Energy storage battery state evaluation method based on generated impedance spectrum and related device

Abstract

The invention belongs to the technical field of energy storage batteries, and discloses an energy storage battery state evaluation method based on a generated impedance spectrum and a related device. The method comprises the following steps: acquiring impedance data of a to-be-evaluated energy storage battery; equivalent circuit fitting is carried out based on the impedance data of the energy storage battery to be evaluated, and equivalent element parameters are obtained; the correlation degree between the equivalent element parameters and the battery health state is calculated, and parameters with the correlation degree larger than a threshold value are selected as characteristic parameters; and inputting the characteristic parameters into a pre-established energy storage battery SOH evaluation model based on the generative impedance spectroscopy to obtain the health state of the energy storage battery to be evaluated. According to the invention, the test speed of the AC impedance spectroscopy and the interpretability of the evaluation model are considered; a full-spectrum alternating-current impedance spectrum is generated by using an adversarial neural network through a few points in different frequency ranges, so that the problems that full-spectrum testing time is long and engineering application is not facilitated are solved; and performing equivalent circuit analysis by using the generated full spectrum, and performing health state evaluation of the energy storage battery.

Description

Technical Field

[0001] This invention belongs to the field of energy storage battery technology, and specifically relates to a method and related apparatus for energy storage battery state assessment based on generated impedance spectrum. Background Technology

[0002] With the transformation of the global energy structure and the rapid development of renewable energy, energy storage technology, as an important component of the energy system, is receiving increasing attention. Energy storage batteries, especially lithium-ion batteries, are widely used in power system peak shaving and valley filling, renewable energy grid integration, and electric vehicles due to their high efficiency, flexibility, and scalability. However, the performance of energy storage batteries directly affects the economics and reliability of the entire energy storage system. Therefore, accurately evaluating and predicting the State of Health (SOH) of energy storage batteries has become a key technical problem that the industry urgently needs to solve.

[0003] Currently, the state assessment of energy storage batteries mainly utilizes test data such as voltage and electrochemical impedance spectroscopy (EIS). EIS is an electrical measurement method that uses a small-amplitude sine wave as a perturbation signal. Combined with electrochemical theory and other related theories, it can obtain information about the test system in terms of materials science, kinetics, and reaction mechanisms. Because EIS has the advantage of reflecting the internal electrochemical reaction mechanism of the battery in situ and non-destructively, it is widely used in the mechanism analysis and state assessment of energy storage batteries. However, in the test frequency range of 10kHz to 0.01Hz, the test time is relatively long, with the test time for a single battery ≥20 minutes, making it unsuitable for engineering applications. To achieve engineering applications, the number of test frequency points is reduced, but this results in some information loss, leading to poor accuracy in the state estimation results of energy storage batteries. Summary of the Invention

[0004] The purpose of this invention is to provide a method and related apparatus for assessing the state of energy storage batteries based on generated impedance spectroscopy, which can shorten the AC impedance spectroscopy test time while ensuring the accuracy of energy storage battery state assessment.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for assessing the state of a storage battery based on generated impedance spectroscopy, comprising: Obtain the impedance full spectrum data of the energy storage battery to be evaluated; Equivalent circuit fitting is performed based on the impedance full spectrum data of the energy storage battery to be evaluated to obtain the equivalent component parameters; Calculate the correlation between equivalent component parameters and battery health status, and select parameters with a correlation greater than a threshold as feature parameters; The characteristic parameters are input into a pre-established energy storage battery SOH evaluation model based on generative impedance spectroscopy to obtain the health status of the energy storage battery to be evaluated. The step of obtaining the impedance full spectrum data of the energy storage battery to be evaluated specifically includes: Within the range of 1000Hz to 0.1Hz, the impedance at two frequencies is tested at each order of magnitude to obtain the impedance of the energy storage battery under evaluation at eight frequencies. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The 8×4 matrix is ​​input into a pre-trained generative impedance model to obtain impedance data of a 70×4 matrix representing the full spectrum of the energy storage battery to be evaluated.

