CN112234673B - Battery energy balancing method suitable for balancing circuit - Google Patents

Battery energy balancing method suitable for balancing circuit Download PDF

Info

Publication number
CN112234673B
CN112234673B CN202011065240.1A CN202011065240A CN112234673B CN 112234673 B CN112234673 B CN 112234673B CN 202011065240 A CN202011065240 A CN 202011065240A CN 112234673 B CN112234673 B CN 112234673B
Authority
CN
China
Prior art keywords
soc
battery
layer
output
neural network
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
Application number
CN202011065240.1A
Other languages
Chinese (zh)
Other versions
CN112234673A (en
Inventor
雷旭
杨越皓
唐鑫
樊临倩
禾建平
于明加
陈潇阳
陈静夷
高钊
高雪
于胜广
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.)
Changan University
Original Assignee
Changan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Changan University filed Critical Changan University
Priority to CN202011065240.1A priority Critical patent/CN112234673B/en
Publication of CN112234673A publication Critical patent/CN112234673A/en
Application granted granted Critical
Publication of CN112234673B publication Critical patent/CN112234673B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J7/00—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries
    • H02J7/0013—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries acting upon several batteries simultaneously or sequentially
    • H02J7/0014—Circuits for equalisation of charge between batteries
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
    • B60L58/00—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles
    • B60L58/10—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries
    • B60L58/12—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries responding to state of charge [SoC]
    • B60L58/14—Preventing excessive discharging
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
    • B60L58/00—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles
    • B60L58/10—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries
    • B60L58/12—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries responding to state of charge [SoC]
    • B60L58/15—Preventing overcharging
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
    • B60L58/00—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles
    • B60L58/10—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries
    • B60L58/18—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries of two or more battery modules
    • B60L58/22—Balancing the charge of battery modules
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
    • 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
    • G06N3/043—Architecture, e.g. interconnection topology based on fuzzy logic, fuzzy membership or fuzzy inference, e.g. adaptive neuro-fuzzy inference systems [ANFIS]
    • 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
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
    • H01M10/4207—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells for several batteries or cells simultaneously or sequentially
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
    • H01M10/44—Methods for charging or discharging
    • H01M10/441—Methods for charging or discharging for several batteries or cells simultaneously or sequentially
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J7/00—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries
    • H02J7/0029—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries with safety or protection devices or circuits
    • H02J7/00302—Overcharge protection
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J7/00—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries
    • H02J7/0029—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries with safety or protection devices or circuits
    • H02J7/00306—Overdischarge protection
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
    • H01M10/425—Structural combination with electronic components, e.g. electronic circuits integrated to the outside of the casing
    • H01M2010/4271—Battery management systems including electronic circuits, e.g. control of current or voltage to keep battery in healthy state, cell balancing
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M2220/00—Batteries for particular applications
    • H01M2220/20—Batteries in motive systems, e.g. vehicle, ship, plane
    • 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
    • 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
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/60—Other road transportation technologies with climate change mitigation effect
    • Y02T10/70—Energy storage systems for electromobility, e.g. batteries

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Power Engineering (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Transportation (AREA)
  • General Engineering & Computer Science (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mechanical Engineering (AREA)
  • Sustainable Energy (AREA)
  • Sustainable Development (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Biophysics (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • Electrochemistry (AREA)
  • Biomedical Technology (AREA)
  • General Chemical & Material Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Manufacturing & Machinery (AREA)
  • Chemical & Material Sciences (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Fuzzy Systems (AREA)
  • Automation & Control Theory (AREA)
  • Operations Research (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Secondary Cells (AREA)

Abstract

The invention discloses a battery energy balancing method suitable for a balancing circuit, relates to the technical field of battery energy, and provides a balancing control algorithm which adopts an AUKF algorithm to improve SOC estimation precision and solves the problem of possible overcharge and overdischarge damages to a battery when only the SOC is used as a parameter of the balancing control algorithm. The fuzzy neural network designed by the invention is a first-order T-S fuzzy neural network, 5 parameters in total of a front part parameter and a back part parameter are required to be determined through learning, the learning algorithm adopts a hybrid algorithm of a BP algorithm and a least square method, the front part parameter is determined through the BP algorithm, and the back part parameter is determined through the least square method. And training the fuzzy neural network through data in the database, and identifying to obtain all parameters.

