Battery EIS electrochemical parameter identification and energy balancing method and equipment

Abstract

The invention provides a battery EIS electrochemical parameter identification and energy equalization method and device, and relates to the field of electrochemical analysis, and the method comprises the steps: generating an impedance data set, and obtaining the ohmic internal resistance of each cell; inputting the impedance data set into an electrochemical impedance model, and calculating to obtain a simulated electrochemical parameter set, wherein the simulated electrochemical parameter set comprises the exchange current density of each battery cell; constructing an SOH estimator by simulating an electrochemical parameter set, and building a neural network model; obtaining SOC and voltage of each battery cell of the battery pack through the trained neural network model; and obtaining the SOH of each battery cell according to the SOH estimator, and carrying out double-layer balance control on the battery pack according to the SOH, the ohmic internal resistance, the exchange current density, the SOC and the voltage of each battery cell of the battery pack. According to the method, the SOH estimator and the neural network model are constructed by simulating the electrochemical parameter set, and the electrochemical parameter consistency is brought into equilibrium management, so that the rapid and accurate evaluation of the battery state is realized.

Description

Technical Field

[0001] The present invention relates to the field of electrochemical analysis, and in particular to a battery EIS electrochemical parameter identification and energy balancing method and equipment. Background Art

[0002] Battery pack balancing technology is one of the key technologies to ensure the performance and life of the battery pack. Its purpose is to ensure that the charge, voltage and other parameters of each single cell in the battery pack remain consistent. At present, battery pack balancing technology is mainly divided into two methods: passive balancing and active balancing. Passive balancing technology achieves balancing by using components such as resistors to dissipate the energy of the battery discharge; active balancing technology relies on energy storage components such as inductors and capacitors through control circuits and algorithms to manage the battery in a targeted manner to achieve balancing; in recent years, a variety of balancing algorithms have been proposed, including classical control algorithms and modern intelligent algorithms. Classical control algorithms such as PID are widely used because of their simple structure, stability and reliability, but they require the selection of appropriate parameters and cannot quickly stabilize after mutations. Intelligent algorithms such as fuzzy logic control and neural networks can provide more flexible and efficient balancing strategies, but the existing electrochemical analysis methods still have the following problems:

[0003] (1) Traditional battery pack balancing technology is mainly divided into two methods: passive balancing and active balancing. Passive balancing discharges or charges through components such as resistors, while active balancing is managed through control circuits and algorithms. The electrochemical parameters of the battery obtained by traditional balancing methods are still not accurate enough.

[0004] (2) Although EIS technology can provide rich electrochemical information, in practical applications it may have problems such as complex data interpretation and long testing time, which limits its widespread application in battery management systems.

[0005] (3) Traditional SOH assessment methods may rely on a single parameter or simple voltage and capacity measurements, which may not accurately reflect the internal state and performance changes of the battery.

[0006] (4) Existing balancing strategies may be mainly based on the consistency judgment of voltage or SOC, and these methods may not fully consider the electrochemical characteristics and internal differences of the battery.

[0007] (5) Traditional battery management systems may be complex in structure, high in cost, and have limitations in battery status monitoring and balancing control. Summary of the invention

[0008] In view of this, the purpose of the present invention is to provide a battery EIS electrochemical parameter identification and energy balancing method and equipment to solve the problem that the electrochemical parameters of the battery obtained by the existing electrochemical analysis method are not accurate enough.

[0009] The present invention provides a battery EIS electrochemical parameter identification and energy balancing method, comprising the steps of:

[0010] S1: Generate an electrochemical impedance spectrum through a variable frequency AC impedance scanning device to detect the battery cell impedance data set and obtain the ohmic internal resistance of each battery cell;

[0011] S2: construct an electrochemical impedance model, input the impedance data set into the electrochemical impedance model to calculate and obtain a simulated electrochemical parameter set, wherein the simulated electrochemical parameter set includes the exchange current density of each battery cell;

[0012] S3: Construct a SOH estimator by simulating a set of electrochemical parameters, and build a neural network model for pre-training and transfer learning to obtain a trained neural network model;

[0013] S4: Identify the health status of the battery cell through the trained neural network model to obtain the SOC and voltage of each battery cell in the battery pack;

[0014] S5: The SOH of each battery cell is obtained according to the SOH estimator, and the battery pack is double-layer balanced according to the SOH, ohmic internal resistance, exchange current density, SOC and voltage of each battery cell of the battery pack, so that the battery pack performs charging and discharging operations.

