EIS-based equivalent circuit model parameter identification method and device, terminal and storage medium

Abstract

The invention discloses an EIS-based equivalent circuit model parameter identification method and device, a terminal and a storage medium, and the method comprises the steps: obtaining the EIS test data of a battery, screening the EIS data corresponding to a preset target SOC, and endowing the EIS data of different frequency points with dynamic weights; obtaining a preset equivalent circuit model, setting a parameter range of each model parameter in the equivalent circuit model, and randomly generating an initial population in the parameter range based on logarithm distribution; performing iterative optimization on the initial population through a differential evolution algorithm, and dynamically adjusting a crossover probability and a scaling factor in an iteration process through an adaptive strategy; and when a preset iteration termination condition is satisfied, outputting an optimal parameter obtained after iteration. According to the method, the equivalent circuit model parameters are identified by using the differential evolution algorithm, so that the parameter identification efficiency and accuracy are improved.

Description

Technical Field

[0001] The present application relates to the field of lithium batteries, and in particular, to a method, device, terminal, and storage medium for identifying parameters of an equivalent circuit model based on EIS. Background Art

[0002] The battery management system plays a key role in ensuring the safe and efficient operation of the battery. The method for identifying parameters of an equivalent circuit model based on electrochemical impedance spectroscopy (EIS) can provide important support for functions such as analyzing the aging mode of the battery and estimating the state of health (SOH), which helps to improve the performance and reliability of the battery management system, and further promotes the further development of the entire battery application field.

[0003] Currently, traditional methods for identifying parameters of an equivalent circuit model based on EIS face many challenges in practical applications. For example, directly selecting EIS data at a specific state of charge (SOC), without considering the differences in data at different frequency points and treating all frequency point data uniformly, will cause the final parameter identification result to be easily affected by low-quality data, reducing the identification accuracy. Or, for the parameter setting of the equivalent circuit model, only giving a rough range based on experience, not accurately setting the model parameter range, or using an inappropriate initial population generation method, will cause the algorithm to fall into a local optimal solution during the iteration process and it is difficult to find the globally optimal parameters. It may also use fixed crossover probabilities and scaling factors during iterative optimization, resulting in its inability to adapt to changes during the iteration process and not being able to converge to the optimal solution quickly and effectively, thus affecting the efficiency and accuracy of the entire parameter identification.

[0004] Therefore, there is an urgent need for a method for identifying parameters of an equivalent circuit model based on EIS to improve the accuracy and reliability of battery equivalent circuit model parameter identification and provide more accurate data support for various functions of the battery management system. Summary of the Invention

[0005] In order to solve the above problems existing in the traditional method and improve the accuracy and reliability of battery equivalent circuit model parameter identification, the present application provides a method, device, terminal, and storage medium for identifying parameters of an equivalent circuit model based on EIS.

[0006] In the first aspect, the present application provides a method for identifying parameters of an equivalent circuit model based on EIS, adopting the following technical means: A method for identifying parameters of an equivalent circuit model based on EIS includes the following steps: Obtain the EIS test data of the battery, screen the EIS data corresponding to the preset target SOC, and assign dynamic weights to the EIS data at different frequency points; Obtain a preset equivalent circuit model, set the parameter ranges of the model parameters in the equivalent circuit model, and randomly generate an initial population based on a logarithmic distribution within the parameter ranges; Iteratively optimize the initial population through a differential evolution algorithm, and dynamically adjust the crossover probability and scaling factor during the iteration process through an adaptive strategy; When a preset iteration termination condition is met, output the optimal parameters obtained after iteration.

[0007] By adopting the above technical solutions, dynamic weights are assigned to the EIS data at different frequency points to improve the influence of the errors existing in the EIS data collected by the chip itself on the identification of model parameters; by setting the model parameter ranges and randomly generating an initial population based on a logarithmic distribution, the accuracy of model parameter identification is significantly improved; by dynamically adjusting the crossover probability and scaling factor through an adaptive strategy, the algorithm converges faster, and combining all operations to output the optimal parameters after meeting the iteration termination condition can complete the accurate identification of the parameters of the battery equivalent circuit model.

[0008] Preferably, the steps of obtaining the EIS test data of the battery, screening the EIS data corresponding to a preset target SOC, and assigning dynamic weights to the EIS data at different frequency points are specifically as follows: Collect the EIS test data under different states of charge through an EIS detection chip mounted on the battery, screen the EIS data corresponding to a preset target SOC, and preprocess the collected EIS data; Divide the frequency range of the EIS data into multiple frequency intervals according to a preset frequency division rule, each frequency interval has a corresponding basic weight, obtain the measured data quality of each frequency point, and adjust the basic weight of each frequency point according to the measured data quality to obtain a dynamic weight.

[0009] By adopting the above technical solutions, adjusting the basic weights of each frequency point according to the measured data quality to obtain dynamic weights can improve the influence of the errors existing in the EIS data collected by the chip itself on the identification of model parameters, optimize the sensitivity of the model to key information, and suppress noise interference.

[0010] Preferably, the steps of obtaining a preset equivalent circuit model, setting the parameter ranges of the model parameters in the equivalent circuit model, and randomly generating an initial population based on a logarithmic distribution within the parameter ranges are specifically as follows: Obtain a fractional-order equivalent circuit model including a Warburg impedance, where the fractional-order equivalent circuit model includes at least one constant phase element and one Warburg element, The fractional-order equivalent circuit model is, ; Among them, is the equivalent resistance, is the equivalent inductance, is the charge transfer resistance, and are the parameters of the constant phase equivalent element, is the pseudo-capacitance, is the fractional order exponent, , and are the parameters of the Weber element; Determine the parameter ranges of the various model parameters in the equivalent circuit model according to the physical characteristics of the battery; Within the parameter ranges, randomly generate an initial population based on a logarithmic distribution. Each initial population contains the model parameters to be identified. Specifically, Generate a corresponding logarithmic space based on the parameter range of each model parameter, and uniformly generate random numbers corresponding to the model parameters within each logarithmic space until an initial population of a preset scale is generated. The preset scale of the initial population is 1000, and the parameter dimension of the model parameters to be identified in each individual is 8.