Description

Battery energy balancing method suitable for balancing circuit
Technical Field
The invention relates to the technical field of battery energy, in particular to a battery energy balancing method suitable for a balancing circuit.
Background
In recent years, electric vehicles have been rapidly developed, and batteries for supplying energy to the electric vehicles naturally receive wide attention from researchers. The types of batteries used by electric vehicles are various, and among them, lithium batteries have become the mainstream battery type used by electric vehicles due to the advantages of good power performance, long cycle life, high energy density, no memory effect and the like. However, the parameters of the single lithium battery, such as the capacitance, the voltage and the instantaneous discharge power, are far from the parameters required by the design of the electric automobile. Therefore, to meet the requirements of electric vehicles, hundreds of single batteries are required to be combined into a battery pack in a series or parallel manner.
Unfortunately, the series-parallel connection of the unit cells creates a new problem of charge imbalance, and the charge imbalance is more pronounced in the series-connected battery pack. The imbalance of the charge amount causes the reduction of the comprehensive performance of the battery, the service life of the battery pack is greatly shortened, and the secondary explosion occurs. Researchers have found that the initial cause of the charging imbalance is due to manufacturing variability resulting in subtle cell-to-cell variability that is amplified during cell cycling. The production difference of the battery cannot be eliminated by applying the technology, and the unbalance of the battery charging can be improved only by utilizing the equalization technology in the use process of the battery.
The essence of the equalization technology is that the charged difference is eliminated to a certain extent through an equalization circuit and an equalization control algorithm. The equalization technology can be divided into a passive equalization technology and an active equalization technology according to two different ways of energy consumption and energy transfer. At present, the SOC is mostly adopted by the equalization algorithm to reflect the state of charge of the battery, and if the SOC can accurately reflect the state of charge of the battery, the performance of the control algorithm with the SOC as the only discrimination parameter can meet the actual requirement. However, when the battery is discharged in a low state, the voltage drop of the battery will increase rapidly, and when the battery is discharged with a large current, if the battery is not provided with the minimum voltage limiting protection, the battery will be in an overdischarge state for a long time, which may cause damage to the battery. Similarly, when the battery is discharged to a high state of charge, the voltage rise of the battery increases, and when the battery is charged with a large current, the battery is easily overcharged for a long time, and the battery is damaged. In addition, when the battery is severely aged, the polarization effect and the ohmic effect of the battery are more remarkable, and the voltage drop of the battery in a low-charge state and the voltage rise of a high-charge state are more obvious.
In order to solve the problems, the application provides a battery energy balancing method suitable for a balancing circuit, and solves the problem that overcharge and overdischarge damage to a battery can be caused when only SOC is used as a parameter of a balancing control algorithm.
Disclosure of Invention
The invention aims to provide a battery energy balancing method suitable for a balancing circuit, which solves the problem of possible overcharge and overdischarge damages to a battery when only SOC is used as a parameter of a balancing control algorithm.
The invention provides a battery energy balancing method suitable for a balancing circuit, which comprises the following steps:
s1: collecting the voltage and current of the equalizing circuit;
s2: establishing a second-order equivalent circuit model, and estimating the SOC by using an AUKF algorithm;
s3: calculating the difference value of the SOC of different batteries, and judging whether the conditions of one low or one high or one low exist;
s4: establishing a neural network, and if the situation described in the step S3 exists, taking the terminal voltage as the input of the neural network, and determining whether the output equalization current value reaches an equalization threshold value;
s5: if the situation described in the step S3 does not exist, calculating an average value of the SOC of each battery, and determining whether the average value is between 0.2 and 0.8, and if so, determining whether the output equalization current value reaches an equalization threshold value by using the SOC as an input of the neural network; if the voltage is not between 0.2 and 0.8, the terminal voltage is used as the input of the neural network, and whether the output equalization current value reaches the equalization threshold value is judged;
s6: and when the output equalization current value reaches an equalization threshold value, the battery energy is equalized.