[0015] Preferably, step S2 specifically comprises:

[0016] S21: Electrochemical impedance models include: EIS fractional-order model and finite element battery physical field impedance model;

[0017] S22: Obtain the frequency set, real part set and negative imaginary part set of the impedance data set, input the frequency set, real part set and negative imaginary part set into the EIS fractional-order model and the finite element battery physical field impedance model, and calculate and obtain the simulated electrochemical parameter set.

[0018] Preferred:

[0019] The exchange current density is calculated as:

[0020]

[0021] Where I0 is the exchange current density, R is the ideal gas constant, F is the Faraday constant, T is the thermodynamic temperature, α a is the charge transfer coefficient in the positive direction, k rate is the electrochemical reaction rate constant, E a is the activation energy, c1 is the solid phase lithium ion concentration, c2 is the liquid phase lithium ion concentration, c 2,max is the maximum lithium ion concentration in the liquid phase, c 1,max is the maximum lithium ion concentration in the solid phase, c1,surf is the lithium ion concentration on the solid surface.

[0022] Preferably, step S3 specifically comprises:

[0023] S31: obtaining an internal resistance parameter set from a simulated electrochemical parameter set, and constructing a SOH estimator through the internal resistance parameter set;

[0024] S32: pre-training the neural network model using the ohmic internal resistance and exchange current density of each battery cell as training data;

[0025] S33: Repeat step S32 until the training fitting error is lower than the training threshold, and obtain a pre-training model;

[0026] S34: Using the pre-trained model as the basic model and the cloud database as the target model data, transfer learning is performed on the pre-trained model. Meanwhile, the ohmic internal resistance, exchange current density and SOH of each cell in the transfer learning process are reported to the local end.

[0027] S35: Repeat step S34 until the transfer learning is completed to obtain a trained neural network model.

[0028] Preferably, step S5 is specifically:

[0029] S51: charging the battery pack, and starting the discharging operation after the battery pack is fully charged;

[0030] S52: Inputting the ohmic internal resistance and exchange current density of each battery cell into the SOH estimator to calculate and obtain the SOH of each battery cell;

[0031] S53: Obtain SOC dispersion by calculating the SOC of each cell. If the SOC dispersion exceeds the balancing threshold, the process proceeds to step S54 after completing the upper-layer balancing; otherwise, the process returns to step S52;

[0032] S54: If the SOC of each battery cell is 20-80%, then proceed to step S55; otherwise, proceed to step S57 after completing voltage difference balancing;

[0033] S55: After completing voltage difference balancing, proceed to step S56;

[0034] S56: Calculate the SOH range of each cell to obtain the SOH range. If the SOH range exceeds the range threshold, complete the lower layer balancing and proceed to step S57; otherwise, return to step S55;

[0035] S57: End the discharge operation and return to step S51.

[0036] Preferred:

[0037] The calculation formula of the SOH of the battery cell is:

[0038]

[0039] Among them, SOH i is the SOH of the ith cell, R End The ohmic internal resistance when the remaining capacity of the battery reaches 80% of the factory rated capacity, R New The internal resistance in ohms measured for a new battery, R c,i is the current ohmic internal resistance of the i-th cell in the battery pack, is the exchange current density of the ith cell, R is the ideal gas constant, F is the Faraday constant, T is the thermodynamic temperature, α a is the charge transfer coefficient in the positive direction.

[0040] Preferably, the process of lower-layer balancing is as follows:

[0041] S561: Obtaining the ohmic internal resistance and exchange current density of each battery cell;

[0042] S562: Get the maximum ohmic internal resistance Minimum ohmic internal resistance Ohm internal resistance balance threshold S Ω , Maximum exchange current density Minimum exchange current density and the exchange current density equilibrium threshold S I ,like and Then proceed to step S563, otherwise return to step S561;

[0043] S563: Calculate and obtain a correction factor and a balancing center point, and perform balancing electrical operation on the battery pack using the correction factor and the balancing center point.