[0011] By adopting the above technical solution, by obtaining a fractional order equivalent circuit model containing specific components and simultaneously determining the model parameter ranges according to the battery physical characteristics, it is possible to reduce the ineffective search space, improve the algorithm efficiency and accuracy, and significantly improve the model parameter identification accuracy; by randomly generating an initial population of a preset scale with a parameter dimension of 8 based on a logarithmic distribution within the parameter ranges, it is possible to make some of the population parameters randomly generated follow a logarithmic distribution, which helps to improve the model fitting effect.

[0012] Preferably, the initial population is iteratively optimized by a differential evolution algorithm, and the crossover probability and scaling factor in the iterative process are dynamically adjusted by an adaptive strategy, which specifically includes the following steps: Randomly select three different individuals from the currently generated initial population, , and calculate the mutation vector, that is, ; where k is the iteration number and F is the scaling factor; Perform crossover on the randomly selected target individual and the mutation vector according to the crossover probability to generate a trial individual, that is, When holds, ; Otherwise, ; Among them, is a random number, CR is the crossover probability, is the index of the parameter dimension, is a random integer, is the target individual; Calculate the fitness of the target individual and the trial individual respectively according to a preset fitness function, and the fitness function is, ; wherein, is the dynamic weight of each of the frequency points, is the model prediction value, is the measured value; If the fitness of the trial individual is less than the fitness of the target individual, let k = k + 1 and enter the next iteration, otherwise retain the target individual.

[0013] By adopting the above technical solution, the initial population is iteratively optimized through mutation, crossover, fitness calculation and selection operations in the differential evolution algorithm, gradually approaching the optimal solution; by combining the fitness function with the allocated dynamic weights, the optimization effect can be improved, the efficiency and accuracy of model parameter identification can be improved, and thus the parameter identification of the equivalent circuit model based on EIS can be better realized.

[0014] Preferably, the method further includes the following steps: According to the method for identifying parameters of an equivalent circuit model based on EIS as described in claim 4, it is characterized in that the method further includes the following steps: Obtain a preset initial crossover probability and an initial scaling factor, and perform dynamic updates on the crossover probability and the scaling factor. The formula is, ; ; wherein, is the initial crossover probability, is the initial scaling factor, K is the total number of iterations, is the preset crossover lower limit.

[0015] By adopting the above technical solution, performing dynamic updates on the crossover probability and the scaling factor can enable the differential evolution algorithm to adaptively adjust these two key parameters during the iteration process, making the algorithm converge faster, and thus improving the efficiency of identifying parameters of the equivalent circuit model.

[0016] Preferably, when the preset iteration termination condition is met, output the optimal parameters obtained after iteration, which specifically includes the following steps: During the iteration process, the individual with the optimal fitness in the global iteration process is stored, denoted as the global optimal individual, and the model parameters corresponding to the global optimal individual are the optimal parameters. After each new population is generated in the iteration, the fitness of the optimal individual in the current population is compared with the fitness of the global optimal individual, and the better fitness is used to update the global optimal individual; After each iteration ends, it is checked whether the current iteration count exceeds a preset iteration threshold, or whether the fitness of the current global optimal individual is less than a preset fitness threshold. If either of the above conditions is met, the iteration is terminated; otherwise, the iteration continues, and it is determined whether to terminate the iteration according to the preset iteration termination conditions; After the iteration is terminated, parameter verification and correction are performed based on the optimal parameters and the parameter range, and the corrected optimal parameters are output.

[0017] By adopting the above technical solution, during the iteration process, the current optimal parameters are always retained through continuous update, and it is determined whether to terminate the iteration according to the iteration count and the fitness threshold, so that the iteration can be stopped in time when the expected effect or the specified iteration count is reached, avoiding ineffective calculations; parameter verification, correction, and output are performed on the optimal parameters according to the parameter range, which can ensure that the output parameters conform to the battery physical characteristics and improve the accuracy of equivalent circuit model parameter identification.

[0018] Preferably, after the optimal parameters are output, the following steps are further included: Within the parameter range of each model parameter, the corresponding EIS curve graph is drawn in combination with the equivalent circuit model, and the intuitive mapping relationship between the parameter and the graph features is analyzed from the EIS curve graph; The battery health assessment is carried out by combining the optimal parameters for comparison and the intuitive mapping relationship, and a battery health assessment report is generated and output according to the results of the health assessment.

[0019] By adopting the above technical solution, parameter analysis is performed on the equivalent circuit model, the corresponding EIS curve graph is drawn in combination with the equivalent circuit model, the intuitive mapping relationship between the parameter and the graph features is obtained, and then the battery health assessment is carried out by using this mapping relationship and the optimal parameters, and a battery health assessment report is generated and output, which is beneficial to the development of subsequent application scenarios based on EIS data.

[0020] In a second aspect, the present application provides an equivalent circuit model parameter identification device based on EIS, adopting the following technical means: An equivalent circuit model parameter identification device based on EIS includes the following modules: A data preprocessing module, configured to obtain the EIS test data of the battery, screen the EIS data corresponding to a preset target SOC, and assign dynamic weights to the EIS data at different frequency points; A population initialization module, which is used to obtain a preset equivalent circuit model, set the parameter ranges of each model parameter in the equivalent circuit model, and randomly generate an initial population based on a logarithmic distribution within the parameter ranges; An iterative optimization module, which is used to iteratively optimize the initial population through a differential evolution algorithm, and dynamically adjust the crossover probability and scaling factor during the iteration process through an adaptive strategy; An optimal parameter output module, which is used to output the optimal parameters obtained after iteration when a preset iteration termination condition is satisfied.