Further, the second-order equivalent circuit model in step S2 describes the concentration polarization effect and the electrochemical polarization effect by using 2 RC networks, respectively, and according to KCL and KVL laws, a circuit state expression can be obtained as follows:
UO(t)＝ROI(t) (1)
Figure BDA0002713567000000031
Figure BDA0002713567000000032
wherein, UO(t) is the ohmic voltage, Ue(t) electrochemical polarization voltage, Ud(t) is concentration polarization voltage;
U(t)＝UOCV(t)+UO(t)+Ue(t)+Ud(t) (4)
wherein U (t) is the battery terminal voltage;
obtaining an expression of the change of the SOC of the battery along with time by using a current integration method:
Figure BDA0002713567000000033
therein, SOC0Is an initial SOC, QNThe standard capacitance of the battery is used, and eta is the charge-discharge efficiency of the battery;
discretization of formulae (2), (3) and (5) gives:
Figure BDA0002713567000000034
Figure BDA0002713567000000035
Figure BDA0002713567000000036
wherein, tau is a sampling period, taue＝ReCe，τd＝RdCd；
A discretized state space observation equation is established by equations (6), (7) and (8), as follows:
Figure BDA0002713567000000041
wherein w (k-1) is the observation noise of the model;
the discretized state output equation of the model is shown in equation (10):
U(k)＝UOCV(k)+UO(k)+Ue(k)+Ud(k)+v(k-1) (10)
where v (k-1) is the observation noise of the terminal voltage.
Further, the AUKF algorithm in step S2 first constructs sampling points, and the state vector X (k) is composed of state variables SOC (k), Ue(k) And Ud(k) The composition is expressed as follows:
X(k)＝[SOC(k) Ue(k) Ud(k)]T (11)
further, the step of estimating the SOC by using the AUKF algorithm in step S2 includes:
s21: initializing a nonlinear system:
Figure BDA0002713567000000042
Figure BDA0002713567000000043
s22: constructing 2n +1 sampling point sets and carrying out nonlinear conversion:
Figure BDA0002713567000000044
s23: the weight in the iterative calculation process is calculated as follows:
Figure BDA0002713567000000054
wherein,
Figure BDA0002713567000000056
for a scale parameter, α represents the sampling point at
Figure BDA0002713567000000055
Distribution of the surroundings, i.e. 1>α>e-4；
Figure BDA0002713567000000058
Generally takes on a value of
Figure BDA0002713567000000057
Beta is used to integrate X prior estimates, generally chosen to be 2; wi (m)And Wi (c)Respectively calculating the weighted factors used for the mean value and the covariance of the ith sampling point;
s24: the one-step prediction for computing the 2n +1 sample point sets is:
Xi(k-1)＝f(Xi,k-1),i＝0,1,…2n (16)
s25: the time update in the AUKF algorithm is:
Figure BDA0002713567000000051
s26: the measurements in the AUKF algorithm are updated as:
Figure BDA0002713567000000052
s27: the battery state variables and covariance are estimated as:
Figure BDA0002713567000000053
s28: adaptive law covariance matching:
Figure BDA0002713567000000061
where q (k) is the covariance of the model noise w (k), and r (k) is the covariance of the terminal voltage measurements v (k).
Further, the neural network in step S4 includes: the fuzzy rule strength normalization layer comprises a fuzzy layer, a fuzzy rule strength release layer, a rule strength normalization layer, a fuzzy rule output layer and an output layer.
Further, the input amount of the blurring layer is 2, which are the SOC difference value and the SOC average value, and can be expressed as:
Figure BDA0002713567000000062
Figure BDA0002713567000000063
therein, SOCiThe SOC value, SOC, of the ith battery in the m single batteries on the left side of the equalizing unitjFor balancing n batteries at the right side of the unitThe SOC value of the jth battery, delta SOC represents the difference of the average SOC values of the left and right sides of the equalizing unit, and SOCavgRepresents the average value of the SOC on both sides of the equalizing unit;
defining the universe of the two input quantities according to the selected battery characteristics and respectively named A and B, wherein the universe of the two input quantities delta SOC of the SOC-based fuzzy neural network is [0,0.6 ]]，SOCavgHas a discourse field of [0,1](ii) a Delta V in two input quantities of the fuzzy neural network based on terminal voltage is [0,1]，VavgIs [2.6,4.2 ]](ii) a Considering the unbalanced state and the battery characteristics of the battery pack, dividing discourse domains of two input quantities into 5 fuzzy sets which are VS, S, M, L and VL respectively, and determining the membership function number of each input quantity; the membership function types are all Gaussian functions and are expressed as follows:
Figure BDA0002713567000000064
wherein a is the center of the membership function, b is the width of the membership function, and a and b are used as the front part parameters of the fuzzy neural network and are obtained by database training and learning;
input quantities delta SOC/delta V and SOCavg/VavgAre respectively defined as x1And x2: and obtaining an input and output expression of the first layer:
Figure BDA0002713567000000071
Figure BDA0002713567000000072
wherein,
Figure BDA0002713567000000077
is an input quantity x1The ith membership function of (a),
Figure BDA0002713567000000078
is a number x2The jth membership function of (a).