[0044] Preferred:

[0045] The correction factor is calculated as:

[0046]

[0047] Where f(R0) is the correction factor, R0 is the ohmic internal resistance of the cell to be corrected, I, Π and III are the intervals of SOC parameters, R Π is the ohmic internal resistance of the battery cell at the beginning of the Π interval, and l is the length of the power range in the Π area;

[0048] The calculation formula of the equilibrium center point is:

[0049]

[0050] Among them, R c is the equilibrium center point, Δ(R c) represents the degree of deviation from the equilibrium center point, R i is the ohmic internal resistance of the i-th cell, R j is the ohmic internal resistance of the jth cell, δ is 5% of R c , η is the energy conversion efficiency, and h is the equilibrium threshold correction factor.

[0051] A storage medium, wherein the storage medium stores instructions and data for implementing the battery EIS electrochemical parameter identification and energy balancing method.

[0052] A battery EIS electrochemical parameter identification and energy balancing device comprises: a processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement the battery EIS electrochemical parameter identification and energy balancing method.

[0053] The present invention has the following beneficial effects:

[0054] (1) Combined with electrochemical impedance spectroscopy (EIS), the impedance data of the battery pack can be obtained. The impedance data can more accurately identify and evaluate the battery status, thereby achieving more effective energy balance management.

[0055] (2) An electrochemical impedance model consisting of an EIS fractional-order model and a finite element battery physical field impedance model was constructed. The impedance data set was input into the electrochemical impedance model, and a simulated electrochemical parameter set was obtained by calculation. A SOH estimator was constructed by simulating the electrochemical parameter set and a neural network model was built. It was also proposed to incorporate the consistency of electrochemical parameters into the balance management, thereby achieving a rapid and accurate assessment of the battery status.

[0056] (3) The equalization threshold correction factor is introduced to adaptively optimize the equalization threshold according to the working conditions of the battery pack, thereby improving the accuracy and efficiency of equalization.

[0057] (4) A double-layer balancing strategy based on the consistency of SOC, voltage and electrochemical parameters is proposed, combined with the battery mechanism characteristics to achieve more precise and effective battery pack energy balancing management.

[0058] (5) Only ohmic internal resistance and exchange current density are used as electrochemical parameters, which simplifies the process of battery state evaluation and balance control, reduces system complexity and cost, and improves system reliability and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 This is a flow chart of a method according to an embodiment of the present invention;

[0060] Figure 2 This is a structural diagram of the device according to an embodiment of the present invention;

[0061] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0062] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0063] Reference Figure 1 The present invention provides a battery EIS electrochemical parameter identification and energy balance method, comprising the steps of:

[0064] S1: Generate an electrochemical impedance spectrum through a variable frequency AC impedance scanning device to detect the battery cell impedance data set and obtain the ohmic internal resistance of each battery cell;

[0065] As an example, the working current of the lithium battery pack is monitored by the energy management system. If the following sampling conditions are met, the impedance online measurement device applies an AC current excitation signal of a certain frequency to each monomer in the pack, observes and records the response voltage of the monomer in the pack at the same frequency, and the ratio of the response voltage to the excitation current is used as the electrochemical impedance at the frequency. Two sampling conditions are defined as follows:

[0066] (1) The working current fluctuation does not exceed the sampling threshold and the duration exceeds a certain value.

[0067] (2) The lithium battery pack is charged and discharged several times.

[0068] If only sampling condition (1) is met, the medium-frequency and high-frequency impedances of each battery cell in the group are sampled respectively. If both sampling conditions (1) and (2) are met, the low-frequency and medium-frequency impedances of each battery cell in the group are sampled. The i-th battery in the battery group is sorted by frequency, and its sampling data format is where ν n is the nth frequency point divided by sampling time in the specified frequency range, is the real part of the impedance at a given frequency, It is the negative imaginary part of the impedance at a given frequency.

[0069] The average value of the two points on the left and right of the zero-crossing point in the real part data of the i-th battery cell extracted is taken as the ohmic internal resistance R Ω .

