting the battery capacity measurement data and the measured battery structure condition data into a combined K-Means clustering algorithm to grade the retired power battery;
and 6, outputting the complementary energy evaluation result and the rating result.
2. The method for rapidly detecting and grading the retired power battery residual energy according to claim 1, wherein: the battery capacity measurement data includes charge temperature curve data, charge curve data, discharge temperature curve data, and discharge curve data.
3. The method for rapidly detecting and grading the retired power battery residual energy according to claim 1 or 2, wherein: in step 1, the MPSOBP model comprises an input layer, a multi-level hidden layer and an output layer; wherein the MPSOBP model input content comprises battery capacity measurement data; optimizing the number of layers of the multi-stage hidden layer, the number of neurons in each layer and a related threshold value through an MPSO algorithm, initially setting the number as one hidden layer, wherein the number of the neurons is 2n +1, and n represents the number of the neurons in an input layer; the output layer includes battery capacity measurement data.
4. The method for rapidly detecting and grading the retired power battery residual energy according to claim 1 or 2, wherein: in step 2, dividing the retired power battery into a training sample and a testing sample, performing charging and discharging for many times, and collecting battery capacity measurement data.
5. The method for rapidly detecting and grading the retired power battery residual energy according to claim 4, wherein: charging and discharging were performed in a cycle of 3 minutes.
6. The method for rapidly detecting and grading the retired power battery residual energy according to claim 4, wherein: in step 3, the training sample adopts complete battery capacity measurement data, the test sample adopts charge and discharge segments in the battery capacity measurement data, and the overall average error is used for precision evaluation to obtain an MPSOBP model when the precision requirement is met.
7. The method for rapidly detecting and grading the retired power battery residual energy according to claim 1 or 2, wherein: in step 4, the first layer of the MPSOBP-BP composite neural network is an MPSOBP model, the second layer of the MPSOBP-BP composite neural network is a BP neural network, the retired power battery to be tested is charged and discharged for several times in a certain time as a period, the charging and discharging segments in the battery capacity measuring data are input into the MPSOBP model and output to obtain complete battery capacity measuring data, and then the battery capacity measuring data are input into the BP neural network and output to obtain the evaluation result of the retired power battery.
8. The method for rapidly detecting and grading the retired power battery residual energy according to claim 7, wherein: and (3) carrying out 2 times of charging and discharging on the retired power battery to be tested in a period of 3 minutes.
9. The method for rapidly detecting and grading the retired power battery residual energy according to claim 1 or 2, wherein: in step 5, the rating result obtained by the K-Means clustering analysis method is detected through a training battery, if the actual result is detected to indicate that the clustered classification is reasonable, the clustering is finished, otherwise, the clustering is re-clustered through manual intervention.
CN202010512176.0A 2020-06-08 2020-06-08 Rapid detection and rating method for residual energy of retired power battery Active CN111814826B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010512176.0A CN111814826B (en) 2020-06-08 2020-06-08 Rapid detection and rating method for residual energy of retired power battery

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010512176.0A CN111814826B (en) 2020-06-08 2020-06-08 Rapid detection and rating method for residual energy of retired power battery

Publications (2)

Publication Number Publication Date
CN111814826A true CN111814826A (en) 2020-10-23
CN111814826B CN111814826B (en) 2022-06-03

Family

ID=72844762

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010512176.0A Active CN111814826B (en) 2020-06-08 2020-06-08 Rapid detection and rating method for residual energy of retired power battery

Country Status (1)

Country Link
CN (1) CN111814826B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113281657A (en) * 2021-05-21 2021-08-20 张家港清研检测技术有限公司 Intelligent assessment retired battery complementary energy classification and echelon utilization method
CN114264967A (en) * 2021-12-14 2022-04-01 哈尔滨工业大学 Method and system for rapidly estimating retired battery residual energy based on capacity loss mechanism
CN114720878A (en) * 2022-03-24 2022-07-08 长安大学 Method for detecting state of retired battery

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105574586A (en) * 2015-12-29 2016-05-11 沈阳航空航天大学 General airplane air-material demand prediction method based on MPSO-BP network
CN106920008A (en) * 2017-02-28 2017-07-04 山东大学 A kind of wind power forecasting method based on Modified particle swarm optimization BP neural network
WO2018076475A1 (en) * 2016-10-26 2018-05-03 广东产品质量监督检验研究院 Photovoltaic assembly accelerated degradation model established based on deep approach of learning, and method for predicting photovoltaic assembly lifetime
WO2018126984A2 (en) * 2017-01-06 2018-07-12 江南大学 Mea-bp neural network-based wsn abnormality detection method
CN109636054A (en) * 2018-12-21 2019-04-16 电子科技大学 Solar energy power generating amount prediction technique based on classification and error combination prediction
CN110187287A (en) * 2019-06-24 2019-08-30 安徽师范大学 A kind of retired lithium battery complementary energy rapid detection method
CN110501651A (en) * 2019-08-19 2019-11-26 国网河北省电力有限公司石家庄供电分公司 Retired battery core holds detection method and device
CN110570091A (en) * 2019-08-12 2019-12-13 国网上海市电力公司 Load identification method based on improved F-score feature selection and particle swarm BP neural network
CN110687452A (en) * 2019-09-05 2020-01-14 南京理工大学 Lithium battery capacity online prediction method based on K-means clustering and Elman neural network
CN110752410A (en) * 2019-10-30 2020-02-04 上海理工大学 Method for rapidly sorting and recombining retired lithium batteries