Further, the fuzzy rule strength release layer multiplies output values of input quantities of the first layer by two to obtain output values of the second layer, so that 25 nodes of the layer can be determined according to the output values of the first layer, and input and output expressions of the layer are as follows:
Figure BDA0002713567000000073
wherein, ω iskRepresenting the activation strength of one of the fuzzy rules.
Further, the rule strength normalization layer calculates the proportion of the activation strength of each fuzzy rule, and the output expression is as follows:
Figure BDA0002713567000000074
further, the fuzzy rule output layer is used for calculating the output of each fuzzy rule, and the calculation expression is as follows:
Figure BDA0002713567000000075
wherein p isk、qkAnd rkThe parameters are acquired through training as the back-piece parameters of the fuzzy neural network.
Further, the output layer is used for calculating the output equalization current value IeqThe calculation expression is as follows:
Figure BDA0002713567000000076
compared with the prior art, the invention has the following remarkable advantages:
the equalization control algorithm provided by the invention adopts the AUKF algorithm to improve the SOC estimation precision and solves the problem of possible overcharge and overdischarge damages to the battery when only the SOC is used as a parameter of the equalization control algorithm. And both the SOC and the terminal voltage are input into the adaptive fuzzy neural network. The self-adaptive fuzzy neural network combines the advantages of fuzzy logic control and the neural network, has strong self-adaptive capacity and high fault tolerance, and can adjust the membership function parameters and the weight between the neurons; the reasoning process is completed by the neuron, so that the speed is higher; the reliability of the algorithm can be improved by using expert experience. The fuzzy neural network designed by the invention is a first-order T-S fuzzy neural network, 5 parameters in total of a front part parameter and a back part parameter are required to be determined through learning, the learning algorithm adopts a hybrid algorithm of a BP algorithm and a least square method, the front part parameter is determined through the BP algorithm, and the back part parameter is determined through the least square method. And training the fuzzy neural network through data in the database, and identifying to obtain all parameters.
Drawings
FIG. 1 is a flow chart of an equalization control algorithm provided by the present invention;
FIG. 2 is a diagram of a second-order equivalent circuit model according to the present invention;
FIG. 3 is a SOC-OCV characteristic curve diagram of a lithium ion battery provided by the present invention;
FIG. 4 is a diagram of a low-high imbalance state according to the present invention;
FIG. 5 is a high-low imbalance state diagram according to the present invention;
fig. 6 is a diagram of a fuzzy neural network structure provided by the present invention.
Detailed Description
The technical solutions of the embodiments of the present invention are clearly and completely described below with reference to the drawings in the present invention, and it is obvious that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, shall fall within the scope of protection of the present invention.
For ease of understanding and explanation, as shown in fig. 1-6, the present invention provides a battery energy equalization method suitable for an equalization circuit, comprising the steps of:
s1: collecting the voltage and current of the equalizing circuit;
s2: establishing a second-order equivalent circuit model, and estimating the SOC by using an AUKF algorithm;
s3: calculating the difference value of the SOC of different batteries, and judging whether the conditions of one low or one high or one low exist;
s4: establishing a neural network, and if the situation described in the step S3 exists, taking the terminal voltage as the input of the neural network, and determining whether the output equalization current value reaches an equalization threshold value;
s5: if the situation described in the step S3 does not exist, calculating an average value of the SOC of each battery, and determining whether the average value is between 0.2 and 0.8, and if so, determining whether the output equalization current value reaches an equalization threshold value by using the SOC as an input of the neural network; if the voltage is not between 0.2 and 0.8, the terminal voltage is used as the input of the neural network, and whether the output equalization current value reaches the equalization threshold value is judged;
s6: and when the output equalization current value reaches an equalization threshold value, the battery energy is equalized.
Example 1