[0070] S2: construct an electrochemical impedance model, input the impedance data set into the electrochemical impedance model to calculate and obtain a simulated electrochemical parameter set, wherein the simulated electrochemical parameter set includes the exchange current density of each battery cell;

[0071] As an example,

[0072] Step S2 is specifically as follows:

[0073] S21: Electrochemical impedance models include: EIS fractional-order model and finite element battery physical field impedance model;

[0074] Specifically, the finite element battery physical field impedance model:

[0075] The basic structure of lithium batteries includes positive and negative current collectors, positive and negative porous electrodes and separators. First, from the perspective of electrochemical processes, lithium batteries are simplified in one dimension, which includes three domains with different thicknesses (positive electrode, negative electrode and separator). The size of the current collector is ignored, and the change in lithium ion concentration in the longitudinal direction of the battery is not taken into account. The electrochemical impedance model is established through simulation software. The frequency, real part and negative imaginary part of the impedance data are input into the model using a typical electrochemical process, and the global least squares is used for optimization and solution, with the ohmic internal resistance R Ω and exchange current density I0 as electrochemical parameters to describe battery characteristics. In the electrochemical reaction of the experimental lithium battery, the local current density I n The expression of is as follows:

[0076]

[0077] Where I0 is the exchange current density, a function of the battery temperature and the lithium ion concentration of the electrolyte and solid active material, R is the ideal gas constant, F represents the Faraday constant, T is the thermodynamic temperature, and α a and α b They represent the charge transfer coefficient in the positive (negative) direction, and γ is the activation overpotential.

[0078] The mean absolute error (MAE) and mean relative error (MSE) are used as the criteria for evaluating the simulation accuracy of the electrochemical model. The calculation formula is as follows:

[0079]

[0080] Where Z measurement and Z emulation Represent the impedance of experimentally measured EIS and simulated EIS at different frequencies. MAE z and MRE z If it is lower than the given threshold, it means that the model fitting effect is good. In the optimization process, the ohmic internal resistance R is solved. Ω And exchange current density I0 are important parameters for battery state evaluation and balancing management.

[0081] S22: Obtain the frequency set, real part set and negative imaginary part set of the impedance data set, input the frequency set, real part set and negative imaginary part set into the EIS fractional-order model and the finite element battery physical field impedance model, and calculate and obtain the simulated electrochemical parameter set.

[0082] As an example,

[0083] The exchange current density is calculated as:

[0084]

[0085] Where I0 is the exchange current density, R is the ideal gas constant, F is the Faraday constant, T is the thermodynamic temperature, α a is the charge transfer coefficient in the positive direction, k rate is the electrochemical reaction rate constant, E a is the activation energy, c1 is the solid phase lithium ion concentration, c2 is the liquid phase lithium ion concentration, c 2,max is the maximum lithium ion concentration in the liquid phase, c 1,max is the maximum lithium ion concentration in the solid phase, c 1 ,sur f is the lithium ion concentration on the solid surface.

[0086] S3: Construct a SOH estimator by simulating a set of electrochemical parameters, and build a neural network model for pre-training and transfer learning to obtain a trained neural network model;

[0087] As an example:

[0088] Step S3 is specifically as follows:

[0089] S31: obtaining an internal resistance parameter set from a simulated electrochemical parameter set, and constructing a SOH estimator through the internal resistance parameter set;

[0090] Specifically, the internal resistance definition of the lithium battery health state SOH is as follows:

[0091]

[0092] Where: R EOL is the internal resistance of the battery at the end of its service life; R C is the current internal resistance of the battery; R new The internal resistance of a new battery.

[0093] S32: pre-training the neural network model using the ohmic internal resistance and exchange current density of each battery cell as training data;

[0094] Specifically, a small sample experiment was conducted on a single model of lithium battery, and the battery was subjected to a cross experiment of cyclic charge and discharge and EIS test. The battery tester was used for cyclic charge and discharge, and the battery SOH gradually decreased during the process. An AC impedance scanner was used to perform EIS test at every 2% SOH interval. The impedance data extraction format was as in step S1, with SOH as the label and the impedance processing data of each sample battery cell as the input, and the training was performed through the neural network algorithm.

[0095] During the training process, the exchange current density is obtained by solving the SOH estimator, and then the ohmic internal resistance under each specific SOH is solved based on the EIS raw data. If the training fitting error is lower than the training threshold, the training requirements are met, and the model and the solved electrochemical parameters are saved.

[0096] S33: Repeat step S32 until the training fitting error is lower than the training threshold, and obtain a pre-training model;

[0097] S34: Using the pre-trained model as the basic model and the cloud database as the target model data, transfer learning is performed on the pre-trained model. Meanwhile, the ohmic internal resistance, exchange current density and SOH of each cell in the transfer learning process are reported to the local end.