[0021] By adopting the above technical solution, a parameter identification system based on an equivalent circuit model is built, providing necessary software technical support for the accurate and efficient identification of battery parameters, and meeting the requirements of the technological progress of battery management.

[0022] In a third aspect, the present application provides an intelligent terminal, adopting the following technical solution: An intelligent terminal includes a memory and a processor. At least one instruction, at least one program, a code set or an instruction set is stored in the memory, and the at least one instruction, at least one program, the code set or the instruction set is loaded and executed by the processor to implement the equivalent circuit model parameter identification method based on EIS as described above.

[0023] In a fourth aspect, the present application provides a computer-readable storage medium, adopting the following technical solution: A computer-readable storage medium, in which at least one instruction, at least one program, a code set or an instruction set is stored, and the at least one instruction, at least one program, the code set or the instruction set is loaded and executed by a processor to implement the equivalent circuit model parameter identification method based on EIS as described above.

[0024] In summary, the present application has at least the following beneficial effects: (1) By assigning dynamic weights to the EIS data at different frequency points, the present application can improve the influence of the errors existing in the chip's acquisition of EIS data itself on the model parameter identification, and improve the parameter identification accuracy; (2) By setting the parameter ranges of each model parameter in the equivalent circuit model and randomly generating an initial population based on a logarithmic distribution within the parameter ranges, the present application can significantly improve the model parameter identification accuracy and avoid the algorithm from falling into a local optimal solution; (3) By dynamically adjusting the crossover probability and scaling factor during the iteration process through an adaptive strategy, the present application can make the algorithm converge faster, improve the efficiency and accuracy of parameter identification, and realize the identification of fractional-order equivalent circuit model parameters by using the differential evolution algorithm. Description of the Drawings

[0025] Figure 1 is the flowchart of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 2 is the EIS test data curve graph under different SOC states of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 3 is the schematic diagram of the parameter ranges of each model parameter of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 4 is the fitness schematic diagram corresponding to different crossover probabilities and scaling factors of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 5 is one of the corresponding relationship graphs between the fitness and the number of iterations of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 6 is one of the curve graphs of the model predicted value and the measured value of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 7 is the second corresponding relationship graph between the fitness and the number of iterations of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 8 is the second curve graph of the model predicted value and the measured value of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 9 is the schematic diagram of the values of each model parameter of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 10 is the EIS curve graph corresponding to the equivalent resistance R1 of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 11 is the EIS curve graph corresponding to the equivalent inductance L1 of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 12 is the EIS curve graph corresponding to the charge transfer resistance R2 of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 13 is the EIS curve graph corresponding to the constant phase equivalent element parameter Q of the equivalent circuit model parameter identification method based on EIS in the embodiment; Figure 14 is the constant phase equivalent element parameter of the equivalent circuit model parameter identification method based on EIS in the embodiment corresponding EIS curve graph; Figure 15 is the Weber element parameter W of the equivalent circuit model parameter identification method based on EIS in the embodiment R corresponding EIS curve graph; Figure 16 is the Warburg element parameter W of the equivalent circuit model parameter identification method based on EIS in the embodiment T The corresponding EIS curve diagram; Figure 17 is the Warburg element parameter W of the equivalent circuit model parameter identification method based on EIS in the embodiment P The corresponding EIS curve diagram; Figure 18 is the structural diagram of the equivalent circuit model parameter identification device based on EIS in the embodiment Specific implementation manners

[0026] Next, in conjunction with the accompanying drawings, the technical solutions in the embodiments of a method for processing a channel party service request based on a key pair of the present invention will be clearly and completely described. The described embodiments are only possible technical implementations of the present invention, not all possible implementations. Those skilled in the art can fully combine the embodiments of the present invention to obtain other embodiments without creative labor, and these embodiments are also within the protection scope of the present invention.

[0027] The equivalent circuit model parameter identification method based on EIS provided in the embodiments of the present application, as Figure 1 shown, includes the following steps: S1. Obtain the EIS test data of the battery, screen the EIS data corresponding to the preset target SOC, and assign dynamic weights to the EIS data at different frequency points, specifically including the following steps: S11. Through the EIS detection chip mounted on the battery, collect the EIS test data under different states of charge, and screen out the EIS data corresponding to the preset target SOC.

[0028] In this embodiment, the EIS detection chip is composed of 15 strings of 100Ah battery cells and an NXP chip, and the collected EIS test data is also collected based on this NXP chip. Mounting the EIS detection chip on a 280Ah high-voltage standard box gives the battery.

[0029] In this embodiment, the EIS test data collected by the NXP chip is as Figure 2 shown. In this embodiment, the target SOC is set to 90% SOC. Therefore, the data of 90% SOC is selected as the AC impedance data for equivalent circuit model parameter identification.

[0030] S12. Preprocess the collected EIS data.

[0031] The preprocessing includes excluding abnormal frequency points, such as outliers caused by chip acquisition noise.

[0032] S13. Divide the frequency range of the EIS data into multiple frequency intervals according to a preset frequency division rule. In this embodiment, the frequency intervals include a high-frequency band, a medium-frequency band, and a low-frequency band, and each frequency interval has a corresponding basic weight.

[0033] In this implementation, the frequency division rule is as follows: The frequencies from 0.01 Hz to 1 Hz are divided into the low-frequency band, and the corresponding basic weight is 1; the frequencies from 1 Hz to 100 Hz are divided into the medium-frequency band, and the corresponding basic weight is 0.8; the frequencies exceeding 100 Hz are divided into the high-frequency band, and the corresponding basic weight is 0.5.

[0034] The data in the high-frequency band is vulnerable to measurement noise, while the measurement time in the low-frequency band is relatively long, and the signal-to-noise ratio of the data may be relatively low. Therefore, the weight of the high-frequency band data needs to be reduced due to large noise, and the weight of the low-frequency band needs to be increased due to key information. The weight value can be adjusted.

[0035] S14. Obtain the measured data quality of each frequency point, and adjust the basic weight of each frequency point according to the measured data quality to obtain the dynamic weight.