The concentration polarization effect and the electrochemical polarization effect exist in the battery in the charging and discharging process, and the two effects can lead the terminal voltage and the SOC of the battery to present hysteresis loop. Therefore, a second-order RC equivalent circuit model is established, and the second-order equivalent circuit model in step S2 describes the concentration polarization effect and the electrochemical polarization effect by using 2 RC networks, so as to improve the model accuracy. The second order RC equivalent circuit model is shown in FIG. 2, UOCV(t) is the open circuit voltage of the battery, and the two RC networks have four parameters in total, wherein Re、CeRespectively represents the electrochemical polarization internal resistance and the capacitance of the battery, Rd、CdRespectively representing concentration polarization resistance and capacitance; r0The ohmic internal resistance of the battery is shown, and I (t) shows the main circuit current, the discharging is positive, and the charging is negative.
From KCL and KVL laws, a circuit state expression can be derived as follows:
UO(t)＝ROI(t) (1)
Figure BDA0002713567000000091
Figure BDA0002713567000000092
wherein, UO(t) is the ohmic voltage, Ue(t) electrochemical polarization voltage, Ud(t) is concentration polarization voltage;
U(t)＝UOCV(t)+UO(t)+Ue(t)+Ud(t) (4)
wherein U (t) is the battery terminal voltage;
obtaining an expression of the change of the SOC of the battery along with time by using a current integration method:
Figure BDA0002713567000000101
therein, SOC0Is an initial SOC, QNThe standard capacitance of the battery is used, and eta is the charge-discharge efficiency of the battery;
discretization of formulae (2), (3) and (5) gives:
Figure BDA0002713567000000102
Figure BDA0002713567000000103
Figure BDA0002713567000000104
wherein, tau is a sampling period, taue＝ReCe，τd＝RdCd；
A discretized state space observation equation is established by equations (6), (7) and (8), as follows:
Figure BDA0002713567000000105
wherein w (k-1) is the observation noise of the model;
the discretized state output equation of the model is shown in equation (10):
U(k)＝UOCV(k)+UO(k)+Ue(k)+Ud(k)+v(k-1) (10)
where v (k-1) is the observation noise of the terminal voltage.
Example 2
The AUKF algorithm adopts an iterative updating method to estimate the system state, and in the iterative process, the maximum likelihood estimation and the expectation maximization method are used to continuously correct the system noise and the observation noise, so that the self-adaptive process is realized, and a better estimation effect is obtained. Based on the established battery equivalent circuit model, the AUKF algorithm in step S2 first constructs sampling points, and the state vector X (k) is composed of state variables SOC (k), Ue(k) And Ud(k) The composition is expressed as follows:
X(k)＝[SOC(k) Ue(k) Ud(k)]T (11)
the step of estimating the SOC by using the AUKF algorithm in step S2 includes:
s21: initializing a nonlinear system:
Figure BDA0002713567000000111
Figure BDA0002713567000000112
s22: constructing 2n +1 sampling point sets and carrying out nonlinear conversion:
Figure BDA0002713567000000113
s23: the weight in the iterative calculation process is calculated as follows:
Figure BDA0002713567000000114
wherein,
Figure BDA0002713567000000116
for a scale parameter, α represents the sampling point at
Figure BDA0002713567000000115
Distribution of the surroundings, i.e. 1>α>e-4；
Figure BDA0002713567000000117
Generally takes on a value of
Figure BDA0002713567000000118
Beta is used to integrate X prior estimates, generally chosen to be 2; wi (m)And Wi (c)Respectively calculating the weighted factors used for the mean value and the covariance of the ith sampling point;
s24: the one-step prediction for computing the 2n +1 sample point sets is:
Xi(k-1)＝f(Xi,k-1),i＝0,1,…2n (16)
s25: the time update in the AUKF algorithm is:
Figure BDA0002713567000000121
s26: the measurements in the AUKF algorithm are updated as:
Figure BDA0002713567000000122
s27: the battery state variables and covariance are estimated as:
Figure BDA0002713567000000123
s28: adaptive law covariance matching:
Figure BDA0002713567000000124
where q (k) is the covariance of the model noise w (k), and r (k) is the covariance of the terminal voltage measurements v (k).
In the prior art, the SOC can accurately reflect the state of charge of the battery, and the performance of a control algorithm which selects the SOC as a unique distinguishing parameter can meet the actual requirement. However, when the battery is discharged in a low state, the voltage drop of the battery will increase rapidly, and when the battery is discharged with a large current, if the battery is not provided with the minimum voltage limiting protection, the battery will be in an overdischarge state for a long time, which may cause damage to the battery. Similarly, when the battery is discharged to a high state of charge, the voltage rise of the battery increases, and when the battery is charged with a large current, the battery is easily overcharged for a long time, and the battery is damaged. In addition, when the battery is severely aged, the polarization effect and the ohmic effect of the battery are more remarkable, and the voltage drop of the battery in a low-charge state and the voltage rise of a high-charge state are more obvious.
In summary, to prevent possible overcharge and overdischarge damage to the battery when only the SOC is used as a parameter of the equalization control algorithm. The invention adopts an equalization control algorithm based on SOC and terminal voltage, and uses the average value of SOC of the series battery pack as the basis of the equalization control algorithm of the terminal voltage or SOC. In combination with the SOC-OCV characteristic curve of the lithium ion battery determined in the figure 3, if the mean value of the SOC is between 0.2 and 0.8, an SOC-based equalization control algorithm is adopted; and if the mean value of the SOC is between 0 and 0.2 or 0.8 and 1, adopting a terminal voltage-based equalization control algorithm.