[0098] Specifically, the transfer learning of the model:

[0099] The pre-trained model is implanted into the cloud-based battery energy management system. The pre-trained model is used as the basic model, and the cloud database is used as the target model data for transfer learning. The ohmic internal resistance and AC current density during the training process are saved to complete the health status assessment from the battery mechanism level, and the SOH of each battery cell is saved.

[0100] Report SOH to the local end:

[0101] To accurately obtain the health status value of each cell in the battery pack, the most direct way is to design a state estimator for each battery, but multiple estimates working at the same time will cause a huge burden on the system microprocessor. The patent of this invention uses a dual parallel EKF-global least squares algorithm to estimate the health status of the battery cells in the lithium battery pack. The algorithm tracks the SOH values ​​of all the cells in the group at a microscopic scale (i.e., at each sampling time point), and the corresponding SOH values ​​and battery serial numbers are stored in the corresponding series; secondly, the sorting traversal algorithm is used to arrange them according to the SOH size, and the battery serial numbers with the first and second largest SOH values ​​and the penultimate and second to last SOH values ​​are found, and the largest and second largest SOH values ​​and the corresponding battery serial numbers in the battery pack, as well as the smallest and second smallest SOH values ​​and the corresponding battery serial numbers are reported to the local end.

[0102] By using the dual parallel EKF-global least squares algorithm, the health status of the battery pack can be estimated with low computational cost and high reliability, and the inconsistent information of cells in the group can be accurately obtained.

[0103] S35: Repeat step S34 until the transfer learning is completed to obtain a trained neural network model.

[0104] S4: Identify the health status of the battery cell through the trained neural network model to obtain the SOC and voltage of each battery cell in the battery pack;

[0105] S5: The SOH of each battery cell is obtained according to the SOH estimator, and the battery pack is double-layer balanced according to the SOH, ohmic internal resistance, exchange current density, SOC and voltage of each battery cell of the battery pack, so that the battery pack performs charging and discharging operations.

[0106] As an example, during the charge and discharge process of the battery pack, in order to maximize the capacity of the battery pack, it is necessary to discharge and balance the cells that reach the charge and discharge cut-off voltage first. Most existing studies classify the cells in the group based on the working voltage, and perform battery balancing based on the charge and discharge cut-off voltage or the SOC difference in the battery pack during the charge and discharge process, that is, the consistency of the working voltage or the SOC consistency is used to determine whether balancing is required. The patent of this invention combines the characteristics of the battery mechanism and proposes a double-layer balancing control method based on the consistency of SOC, voltage and electrochemical parameters.

[0107] Step S5 is specifically as follows:

[0108] S51: charging the battery pack, and starting the discharging operation after the battery pack is fully charged;

[0109] S52: Inputting the ohmic internal resistance and exchange current density of each battery cell into the SOH estimator to calculate and obtain the SOH of each battery cell;

[0110] Specific:

[0111] The calculation formula of the SOH of the battery cell is:

[0112]

[0113] Among them, SOH i is the SOH of the ith cell, R End R is the ohmic internal resistance when the remaining capacity of the battery reaches 80% of the factory rated capacity. New The internal resistance in ohms measured for a new battery, R c,i is the current ohmic internal resistance of the i-th cell in the battery pack, is the exchange current density of the ith cell, R is the ideal gas constant, F is the Faraday constant, T is the thermodynamic temperature, α a is the charge transfer coefficient in the positive direction.

[0114] S53: Obtain SOC dispersion by calculating the SOC of each cell. If the SOC dispersion exceeds the balancing threshold, the process proceeds to step S54 after completing the upper-layer balancing; otherwise, the process returns to step S52;

[0115] Specifically, the upper-level balancing strategy:

[0116] The battery management system collects battery pack voltage, current, temperature and other parameter information, evaluates the SOC of n batteries, and records the SOCi (i=1,2,3…,n). Sort the acquired SOC values ​​and find the SOC of the cell with the largest remaining capacity Max , the SOC of the minimum remaining capacity monomer Min , calculate the SOC dispersion of the battery pack ΔSOC Max , and determine ΔSOC Max Is it greater than or equal to the set balance start threshold S? t .

[0117] When the SOC of the battery working state is between 20-80%, the battery cells in the first three SOC values ​​are discharged, and the battery cells in the last three SOC values ​​are charged; if the SOC is lower than 20% or higher than 80%, voltage difference is used for balancing, and the battery cells in the first three voltage values ​​are discharged, and the battery cells in the last three voltage values ​​are charged.