[0036] The measured data quality in this embodiment is the signal-to-noise ratio. For each frequency point, calculate the signal-to-noise ratio of its measured impedance, that is: ; Among them, is the signal strength of the measured impedance, is the noise standard deviation, and the noise standard deviation can be obtained through multiple measurements or instrument accuracy estimation.

[0037] Associate the basic weight with the signal-to-noise ratio, that is, ; Among them, is the adjusted dynamic weight of the current frequency point, is the basic weight of the current frequency point.

[0038] In a specific implementation manner, if the signal-to-noise ratio of a certain frequency point is lower than a preset threshold, such as SNR < 2, choose to further reduce its weight or directly eliminate it.

[0039] The above steps rationally allocate the data contribution degrees of each frequency point through frequency segmentation and signal-to-noise ratio evaluation, so as to achieve a balance between noise suppression and information extraction, and realize the dynamic adjustment of weights.

[0040] S2. Obtain a preset equivalent circuit model, and set the parameter ranges of each model parameter in the equivalent circuit model. Randomly generate an initial population based on a logarithmic distribution within the parameter ranges, which specifically includes the following steps: S21. Obtain a fractional-order equivalent circuit model including Warburg impedance. The fractional-order equivalent circuit model includes at least one constant phase element and one Warburg element. In this embodiment, from the perspectives of comprehensive accuracy and computing power, the fractional-order equivalent circuit model is a second-order equivalent circuit model, including an equivalent resistance R1, an equivalent inductance L1, a charge transfer resistance R2, a constant phase equivalent element, and a Warburg element W.

[0041] The fractional-order equivalent circuit model is ; where is the equivalent resistance, is the equivalent inductance, is the charge transfer resistance, and are the parameters of the constant phase equivalent element, is the pseudo-capacitance, is the fractional-order exponent, , and are the parameters of the Warburg element.

[0042] S22. Determine the parameter ranges of each model parameter in the equivalent circuit model according to the physical characteristics of the battery.

[0043] For example, for the equivalent resistance R1, a reasonable interval is set according to the internal resistance characteristics of the battery. In this embodiment, the typical internal resistance range of a 280 Ah battery is referred to. For the equivalent inductance L1, a reasonable interval is set considering the polarization effect and material characteristics of the battery. For the constant phase equivalent element, a reasonable interval is set based on the double-layer capacitance characteristics of the battery. For example, according to the capacitance strength and the degree of deviation of the capacitance phase from the ideal value for setting.

[0044] In this embodiment, the parameter ranges of each model parameter are referred to Figure 3 as shown.

[0045] S23. Within the parameter ranges, randomly generate an initial population through logarithmic distribution. Each initial population contains the model parameters to be identified. Specifically, S231. Generate a corresponding logarithmic space based on the parameter range of each model parameter. For example, if the range of the model parameter x is [a, b], the corresponding logarithmic space is ; S232. Uniformly generate random numbers of the corresponding model parameters within each logarithmic space. ; Among them, is a uniformly distributed random number.

[0046] S233. Obtain the original parameter value through exponential reduction: .

[0047] If the generated parameter exceeds the range, it is directly randomly regenerated until all parameters are within the preset range, thereby ensuring the legality of the initial population and avoiding invalid solutions from participating in subsequent calculations.

[0048] S234. Repeat the above steps until an initial population of a preset scale is generated. The preset scale of the initial population is 1000, that is, 1000 individuals are generated. Among them, the parameter dimension of the model parameters to be identified in each individual is 8, that is, each individual includes 8 model parameters to be identified, that is, the 8 model parameters in the above step S21. Each individual in the population represents a complete set of model parameter combinations; In the current initial population, the i-th individual is represented as ; where D is the parameter dimension of the model parameters to be identified.

[0049] Restricting the parameter range limits the search space, can avoid exploring in areas without physical meaning, avoid waste of resources, improve efficiency, and at the same time ensure that the identification result conforms to the actual characteristics of the battery.

[0050] S3. Iteratively optimize the initial population through a preset differential evolution algorithm and a preset iteration rule, and adjust the crossover probability and scaling factor in the iteration process through an adaptive strategy, specifically including the following steps: S31. Randomly select three different individuals from the currently generated initial population, , and scale the vector difference between two of the individuals and synthesize it with the target individual to calculate the mutant vector, that is, ; where is the mutant vector, k is the number of iterations, and F is the scaling factor.

[0051] The above mutation operation generates new solutions using the differences between individuals, which can enhance the global search ability.

[0052] S32. Perform crossover recombination on the randomly selected target individual and the mutant individual corresponding to the mutant vector according to the crossover probability, exchange some information between the two, and generate a trial individual, that is, When , ; Otherwise, ; Among them, is a random number, is a random integer, j is the parameter dimension index, is the target individual, and CR is the crossover probability; The above steps recombine the target individual and the mutation vector through the crossover operation, which can increase the diversity of the population.

[0053] In a specific implementable manner, before the crossover operation, calculate the parameter similarity between the individuals to be crossed and perform inbreeding judgment: if the similarity between two individuals exceeds a preset threshold, in this embodiment, if the parameter difference is less than 5%, reject the crossover and instead cross with other individuals with larger differences in the population, which can prevent the population from losing its exploration ability due to excessive convergence, thereby maintaining the balance between global search and local development.

[0054] In a specific implementable manner, after the crossover operation, check whether the parameters of the newly generated trial individual exceed the parameter range. If so, randomly reset the parameter to dynamically correct the illegal parameter and maintain the feasibility of the population, avoiding interference from invalid solutions in fitness evaluation.

[0055] S33. Calculate the fitness of the target individual and the trial individual respectively according to the preset fitness function, and select one from the target individual and the trial individual to enter the next generation population according to the size of the fitness. The fitness function is ; Among them, is the dynamic weight of each frequency point, is the model prediction value, corresponding to the real part and the imaginary part of the impedance calculated by the model in sequence, is the measured value, corresponding to the real part and the imaginary part of the measured impedance in sequence. In the above formula, for the convenience of comparison, the fitness is enlarged by 10 12 times.