However, the SOC average value is not used as the only judgment basis of the selection algorithm in the invention in consideration of the unbalanced state diversity of the battery pack. When the charge amount of the battery pack is in a high-low unbalance state, as shown in fig. 4, although the SOC is equalized between 0.2 and 0.8, a high-charge-amount battery is close to a full charge state, and therefore, in order to avoid overcharging the high-charge-amount battery in the high-low unbalance state, it is necessary to employ a terminal-voltage-based equalization control algorithm. Similarly, when the battery pack is in an unbalanced state of low charge or high charge, as shown in fig. 5, in order to prevent the low-battery cell from over-discharging, an equalization control algorithm based on the terminal voltage is also required.
In summary, the specific flow of the equalization control adopted in the present invention is shown in fig. 1. In fig. 1, both SOC and terminal voltage are input to the adaptive fuzzy neural network. The self-adaptive fuzzy neural network combines the advantages of fuzzy logic control and the neural network, has strong self-adaptive capacity and high fault tolerance, and can adjust the membership function parameters and the weight between the neurons; the reasoning process is completed by the neuron, so that the speed is higher; the reliability of the algorithm can be improved by using expert experience.
Example 3
As shown in fig. 6, the neural network in step S4 includes: the fuzzy rule strength normalization layer comprises a fuzzy layer, a fuzzy rule strength release layer, a rule strength normalization layer, a fuzzy rule output layer and an output layer.
The input quantity of the blurring layer is 2, which are respectively the SOC difference value and the SOC average value, and can be expressed as:
Figure BDA0002713567000000141
Figure BDA0002713567000000142
therein, SOCiThe SOC value, SOC, of the ith battery in the m single batteries on the left side of the equalizing unitjThe SOC value of the jth battery in the n batteries at the right side of the equalizing unit is shown, delta SOC represents the difference value of the average values of the SOC at the left side and the right side of the equalizing unit, and SOCavgRepresents the average value of the SOC on both sides of the equalizing unit;
the universe of discourse is defined for two input quantities based on selected battery characteristics and is designated A and B, respectively, based onThe discourse domain of two input quantities delta SOC of the fuzzy neural network of the SOC is [0,0.6 ]]，SOCavgHas a discourse field of [0,1](ii) a Delta V in two input quantities of the fuzzy neural network based on terminal voltage is [0,1]，VavgIs [2.6,4.2 ]](ii) a Considering the unbalanced state and the battery characteristics of the battery pack, dividing discourse domains of two input quantities into 5 fuzzy sets which are VS, S, M, L and VL respectively, and determining the membership function number of each input quantity; the membership function types are all Gaussian functions and are expressed as follows:
Figure BDA0002713567000000143
wherein a is the center of the membership function, b is the width of the membership function, and a and b are used as the front part parameters of the fuzzy neural network and are obtained by database training and learning;
input quantities delta SOC/delta V and SOCavg/VavgAre respectively defined as x1And x2: and obtaining an input and output expression of the first layer:
Figure BDA0002713567000000144
Figure BDA0002713567000000145
wherein, muAi(x) Is an input quantity x1Of the ith membership function, muBj(x) Is a number x2The jth membership function of (a).
(II) the fuzzy rule strength release layer multiplies output values of input quantities of the first layer by two to obtain output values of the second layer, so that 25 nodes of the layer can be determined according to the output values of the first layer, and input and output expressions of the layer are as follows:
Figure BDA0002713567000000151
wherein, ω iskRepresenting the activation strength of one of the fuzzy rules.
(III) the rule strength normalization layer calculates the proportion of the activation strength of each fuzzy rule, and the output expression is as follows:
Figure BDA0002713567000000152
and (IV) the fuzzy rule output layer is used for calculating the output of each fuzzy rule, and the calculation expression is as follows:
Figure BDA0002713567000000153
wherein p isk、qkAnd rkThe parameters are acquired through training as the back-piece parameters of the fuzzy neural network.
Fifthly, the output layer is used for calculating the output equilibrium current value IeqThe calculation expression is as follows:
Figure BDA0002713567000000154
the above disclosure is only for a few specific embodiments of the present invention, however, the present invention is not limited to the above embodiments, and any variations that can be made by those skilled in the art are intended to fall within the scope of the present invention.