[0118] Set 95% of the lithium battery's charge cut-off voltage as the upper limit of the safety threshold, and 105% of the discharge cut-off voltage as the lower limit of the safety threshold. For example, if a certain type of lithium battery has a charge cut-off voltage of 4.2V, its upper limit of the safety threshold is 3.99V, and its discharge cut-off voltage is 2.6V, then its lower limit of the safety threshold is 2.73V. Set the upper limit of the cell voltage difference to 6mv as one of the conditions for starting the balancing, in order to ensure the balancing effect and not to start the balancing frequently.

[0119] S54: If the SOC of each battery cell is 20-80%, then proceed to step S55; otherwise, proceed to step S57 after completing voltage difference balancing;

[0120] S55: After completing voltage difference balancing, proceed to step S56;

[0121] S56: Calculate the SOH range of each cell to obtain the SOH range. If the SOH range exceeds the range threshold, complete the lower layer balancing and proceed to step S57; otherwise, return to step S55;

[0122] Specifically, the lower-level balancing strategy:

[0123] If the cells with the largest and smallest SOH values ​​exceed a given threshold, the lower-level balancing is turned on to achieve consistent balancing of the cell health status from the perspective of battery mechanism.

[0124] Assume the ohmic internal resistance of the i-th cell Exchange current density State of Health (SOH) i . The local end obtains the battery cells of the battery pack and And mark the maximum value and With minimum and Set the ohmic internal resistance balance threshold S Ω ,. Similarly, set the exchange current density balance threshold S I The specific threshold is determined according to the battery model. For example, for a certain type of 18650 battery, if When the balance is started and the resistance is less than 1mΩ, the balance is ended and charging or discharging is stopped. The switching current threshold setting is similar.

[0125] After the lower-level balancing is turned on, the four batteries with the largest, second largest, smallest, and second smallest SOH mentioned in step 2-3 are balanced and controlled.

[0126] The specific process of lower-level balancing is as follows:

[0127] S561: Obtaining the ohmic internal resistance and exchange current density of each battery cell;

[0128] S562: Get the maximum ohmic internal resistance Minimum ohmic internal resistance Ohm internal resistance balance threshold S Ω , Maximum exchange current density Minimum exchange current density and the exchange current density equilibrium threshold S I ,like and Then proceed to step S563, otherwise return to step S561;

[0129] S563: Calculate and obtain a correction factor and a balancing center point, and perform balancing electrical operation on the battery pack using the correction factor and the balancing center point.

[0130] Specifically, the equilibrium threshold solution:

[0131] This patent introduces the equalization threshold correction factor h to adaptively optimize the set threshold, and the threshold calculation equation is as follows:

[0132] R set =R c +(1+h)δ

[0133] I 0,set =λI c +(1+h)δ

[0134] Where: R set The ohm internal resistance balance threshold setting value; R c is the midpoint of the target cell ohmic internal resistance; h is the equalization threshold correction factor, and its value is determined according to the working conditions of the battery pack; δ is the equalization accuracy, and its value is 5%R c I 0,set is the exchange current density threshold setting value, I cis the target cell exchange current density midpoint; λ is the hardware condition compensation point;

[0135] The charge and discharge characteristics of the target battery are fitted to obtain the one-dimensional equation of the ohmic internal resistance:

[0136] V cells (R) = a i R i +a i-1 R i-1 +L+a1R 1 +a0

[0137] Where: a i is a polynomial; i is the number of cells.

[0138] In order to calculate the battery balancing threshold correction factor h, set e = |V′(R c )|, solve the inequality for |V′(Q)<1.5e|. The correction factor is shown in the following formula:

[0139] The correction factor is calculated as:

[0140]

[0141] Where f(R0) is the correction factor, R0 is the ohmic internal resistance of the cell to be corrected, I, Π and III are the intervals of SOC parameters, R Π is the ohmic internal resistance of the battery cell at the beginning of the Π interval, and l is the length of the power range in the Π area;

[0142] Specifically, considering the differences in cell monomers, the ohmic internal resistance of the battery module is corrected, and a safety factor σ is set at the boundary of adjacent regions to improve the accuracy of battery module balancing. Region III represents cells with an SOC of 20-80% and excludes Class II cells, and region I represents cells with an SOC below 20% or above 80%.