[0056] S34. If the fitness of the trial individual is less than the fitness of the target individual, let k = k + 1, and the trial individual enters the next iteration; otherwise, retain the target individual, that is, When ,[[]]END]] ; Otherwise ; Among them, is the fitness of the trial individual, is the fitness of the target individual.

[0057] The above steps retain the optimal individual through the selection operation, ensuring the quality of the population.

[0058] Such as Figure 4As shown, the fitness corresponding to different crossover probabilities and scaling factors is presented. When the crossover probability is 0.6 and the scaling factor is 0.6, the fitness is optimal, which is 32717. Among the Figure 4 data shown, the fitness is the result after being enlarged by 10 12 times.

[0059] As Figure 5 shown, it is the corresponding relationship between fitness and the number of iterations. Figure 6 It is the curve graph of the model prediction value and the measured value. Combining Figure 5 and Figure 6 , Figure 5 it shows that when the number of iterations is 5000, the optimal fitness is 32717, and the corresponding crossover probability is 0.6 and the scaling factor is 0.6. Combining Figure 6 it can be known that when the optimal fitness is 32717, the fitting effect is not ideal.

[0060] To improve the above results, after each iteration ends, by obtaining the preset initial crossover probability and initial scaling factor, the crossover probability and scaling factor are continuously updated dynamically. The formula is ; ; wherein, is the initial crossover probability, is the initial scaling factor, K is the total number of iterations, is the preset crossover lower limit.

[0061] In this embodiment, according to the results before improvement, the preset initial crossover probability is 0.6 and the initial scaling factor is 0.6.

[0062] In a specific implementable manner, in order to break the premature convergence of the population, increase its diversity, and avoid falling into local optimum, a perturbation operation and a random restart operation are added.

[0063] During the iteration process, every fixed number of generations N, with probability P, randomly select M% of the individuals in the population. For the selected individuals, randomly generate their parameters that satisfy the range constraints again, or replace them with the neighborhood solutions of the current optimal individuals. In this implementable manner, N = 500 generations, P = 10%, and M = 10.

[0064] Through the parameter range constraints and the perturbation operation, it can be jointly ensured that the population always explores within the feasible domain and avoids invalid solutions.

[0065] S4. When the preset iteration termination condition is satisfied, output the optimal parameters obtained after iteration, and perform parameter analysis on the equivalent circuit model based on the optimal parameters to obtain the analysis result. The specific steps are as follows: S41. During the iteration process, store the individual with the optimal fitness in the global iteration process, that is, the individual with the minimum fitness, denoted as the global optimal individual. The model parameters corresponding to the global optimal individual are the optimal parameters. After generating a new population in each iteration, compare the fitness of the optimal individual in the current population with the fitness of the global optimal individual, and update with the better fitness pair of the two.

[0066] In a specific implementable manner, when the fitness change rate is less than the preset fitness change threshold, select the optimal individual in the current population. This process should perform a local search algorithm on the optimal individual before ending the iteration. In this embodiment, the particle swarm optimization algorithm is selected to finely adjust the parameters within its neighborhood, and replace the worst individual in the population with the optimal solution obtained by local search. It can make up for the deficiency of the differential evolution algorithm in local search ability and significantly improve the parameter identification accuracy.

[0067] S42. Judge whether to terminate the iteration according to the preset iteration termination condition. The iteration termination condition is that after each iteration ends, check whether the current iteration number exceeds the preset iteration threshold, or whether the fitness of the current global optimal individual is less than the preset fitness threshold; If any of the above conditions is met, terminate the iteration; otherwise, continue the iteration; In this embodiment, the iteration threshold is 10,000 and the fitness threshold is 1,000, that is, the termination condition is that the iteration number is greater than 10,000 times or the fitness of the current global optimal individual is less than 1,000.

[0068] As Figure 7 and Figure 8 shown, Figure 7 when iterating to 10,000 times in Figure 8 the fitness is 1,667. At this time,

[0069] As Figure 9 shown are the values of the identified model parameters at this time.

[0070] S43. After terminating the iteration, perform parameter verification and correction according to the optimal parameters and the parameter range, and output the corrected optimal parameters.

[0071] In this embodiment, parameter verification includes verifying whether the values of each model parameter are within the parameter range. If the parameter exceeds the boundary, it needs to be corrected to within the range, for example, correcting it to the nearest boundary value. In a specific implementable manner, substitute the parameters into the equivalent circuit model to generate a simulated EIS curve, and obtain the parameter identification results of the model under different SOCs. The average fitness is 1,555, which is equivalent to the fitting effect using ZView software.

[0072] S44. Draw the corresponding EIS curve diagram in combination with the equivalent circuit model within the parameter range of each model parameter.

[0073] The specific implementation method is that for each model parameter, while fixing other parameters, take 6 logarithmically equally spaced points within its corresponding parameter range and draw the EIS curve diagram.

[0074] The range of each parameter is ; Divide the range into 5 equally spaced intervals in logarithmic coordinates to generate 6 sampling points: , j = 0, 1, 2, 3, 4, 5; For each sampling point, substitute it into the equivalent circuit model and calculate the real part of the impedance and the imaginary part ; If the current parameter changes and other parameters are fixed, generate 6 curves and observe the changes in the curves.

[0075] In a specific implementable manner, Figures 10 - 17 in sequence are , , , , , , and the corresponding EIS curve diagram.

[0076] S45. Analyze the EIS curve diagram to obtain the intuitive mapping relationship between parameters and graphical features.

[0077] In a specific implementable manner, referring to Figures 10 - 17 , the intuitive mapping relationship between parameters and graphical features of each model parameter can be obtained as: From Figure 10 , it can be obtained that the larger R1 is, the more the EIS curve shifts to the right, that is, the real-axis intercept in the high-frequency region shifts to the right as R1 increases, and the overall curve shifts.