Claims (7)

1. A battery energy balancing method suitable for a balancing circuit is characterized by comprising the following steps:
s1: collecting the voltage and current of the equalizing circuit;
s2: establishing a second-order equivalent circuit model, and estimating the SOC by using an AUKF algorithm;
s3: calculating the difference value of the SOC of different batteries, and judging whether the conditions of one low or one high or one low exist;
s4: establishing a neural network, and if the situation described in the step S3 exists, taking the terminal voltage as the input of the neural network, and determining whether the output equalization current value reaches an equalization threshold value;
s5: if the situation described in the step S3 does not exist, calculating an average value of the SOC of each battery, and determining whether the average value is between 0.2 and 0.8, and if so, determining whether the output equalization current value reaches an equalization threshold value by using the SOC as an input of the neural network; if the voltage is not between 0.2 and 0.8, the terminal voltage is used as the input of the neural network, and whether the output equalization current value reaches the equalization threshold value is judged;
s6: when the output equalization current value reaches an equalization threshold value, the battery energy is equalized;
the second-order equivalent circuit model in step S2 describes the concentration polarization effect and the electrochemical polarization effect by using 2 RC networks, and according to KCL and KVL laws, a circuit state expression can be obtained as follows:
UO(t)＝ROI(t) (1)
Figure FDA0003536295830000011
Figure FDA0003536295830000012
wherein, UO(t) is the ohmic voltage, Ue(t) electrochemical polarization voltage, Ud(t) is concentration polarization voltage;
U(t)＝UOCV(t)+UO(t)+Ue(t)+Ud(t) (4)
wherein U (t) is the battery terminal voltage;
obtaining an expression of the change of the SOC of the battery along with time by using a current integration method:
Figure FDA0003536295830000021
wherein,SOC0is an initial SOC, QNThe standard capacitance of the battery is used, and eta is the charge-discharge efficiency of the battery;
discretization of formulae (2), (3) and (5) gives:
Figure FDA0003536295830000022
Figure FDA0003536295830000023
Figure FDA0003536295830000024
wherein, tau is a sampling period, taue＝ReCe，τd＝RdCd；
A discretized state space observation equation is established by equations (6), (7) and (8), as follows:
Figure FDA0003536295830000025
wherein w (k-1) is the observation noise of the model;
the discretized state output equation of the model is shown in equation (10):
U(k)＝UOCV(k)+UO(k)+Ue(k)+Ud(k)+v(k-1) (10)
wherein v (k-1) is observation noise of the terminal voltage;
the AUKF algorithm in step S2 first constructs sampling points, and the state vector X (k) is composed of state variables SOC (k), Ue(k) And Ud(k) The composition is expressed as follows:
X(k)＝[SOC(k) Ue(k) Ud(k)]T (11)
the neural network in step S4 includes: the fuzzy rule strength normalization layer comprises a fuzzy layer, a fuzzy rule strength release layer, a rule strength normalization layer, a fuzzy rule output layer and an output layer.
2. The battery energy balancing method applicable to the balancing circuit according to claim 1, wherein the step of estimating the SOC by using the AUKF algorithm in step S2 comprises:
s21: initializing a nonlinear system:
Figure FDA0003536295830000031
Figure FDA0003536295830000032
s22: constructing 2n +1 sampling point sets and carrying out nonlinear conversion:
Figure FDA0003536295830000033
s23: the weight in the iterative calculation process is calculated as follows:
Figure FDA0003536295830000034
wherein,
Figure FDA0003536295830000036
for a scale parameter, α represents the sampling point at
Figure FDA0003536295830000035
Distribution of the surroundings, i.e. 1>α>e-4；
Figure FDA0003536295830000037
Generally takes on a value of
Figure FDA0003536295830000038
Beta is used to integrate X prior estimates, generally chosen to be 2; wi (m)And Wi (c)Respectively calculating the weighted factors used for the mean value and the covariance of the ith sampling point;
s24: the one-step prediction for computing the 2n +1 sample point sets is:
Xi(k-1)＝f(Xi,k-1),i＝0,1,…2n (16)
s25: the time update in the AUKF algorithm is:
Figure FDA0003536295830000041
s26: the measurements in the AUKF algorithm are updated as:
Figure FDA0003536295830000042
s27: the battery state variables and covariance are estimated as:
Figure FDA0003536295830000043
s28: adaptive law covariance matching:
Figure FDA0003536295830000044
where q (k) is the covariance of the model noise w (k), and r (k) is the covariance of the terminal voltage measurements v (k).
3. The battery energy balancing method suitable for the balancing circuit according to claim 1, wherein the input amount of the blurring layer is 2, which are the SOC difference value and the SOC average value, respectively, and can be expressed as:
Figure FDA0003536295830000045
Figure FDA0003536295830000046
therein, SOCiThe SOC value, SOC, of the ith battery in the m single batteries on the left side of the equalizing unitjThe SOC value of the jth battery in the n batteries at the right side of the equalizing unit is shown, delta SOC represents the difference value of the average values of the SOC at the left side and the right side of the equalizing unit, and SOCavgRepresents the average value of the SOC on both sides of the equalizing unit;
defining the universe of the two input quantities according to the selected battery characteristics and respectively named A and B, wherein the universe of the two input quantities delta SOC of the SOC-based fuzzy neural network is [0,0.6 ]]，SOCavgHas a discourse field of [0,1](ii) a Delta V in two input quantities of the fuzzy neural network based on terminal voltage is [0,1]，VavgIs [2.6,4.2 ]](ii) a Considering the unbalanced state and the battery characteristics of the battery pack, dividing discourse domains of two input quantities into 5 fuzzy sets which are VS, S, M, L and VL respectively, and determining the membership function number of each input quantity; the membership function types are all Gaussian functions and are expressed as follows:
Figure FDA0003536295830000051
wherein a is the center of the membership function, b is the width of the membership function, and a and b are used as the front part parameters of the fuzzy neural network and are obtained by database training and learning;
input quantities delta SOC/delta V and SOCavg/VavgAre respectively defined as x1And x2: and obtaining an input and output expression of the first layer:
Figure FDA0003536295830000052
Figure FDA0003536295830000053
wherein,
Figure FDA0003536295830000055
is an input quantity x1The ith membership function of (a),
Figure FDA0003536295830000056
is a number x2The jth membership function of (a).
4. The battery energy balancing method for the balancing circuit as claimed in claim 1, wherein the fuzzy rule strength releasing layer multiplies the output values of the input quantities of the first layer by two to obtain the output values of the second layer, so that 25 nodes of the layer can be determined from the output values of the first layer, and the input and output expressions of the layer are:
Figure FDA0003536295830000054
wherein, ω iskRepresenting the activation strength of one of the fuzzy rules.
5. The battery energy equalization method for the equalization circuit as claimed in claim 1, wherein the rule strength normalization layer calculates the proportion of the activation strength of each fuzzy rule, and the output expression is as follows:
Figure FDA0003536295830000061
6. the battery energy equalization method for an equalization circuit according to claim 1, wherein the fuzzy rule output layer is configured to calculate an output of each fuzzy rule, and the calculation expression is as follows:
Figure FDA0003536295830000062
wherein p isk、qkAnd rkThe parameters are acquired through training as the back-piece parameters of the fuzzy neural network.
7. The battery energy balancing method for the balancing circuit of claim 1, wherein the output layer is used for calculating the output balancing current value IeqThe calculation expression is as follows:
Figure FDA0003536295830000063
CN202011065240.1A 2020-09-30 2020-09-30 Battery energy balancing method suitable for balancing circuit Active CN112234673B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011065240.1A CN112234673B (en) 2020-09-30 2020-09-30 Battery energy balancing method suitable for balancing circuit