[0143]

[0144] Where: R * Represents the corrected battery ohmic internal resistance; R i Represents the ohmic internal resistance of each battery cell;

[0145] In order to determine the battery balancing center point R c , solve |Δ(R c The minimum value of )| is calculated as follows:

[0146] The calculation formula of the equilibrium center point is:

[0147]

[0148] Among them, Rc is the equilibrium center point, Δ(R c ) represents the degree of deviation from the equilibrium center point, R i is the ohmic internal resistance of the i-th cell, R j is the ohmic internal resistance of the jth cell, δ is 5% of R c , η is the energy conversion efficiency, and h is the equilibrium threshold correction factor.

[0149] Specifically, R c In [R min ,R max ] internal fluctuation, further deduced:

[0150]

[0151] Wherein: u represents that the ohmic internal resistance of u cells in the target module exceeds the set balancing threshold; v represents that the exchange current density of u cells in the target module exceeds the set balancing threshold.

[0152] Determine R C Finally, the battery pack balancing model is established, and its power transfer equation is:

[0153]

[0154] Where: t i Represents the time required for the battery pack to transfer C units of electricity from the high SOC value cell i to the low SOC value single cell j; T is the pulse signal period; Δq i,j Represents the amount of electricity transferred between single cells during a cycle.

[0155] S57: End the discharge operation and return to step S51.

[0156] See also Figure 2 , Figure 2 4 is a schematic diagram of the working of the hardware device of an embodiment of the present invention, wherein the hardware device specifically comprises: a battery EIS electrochemical parameter identification and energy balancing device 401, a processor 402 and a storage medium 403.

[0157] A battery EIS electrochemical parameter identification and energy balancing device 401: The battery EIS electrochemical parameter identification and energy balancing device 401 implements the battery EIS electrochemical parameter identification and energy balancing method.

[0158] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the battery EIS electrochemical parameter identification and energy balancing method.

[0159] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the battery EIS electrochemical parameter identification and energy balancing method.

[0160] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0161] The serial numbers of the embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. In a unit claim that lists several means, several of these means may be embodied by the same hardware item. The use of the words first, second, and third, etc. does not indicate any order and these words may be interpreted as identifiers.

[0162] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A battery EIS electrochemical parameter identification and energy balance method, characterized in that: Includes steps: S1: Generate an electrochemical impedance spectrum through a variable frequency AC impedance scanning device to detect the battery cell impedance data set and obtain the ohmic internal resistance of each battery cell; S2: construct an electrochemical impedance model, input the impedance data set into the electrochemical impedance model to calculate and obtain a simulated electrochemical parameter set, wherein the simulated electrochemical parameter set includes the exchange current density of each battery cell; S3: Construct a SOH estimator by simulating a set of electrochemical parameters, and build a neural network model for pre-training and transfer learning to obtain a trained neural network model; S4: Identify the health status of the battery cell through the trained neural network model to obtain the SOC and voltage of each battery cell in the battery pack; S5: The SOH of each battery cell is obtained according to the SOH estimator, and the battery pack is double-layer balanced according to the SOH, ohmic internal resistance, exchange current density, SOC and voltage of each battery cell of the battery pack, so that the battery pack performs charging and discharging operations.

2. The battery EIS electrochemical parameter identification and energy balancing method according to claim 1, characterized in that: Step S2 is specifically as follows: S21: Electrochemical impedance models include: EIS fractional-order model and finite element battery physical field impedance model; S22: Obtain the frequency set, real part set and negative imaginary part set of the impedance data set, input the frequency set, real part set and negative imaginary part set into the EIS fractional-order model and the finite element battery physical field impedance model, and calculate and obtain the simulated electrochemical parameter set.

3. The battery EIS electrochemical parameter identification and energy balancing method according to claim 1, characterized in that: The exchange current density is calculated as: Where I0 is the exchange current density, R is the ideal gas constant, F is the Faraday constant, T is the thermodynamic temperature, α a is the charge transfer coefficient in the positive direction, k rate is the electrochemical reaction rate constant, E a is the activation energy, c1 is the solid phase lithium ion concentration, c2 is the liquid phase lithium ion concentration, c 2,max is the maximum lithium ion concentration in the liquid phase, c 1,max is the maximum lithium ion concentration in the solid phase, c 1 ,surf is the lithium ion concentration on the solid surface.