[0078] The intuitive connection is that R1 directly controls the real part of the high-frequency impedance, reflecting the battery body resistance. An increase in R1 indicates a decrease in the conductivity of the electrolyte, electrode corrosion, or poor contact.

[0079] From Figure 11 , it can be obtained that the larger the inductance L1 is, the more the left side of the EIS curve is pulled down, that is, the low-frequency region (the transition section between medium and high frequencies) bends downward, and the larger L1 is, the more obvious the left side of the curve is pulled down.

[0080] The intuitive connection is that L1 is related to electrochemical polarization or parasitic inductance of the test system, and when it increases, it will introduce inductive reactance.

[0081] FromFigure 12 It is obtained that the larger the charge transfer resistance R2 is, the larger the semicircle in the EIS curve is, that is, the radius of the mid-frequency semicircle expands with the increase of R2, and the right endpoint of the semicircle moves to the right.

[0082] The intuitive connection is that R2 represents the difficulty of charge transfer. When it increases, the electrochemical reaction is hindered, reflecting the decrease in electrode activity, such as the thickening of the SEI film and the inactivation of the catalyst.

[0083] From Figure 13 It is obtained that the larger the Q is, the higher the right side of the semicircle in the EIS curve is pulled.

[0084] The intuitive connection is that Q reflects the non-ideality of the capacitor. The smaller Q is, the flatter the semicircle is, and the weaker the capacitor characteristics are, indicating the degradation of the double-layer capacitance.

[0085] From Figure 14 It is obtained that the larger it is, the larger the diameter of the semicircle and the tangent angle of the imaginary part are. The tangent angle of the imaginary part is the slope in the low-frequency region.

[0086] The intuitive connection is that controls the frequency dependence of the constant phase equivalent element, and the abnormality indicates a change in the porous electrode structure.

[0087] From Figure 15 It is obtained that W R the larger it is, the smaller the connection angle between the semicircle and the straight line is, and the connection is smoother; From Figure 16 It is obtained that W T the larger it is, the shorter the straight line is; From Figure 17 It is obtained that W P affects the slope of the Warburg impedance; The intuitive connection is that W reflects the ion diffusion process. When W R increases or W T decreases, it indicates an increase in the diffusion resistance. When W P is abnormal, it indicates that the diffusion process is hindered, such as the imbalance of the electrolyte concentration gradient.

[0088] S46. Compare the effects of different parameter changes on the curve and identify the sensitive parameters among them.

[0089] In this embodiment, the high-frequency region is significantly affected by R1, the mid-frequency semicircle is controlled by R2 and Q (double-layer capacitance), and the low-frequency slope is determined by Q and the Warburg impedance W P The sensitive parameters are R1, R2, Q, and W P .

[0090] Combine the parameters for sensitivity analysis to obtain the fault diagnosis conclusion.

[0091] High-frequency region fault: If R1 exceeds 50% of the initial value, it is determined that the change is significant, and it is diagnosed as electrolyte dry-out or current collector corrosion; Mid-frequency region fault: If R2 suddenly increases and the semicircle radius expands, it indicates loss of electrode active material or an increase in interfacial side reactions; Low-frequency region fault: If Q drops significantly and the slope deviates from -45°, it indicates collapse of the electrode material structure or degradation of capacitance characteristics.

[0092] Relative to the change of a single parameter, in more cases, the fault type is judged by combining multiple parameter anomalies.

[0093] In a specific embodiment, if R1 increases while Q decreases, it indicates the combined effect of electrolyte aging and electrode degradation.

[0094] If W P Fluctuates and R2 increases, indicating abnormal diffusion process and increased charge transfer resistance.

[0095] S47. Combine the optimal parameters and the intuitive mapping relationship to conduct a health assessment of the battery, generate a health assessment report of the battery based on the results of the health assessment, and output it.

[0096] In this embodiment, sensitive parameters are extracted from the optimal parameters. Since the changes of sensitive parameters in the curve are more significant, in order to save computing resources, sensitive parameters can be extracted for health assessment. In other implementable specific ways, if the accuracy of the results is pursued, health assessment can also be conducted based on all the model parameters.

[0097] Combine the parameter ranges of each sensitive parameter, perform normalization processing on each sensitive parameter respectively to obtain the corresponding parameter values, and eliminate the influence of dimension.

[0098] Obtain the health threshold of each sensitive parameter in the battery, calculate the relative change of the parameter value of the current sensitive parameter and the health threshold, and obtain the degradation factor.

[0099] Combine the degradation factors of all sensitive parameters, and perform weighted comprehensive calculation with their respective preset weight coefficients to obtain the comprehensive health index of each sensitive parameter.

[0100] Evaluate the health status of the battery according to the comprehensive health index and the preset health assessment rules.

[0101] In this embodiment, the health assessment rules divide the battery into 4 grades: excellent, good, critical, and failed, as shown in the following table, Health level SOH range Health assessment Excellent >90% Low internal resistance, high capacitance Good 70%-90% Mild aging Critical 50%-70% Needs maintenance or replacement Failed <50% Loss of function Summarize the optimal parameter list and the corresponding fitness values, the health status grade and degradation factor of the battery, and the fault diagnosis conclusion into a health assessment report and output it.

[0102] Based on the same inventive concept described above, an embodiment of the present application also discloses an equivalent circuit model parameter identification device based on EIS, as Figure 18 shown, which includes the following modules: A data preprocessing module, configured to obtain EIS test data of a battery, screen the EIS data corresponding to a preset target SOC, and assign dynamic weights to the EIS data at different frequency points; A population initialization module, configured to obtain a preset equivalent circuit model, set the parameter ranges of the model parameters in the equivalent circuit model, and randomly generate an initial population based on a logarithmic distribution within the parameter ranges; An iterative optimization module, configured to perform iterative optimization on the initial population through a differential evolution algorithm, and dynamically adjust the crossover probability and scaling factor during the iterative process through an adaptive strategy; An optimal parameter output module, configured to output the optimal parameters obtained after iteration when a preset iterative termination condition is satisfied.