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105574586A (en) * 2015-12-29 2016-05-11 沈阳航空航天大学 General airplane air-material demand prediction method based on MPSO-BP network
WO2018076475A1 (en) * 2016-10-26 2018-05-03 广东产品质量监督检验研究院 Photovoltaic assembly accelerated degradation model established based on deep approach of learning, and method for predicting photovoltaic assembly lifetime
WO2018126984A2 (en) * 2017-01-06 2018-07-12 江南大学 Mea-bp neural network-based wsn abnormality detection method
CN106920008A (en) * 2017-02-28 2017-07-04 山东大学 A kind of wind power forecasting method based on Modified particle swarm optimization BP neural network
CN109636054A (en) * 2018-12-21 2019-04-16 电子科技大学 Solar energy power generating amount prediction technique based on classification and error combination prediction
CN110187287A (en) * 2019-06-24 2019-08-30 安徽师范大学 A kind of retired lithium battery complementary energy rapid detection method
CN110570091A (en) * 2019-08-12 2019-12-13 国网上海市电力公司 Load identification method based on improved F-score feature selection and particle swarm BP neural network
CN110501651A (en) * 2019-08-19 2019-11-26 国网河北省电力有限公司石家庄供电分公司 Retired battery core holds detection method and device
CN110687452A (en) * 2019-09-05 2020-01-14 南京理工大学 Lithium battery capacity online prediction method based on K-means clustering and Elman neural network
CN110752410A (en) * 2019-10-30 2020-02-04 上海理工大学 Method for rapidly sorting and recombining retired lithium batteries

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
TAO LI等: "Research on Logging Evaluation of Reservoir Contamination Based on PSO-BP Neural Network", 《INTERNATIONAL SYMPOSIUM ON NEURAL NETWORKS》 *
袁鲍蕾: "交直流混合微电网的能量管理", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113281657A (en) * 2021-05-21 2021-08-20 张家港清研检测技术有限公司 Intelligent assessment retired battery complementary energy classification and echelon utilization method
CN114264967A (en) * 2021-12-14 2022-04-01 哈尔滨工业大学 Method and system for rapidly estimating retired battery residual energy based on capacity loss mechanism
CN114720878A (en) * 2022-03-24 2022-07-08 长安大学 Method for detecting state of retired battery

Also Published As

Publication number Publication date
CN111814826B (en) 2022-06-03

Similar Documents

Publication Publication Date Title
CN111814826B (en) Rapid detection and rating method for residual energy of retired power battery
CN110045298B (en) Method for diagnosing parameter inconsistency of power battery pack
CN110752410A (en) Method for rapidly sorting and recombining retired lithium batteries
CN109993270A (en) Lithium ion battery residual life prediction technique based on grey wolf pack optimization LSTM network
CN108535656A (en) Lithium ion battery remaining life prediction technique and system based on PCA-NARX neural networks
CN111999649A (en) XGboost algorithm-based lithium battery residual life prediction method
CN111366848A (en) Battery health state prediction method based on PSO-ELM algorithm
CN111027625A (en) Battery screening method based on SAE and K-means clustering algorithm
CN106778846A (en) A kind of method for forecasting based on SVMs
CN112305441B (en) Power battery health state assessment method under integrated clustering
CN108009585B (en) Lead-acid battery health state prediction method based on local information fusion
CN112287980B (en) Power battery screening method based on typical feature vector
CN113917334A (en) Battery health state estimation method based on evolution LSTM self-encoder
CN114545274A (en) Lithium battery residual life prediction method
CN113125960A (en) Vehicle-mounted lithium ion battery charge state prediction method based on random forest model
CN115219906A (en) Multi-model fusion battery state of charge prediction method and system based on GA-PSO optimization
CN115469227A (en) Set variational self-encoder and dynamic regular lithium battery abnormity detection method
CN113376541B (en) Lithium ion battery health state prediction method based on CRJ network
CN110232432B (en) Lithium battery pack SOC prediction method based on artificial life model
CN114460481A (en) Energy storage battery thermal runaway early warning method based on Bi-LSTM and attention mechanism
CN113640690A (en) Method for predicting residual life of power battery of electric vehicle
CN116613745A (en) PSO-ELM electric vehicle charging load prediction method based on variation modal decomposition
CN116736133A (en) Early prediction method for capacity degradation track of lithium ion battery in full life cycle
CN114280490B (en) Lithium ion battery state of charge estimation method and system
CN115993537A (en) Lithium battery capacity prediction method based on correlation analysis and WOA-LSTM

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