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011065240.1A CN112234673B (en) 2020-09-30 2020-09-30 Battery energy balancing method suitable for balancing circuit

Publications (2)

Publication Number Publication Date
CN112234673A CN112234673A (en) 2021-01-15
CN112234673B true CN112234673B (en) 2022-04-22

Family

ID=74120609

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011065240.1A Active CN112234673B (en) 2020-09-30 2020-09-30 Battery energy balancing method suitable for balancing circuit

Country Status (1)

Country Link
CN (1) CN112234673B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112952950A (en) * 2021-03-09 2021-06-11 石河子大学 Battery pack balancing method and system based on flyback converter

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104617623A (en) * 2015-01-30 2015-05-13 武汉理工大学 Balance control method for power battery pack of electric vehicle
CN107064811A (en) * 2017-03-01 2017-08-18 华南理工大学 A kind of lithium battery SOC On-line Estimation methods
CN110395142A (en) * 2019-07-16 2019-11-01 奇瑞新能源汽车股份有限公司 A kind of battery balanced control method of Adaptive Fuzzy Neural-network and its control system
CN111505506A (en) * 2020-05-15 2020-08-07 吉林大学 Battery SOC estimation method integrating multi-scale Kalman filtering and unscented Kalman filtering

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI260808B (en) * 2003-12-18 2006-08-21 Lg Chemical Ltd Apparatus and method for estimating stage of charge of battery using neural network

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104617623A (en) * 2015-01-30 2015-05-13 武汉理工大学 Balance control method for power battery pack of electric vehicle
CN107064811A (en) * 2017-03-01 2017-08-18 华南理工大学 A kind of lithium battery SOC On-line Estimation methods
CN110395142A (en) * 2019-07-16 2019-11-01 奇瑞新能源汽车股份有限公司 A kind of battery balanced control method of Adaptive Fuzzy Neural-network and its control system
CN111505506A (en) * 2020-05-15 2020-08-07 吉林大学 Battery SOC estimation method integrating multi-scale Kalman filtering and unscented Kalman filtering

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
动力电池状态估计及均衡电路的研究;尤森槟等;《电池》;20200625(第03期);全文 *
基于改进Sage-Husa的自适应无迹卡尔曼滤波的锂离子电池SOC估计;杨海学等;《电工电能新技术》;20160115(第01期);全文 *
基于无迹卡尔曼滤波估算电池SOC;石刚等;《计算机应用》;20161210(第12期);第1-4节 *

Also Published As

Publication number Publication date
CN112234673A (en) 2021-01-15

Similar Documents

Publication Publication Date Title
Lee et al. Soft computing for battery state-of-charge (BSOC) estimation in battery string systems
CN108983108B (en) Power battery pack peak power estimation method
Azis et al. State of charge (SoC) and state of health (SoH) estimation of lithium-ion battery using dual extended kalman filter based on polynomial battery model
Yan et al. Fuzzy control for battery equalization based on state of charge
CN108872869B (en) Lithium ion battery degradation classification method based on BP neural network
CN112464571B (en) Lithium battery pack parameter identification method based on multi-constraint-condition particle swarm optimization algorithm
CN115632179B (en) Intelligent quick charging method and system for lithium ion battery
CN112858929A (en) Battery SOC estimation method based on fuzzy logic and extended Kalman filtering
CN111079349B (en) Energy real-time optimization method for lithium battery and super capacitor composite power supply system
CN114217234B (en) IDE-ASRCKF-based lithium ion battery parameter identification and SOC estimation method
CN106026292A (en) A* algorithm-based battery equalization control method and system
CN112234673B (en) Battery energy balancing method suitable for balancing circuit
CN115101836A (en) Cloud-end converged battery system management method and device
Abbasi et al. Joint optimization of electric vehicle fast charging and DC fast charging station
CN114839537A (en) Battery state of charge prediction method and device
CN117799495A (en) Lithium ion battery state of charge estimation and equalization control method
CN114301121A (en) Lithium battery capacity strengthening and balancing maintenance method
Gharavian et al. ZEBRA battery SOC estimation using PSO-optimized hybrid neural model considering aging effect
Fan et al. An adaptive neuro-fuzzy inference system (ANFIS) based model for the temperature prediction of lithium-ion power batteries
CN111762059A (en) Multivariable fusion battery pack balancing method considering battery charging and discharging working conditions
Liu et al. RBF network-aided adaptive unscented kalman filter for lithium-ion battery SOC estimation in electric vehicles
Mawatwal et al. State of charge estimation for rechargeable lithium-ion battery using ANFIS MATLAB
Zhang et al. State-of-charge estimation of li-ion batteries based on a hybrid model using nonlinear autoregressive exogenous neural networks
Lipu et al. Optimal neural network approach for estimating state of energy of lithium-ion battery using heuristic optimization techniques
CN115308623A (en) Battery state of charge estimation method based on particle resampling and searcher optimization algorithm

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