[0103] In a specific feasible implementation, the data preprocessing module includes the following units: A first data preprocessing unit, configured to collect EIS test data under different state of charge through an EIS detection chip mounted on the battery, screen the EIS data corresponding to a preset target SOC, and preprocess the collected EIS data; A second data preprocessing unit, configured to divide the frequency range of the EIS data into multiple frequency intervals according to a preset frequency division rule, each frequency interval having a corresponding basic weight, obtain the measured data quality of each frequency point, and adjust the basic weight of each frequency point according to the measured data quality to obtain a dynamic weight.

[0104] In a specific feasible implementation, the population initialization module includes the following units: A first population initialization unit, configured to obtain a fractional-order equivalent circuit model including a Warburg impedance, the fractional-order equivalent circuit model including at least one constant phase element and one Warburg element, The fractional-order equivalent circuit model is ; wherein, is the equivalent resistance, is the equivalent inductance, is the charge transfer resistance, and are the constant phase equivalent element parameters, is the pseudocapacitance, is the fractional-order exponent, 、 and are the Warburg element parameters; The second population initialization unit is used to determine the parameter ranges of the model parameters in the equivalent circuit model according to the physical characteristics of the battery; The third population initialization unit is used to randomly generate an initial population within the parameter ranges, and each initial population contains the model parameters to be identified. Specifically, Generate a corresponding logarithmic space based on the parameter range of each model parameter, and uniformly generate random numbers of the corresponding model parameters within each logarithmic space until an initial population of a preset scale is generated. The preset scale of the initial population is 1000, and the parameter dimension of the model parameters to be identified in each individual is 8.

[0105] In a specific implementable solution, the iterative optimization module includes the following units: The first iterative optimization unit is used to randomly select three different individuals from the currently generated initial population, and calculate the mutation vector, that is, ; where k is the number of iterations and F is the scaling factor; The second iterative optimization unit is used to perform crossover on the randomly selected target individual and the mutation vector according to the crossover probability to generate a trial individual, that is, When then, ; otherwise, ; where, is a random number, CR is the crossover probability, is the index of the parameter dimension, is a random integer, is the target individual; The third iterative optimization unit is used to calculate the fitness of the target individual and the trial individual respectively according to the preset fitness function. The fitness function is, ; where, is the dynamic weight at each frequency point, is the model prediction value, is the measured value; If the fitness of the trial individual is less than the fitness of the target individual, let k = k + 1 and enter the next iteration, otherwise retain the target individual.

[0106] If the fitness of the trial individual is less than the fitness of the target individual, let k = k + 1 and enter the next iteration, otherwise retain the target individual.

[0107] The fourth iterative optimization unit is used to obtain a preset initial crossover probability and an initial scaling factor, and dynamically update the crossover probability and the scaling factor. The formula is ; ; where is the initial crossover probability, is the initial scaling factor, K is the total number of iterations, is the preset crossover lower limit.

[0108] In a specific feasible implementation, the optimal parameter output module includes the following units: The first optimal parameter output unit is used to store, during the iteration process, the individual with the best fitness in the global iteration process, denoted as the global optimal individual. The model parameters corresponding to the global optimal individual are the optimal parameters. When a new population is generated in each iteration, compare the fitness of the optimal individual in the current population with the fitness of the global optimal individual, and update with the better fitness pair; The second optimal parameter output unit is used to check, after each iteration ends, whether the current iteration number exceeds the preset iteration threshold, or whether the fitness of the current global optimal individual is less than the preset fitness threshold. If either of the above conditions is met, terminate the iteration; otherwise, continue the iteration, and determine whether to terminate the iteration according to the preset iteration termination condition; The third optimal parameter output unit is used to, after the iteration is terminated, perform parameter verification and correction based on the optimal parameters and the parameter range, and output the corrected optimal parameters.

[0109] In a specific feasible implementation, the equivalent circuit model parameter identification device based on EIS further includes the following modules: The parameter analysis module is used to, within the parameter range of each model parameter, draw the corresponding EIS curve graph in combination with the equivalent circuit model, and analyze the EIS curve graph to obtain the intuitive mapping relationship between the parameter and the graphical feature; perform comparison in combination with the optimal parameters and the intuitive mapping relationship to evaluate the health of the battery, and generate and output a health assessment report of the battery according to the result of the health assessment.

[0110] Based on the same inventive concept as above, an embodiment of the present application also discloses a computer-readable storage medium. At least one instruction, at least one program, a code set or an instruction set is stored in the storage medium, and at least one instruction, at least one program, a code set or an instruction set can be loaded and executed by a processor to implement the equivalent circuit model parameter identification method based on EIS provided in the above method embodiment.

[0111] Based on the same inventive concept as described above, an embodiment of the present application also discloses a computer-readable storage medium, in which at least one instruction, at least one program, a code set or an instruction set is stored, and the at least one instruction, at least one program, the code set or the instruction set is loaded and executed by a processor to implement the equivalent circuit model parameter identification method based on EIS as described above.

[0112] Those of ordinary skill in the art can understand that all or part of the steps to implement the above embodiments can be completed by hardware, or can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable storage medium, and the computer-readable storage medium includes, for example: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks or optical discs that can store program codes.

[0113] The above are only optional embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An equivalent circuit model parameter identification method based on EIS, characterized in that It includes the following steps: Obtain the EIS test data of the battery, screen the EIS data corresponding to the preset target SOC, and assign dynamic weights to the EIS data at different frequency points; Obtain the preset equivalent circuit model, set the parameter ranges of the model parameters in the equivalent circuit model, and randomly generate an initial population based on the logarithmic distribution within the parameter ranges; Iteratively optimize the initial population through the differential evolution algorithm, and dynamically adjust the crossover probability and scaling factor during the iteration process through an adaptive strategy; When the preset iteration termination condition is met, output the optimal parameters obtained after iteration.