4. The battery EIS electrochemical parameter identification and energy balancing method according to claim 1, characterized in that: Step S3 is specifically as follows: S31: obtaining an internal resistance parameter set from a simulated electrochemical parameter set, and constructing a SOH estimator through the internal resistance parameter set; S32: pre-training the neural network model using the ohmic internal resistance and exchange current density of each battery cell as training data; S33: Repeat step S32 until the training fitting error is lower than the training threshold, and obtain a pre-training model; S34: Using the pre-trained model as the basic model and the cloud database as the target model data, transfer learning is performed on the pre-trained model. Meanwhile, the ohmic internal resistance, exchange current density and SOH of each cell in the transfer learning process are reported to the local end. S35: Repeat step S34 until the transfer learning is completed to obtain a trained neural network model.

5. The battery EIS electrochemical parameter identification and energy balancing method according to claim 1, characterized in that: Step S5 is specifically as follows: S51: charging the battery pack, and starting the discharging operation after the battery pack is fully charged; S52: Inputting the ohmic internal resistance and exchange current density of each battery cell into the SOH estimator to calculate and obtain the SOH of each battery cell; S53: Obtain SOC dispersion by calculating the SOC of each cell. If the SOC dispersion exceeds the balancing threshold, the process proceeds to step S54 after completing the upper-layer balancing; otherwise, the process returns to step S52; S54: If the SOC of each battery cell is 20-80%, then proceed to step S55; otherwise, proceed to step S57 after completing voltage difference balancing; S55: After completing voltage difference balancing, proceed to step S56; S56: Calculate the SOH range of each cell to obtain the SOH range. If the SOH range exceeds the range threshold, complete the lower layer balancing and proceed to step S57; otherwise, return to step S55; S57: End the discharge operation and return to step S51.

6. The battery EIS electrochemical parameter identification and energy balancing method according to claim 5, characterized in that: The calculation formula of the SOH of the battery cell is: Among them, SOH i is the SOH of the ith cell, R End The ohmic internal resistance when the remaining capacity of the battery reaches 80% of the factory rated capacity, R New The internal resistance in ohms measured for a new battery, R c,i is the current ohmic internal resistance of the i-th cell in the battery pack, is the exchange current density of the ith cell, R is the ideal gas constant, F is the Faraday constant, T is the thermodynamic temperature, α a is the charge transfer coefficient in the positive direction.

7. The battery EIS electrochemical parameter identification and energy balancing method according to claim 5, characterized in that: The specific process of lower-level balancing is as follows: S561: Obtaining the ohmic internal resistance and exchange current density of each battery cell; S562: Get the maximum ohmic internal resistance Minimum ohmic internal resistance Ohm internal resistance balance threshold S Ω , Maximum exchange current density Minimum exchange current density and the exchange current density equilibrium threshold S I ,like and Then proceed to step S563, otherwise return to step S561; S563: Calculate and obtain a correction factor and a balancing center point, and perform balancing electrical operation on the battery pack using the correction factor and the balancing center point.

8. The battery EIS electrochemical parameter identification and energy balancing method according to claim 7, characterized in that: The correction factor is calculated as: Where f(R0) is the correction factor, R0 is the ohmic internal resistance of the cell to be corrected, I, Π and III are the intervals of SOC parameters, R Π is the ohmic internal resistance of the battery cell at the beginning of the Π interval, and l is the length of the power range in the Π area; The calculation formula of the equilibrium center point is: Among them, R c is the equilibrium center point, Δ(R c ) represents the degree of deviation from the equilibrium center point, R i is the ohmic internal resistance of the i-th cell, R j is the ohmic internal resistance of the jth cell, δ is 5% of R c , η is the energy conversion efficiency, and h is the equilibrium threshold correction factor.

9. A storage medium, characterized in that: The storage medium stores instructions and data for implementing the battery EIS electrochemical parameter identification and energy balancing method according to any one of claims 1 to 8.

10. A battery EIS electrochemical parameter identification and energy balancing device, characterized in that: include: Processor and storage medium; the processor loads and executes instructions and data in the storage medium to implement the battery EIS electrochemical parameter identification and energy balancing method according to any one of claims 1 to 8.

Images