2. The equivalent circuit model parameter identification method based on EIS according to claim 1, characterized in that The step of obtaining the EIS test data of the battery, screening the EIS data corresponding to the preset target SOC, and assigning dynamic weights to the EIS data at different frequency points specifically includes the following steps: Collect the EIS test data under different state of charge through the EIS detection chip mounted on the battery, screen the EIS data corresponding to the preset target SOC, and preprocess the collected EIS data; Divide the frequency range of the EIS data into multiple frequency intervals according to the preset frequency division rule, each frequency interval has a corresponding basic weight, obtain the measured data quality of each frequency point, and adjust the basic weight of each frequency point according to the measured data quality to obtain the dynamic weight.

3. The equivalent circuit model parameter identification method based on EIS according to claim 1, characterized in that The step of obtaining the preset equivalent circuit model, setting the parameter ranges of the model parameters in the equivalent circuit model, and randomly generating an initial population based on the logarithmic distribution within the parameter ranges specifically includes the following steps: Obtain a fractional-order equivalent circuit model including the Warburg impedance, the fractional-order equivalent circuit model includes at least one constant phase element and one Warburg element, The fractional-order equivalent circuit model is ； Among them, is the equivalent resistance, is the equivalent inductance, is the charge transfer resistance, and are the constant phase equivalent element parameters, is the pseudo-capacitance, is the fractional order exponent, 、 as well as are the Warburg element parameters; Determine the parameter ranges of the model parameters in the equivalent circuit model according to the physical characteristics of the battery; Within the parameter ranges, randomly generate an initial population through the logarithmic distribution, each initial population includes the model parameters to be identified, specifically, Generate a corresponding logarithmic space based on the parameter range of each model parameter, uniformly generate random numbers corresponding to the model parameters within each logarithmic space until an initial population of a preset scale is generated, the preset scale of the initial population is 1000, and the parameter dimension of the model parameters to be identified in each individual is 8.

4. The equivalent circuit model parameter identification method based on EIS according to claim 3, wherein The step of iteratively optimizing the initial population through the differential evolution algorithm and dynamically adjusting the crossover probability and scaling factor during the iteration process through an adaptive strategy specifically includes the following steps: Randomly select three different individuals from the currently generated initial population, , and calculate the mutation vector, that is, ； Where k is the iteration number and F is the scaling factor; Perform crossover on the randomly selected target individual and the mutant vector according to the crossover probability to generate a trial individual, that is, When then ; Otherwise, ; Among them, is a random number, CR is the crossover probability, is the index of the parameter dimension, is a random integer, is the target individual; Calculate the fitness of the target individual and the trial individual respectively according to the preset fitness function, and the fitness function is ； Among them, is the dynamic weight of each of the said frequency points, is the model prediction value, is the measured value; If the fitness of the trial individual is less than the fitness of the target individual, let k = k + 1 and enter the next iteration, otherwise retain the target individual.

5. The equivalent circuit model parameter identification method based on EIS according to claim 4, characterized in that, It also includes the following steps: Obtain the preset initial crossover probability and initial scaling factor, and dynamically update the crossover probability and the scaling factor. The formula is ； ； Among them, is the initial crossover probability, is the initial scaling factor, K is the total number of iterations, is the preset lower limit of crossover.

6. The equivalent circuit model parameter identification method based on EIS according to claim 4, characterized in that When the preset iteration termination condition is satisfied, output the optimal parameters obtained after iteration. The specific steps are as follows: During the iteration process, store the individual with the best fitness in the global iteration process, denoted as the global optimal individual. The model parameters corresponding to the global optimal individual are the optimal parameters. After generating a new population in each iteration, compare the fitness of the optimal individual in the current population with the fitness of the global optimal individual, and use the better fitness to update the global optimal individual. After each iteration ends, check whether the current iteration number exceeds the preset iteration threshold, or whether the fitness of the current global optimal individual is less than the preset fitness threshold. If either of the above conditions is satisfied, terminate the iteration; otherwise, continue the iteration, and judge whether to terminate the iteration according to the preset iteration termination condition. After terminating the iteration, perform parameter verification and correction based on the optimal parameters and the parameter range, and output the corrected optimal parameters.

7. The equivalent circuit model parameter identification method based on EIS according to claim 1, characterized in that After outputting the optimal parameters, the following steps are further included: Within the parameter range of each model parameter, combine the equivalent circuit model to draw the corresponding EIS curve diagram, and analyze the EIS curve diagram to obtain the intuitive mapping relationship between parameters and graphical features. Combine the optimal parameters for comparison and the intuitive mapping relationship to conduct a health assessment of the battery, and generate and output a health assessment report of the battery according to the results of the health assessment.

8. An equivalent circuit model parameter identification device based on EIS, characterized in that, The steps include: A data preprocessing module for obtaining the EIS test data of the battery, screening the EIS data corresponding to the preset target SOC, and assigning dynamic weights to the EIS data at different frequency points. A population initialization module for obtaining the preset equivalent circuit model, setting the parameter range of each model parameter in the equivalent circuit model, and randomly generating an initial population based on the logarithmic distribution within the parameter range. An iterative optimization module for iteratively optimizing the initial population through the differential evolution algorithm, and dynamically adjusting the crossover probability and scaling factor during the iteration through an adaptive strategy. An optimal parameter output module for outputting the optimal parameters obtained after iteration when the preset iteration termination condition is satisfied.

9. An intelligent terminal, characterized in that, It includes a memory and a processor. At least one instruction, at least one program, a code set or an instruction set is stored in the memory. The at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the method for identifying the parameters of the equivalent circuit model based on EIS according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, At least one instruction, at least one program, a code set or an instruction set is stored in the readable storage medium. The at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the method for identifying the parameters of the equivalent circuit model based on EIS according to any one of claims 1 to 7.

Images