[0006] A further improvement of this invention is that the pre-trained generative impedance model is obtained through the following steps: The energy storage battery is charged and discharged at a preset rate to obtain the rated capacity; After calibration, impedance was tested at two frequencies per order of magnitude within the range of 1000Hz to 0.1Hz to obtain the impedance of the energy storage battery at eight frequencies for different cycle counts until the calibrated capacity of the energy storage battery decayed to below 80% SOH. The full-spectrum AC impedance spectrum was collected simultaneously. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The impedance data of the full spectrum forms a 70×4 matrix. The 8×4 matrix with different SOH and the 100-dimensional Gaussian distributed noise vector are used as input to the generative impedance model, and the 70×4 matrix is ​​used as output to train the generative impedance model until the loss function converges, thus obtaining the pre-trained generative impedance model. The loss function is a hybrid loss function that combines generator loss and discriminator loss.

[0007] A further improvement of the present invention is that: in the step of testing the impedance at two frequencies per order of magnitude within the range of 1000Hz to 0.1Hz to obtain the impedance of the energy storage battery to be evaluated at eight frequencies; the impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix, and the eight frequencies are: 601 Hz, 186 Hz, 73 Hz, 23 Hz, 7 Hz, 2 Hz, 0.6 Hz and 0.2 Hz.

[0008] A further improvement of the present invention is that the pre-trained generative impedance model adopts a conditional generative adversarial neural network, which consists of a generator and a discriminator; the generator adopts an encoder-decoder structure, a fully connected network, or a one-dimensional convolution; the discriminator adopts a fully connected network or LSTM.

[0009] A further improvement of this invention is that the step of performing equivalent circuit fitting based on the impedance data of the energy storage battery to be evaluated to obtain equivalent component parameters specifically includes: The impedance data are fitted to an equivalent circuit using the least squares method to obtain the inductance L and the internal resistance R in ohms. s Charge transfer internal resistance R ct The parameters Y0 and n of the constant phase angle element Q, and the parameter Y of the Weber impedance W. 0w ; The grey relational analysis method is used to calculate the correlation between equivalent component parameters and battery health status, and parameters with a correlation greater than a threshold are selected as feature parameters.

[0010] A further improvement of the present invention is that: in the step of calculating the correlation between the equivalent element parameters and the battery health status, and selecting parameters with a correlation greater than a threshold as feature parameters, the threshold is 0.8.

[0011] A further improvement of this invention is that the pre-established SOH evaluation model for energy storage batteries based on generative impedance spectroscopy is established through the following steps: By using characteristic parameters as input parameters and health status as output parameters in the health status estimation model, a neural network algorithm is employed to construct a state of health (SOH) assessment model for energy storage batteries based on generative impedance spectroscopy.

[0012] Secondly, the present invention provides a battery state assessment device based on generated impedance spectrum, comprising: The acquisition module is used to acquire the impedance full spectrum data of the energy storage battery to be evaluated; The fitting module is used to perform equivalent circuit fitting based on the impedance full spectrum data of the energy storage battery to be evaluated, and obtain the equivalent component parameters. The feature module is used to calculate the correlation between equivalent component parameters and battery health status, and selects parameters with a correlation greater than a threshold as feature parameters. The evaluation module is used to input the characteristic parameters into a pre-established energy storage battery SOH evaluation model based on generative impedance spectroscopy to obtain the health status of the energy storage battery to be evaluated. The step of obtaining the impedance full spectrum data of the energy storage battery to be evaluated specifically includes: Within the range of 1000Hz to 0.1Hz, the impedance at two frequencies is tested at each order of magnitude to obtain the impedance of the energy storage battery under evaluation at eight frequencies. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The 8×4 matrix is ​​input into a pre-trained generative impedance model to obtain impedance data of a 70×4 matrix representing the full spectrum of the energy storage battery to be evaluated.

[0013] Thirdly, the present invention provides an electronic device including a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the aforementioned method for assessing the state of a storage battery based on generated impedance spectroscopy.

[0014] Fourthly, the present invention provides a computer-readable storage medium storing at least one instruction, which, when executed by a processor, implements the aforementioned method for assessing the state of a storage battery based on generated impedance spectra.

[0015] Compared with the prior art, the present invention has the following unexpected beneficial effects: This invention provides a method for assessing the state of energy storage batteries based on generated impedance spectroscopy, comprising: acquiring impedance data of the energy storage battery to be evaluated; performing equivalent circuit fitting based on the impedance data of the energy storage battery to be evaluated to obtain equivalent element parameters; calculating the correlation between the equivalent element parameters and the battery's state of health, and selecting parameters with a correlation greater than a threshold as feature parameters; inputting the feature parameters into a pre-established energy storage battery SOH assessment model based on generated impedance spectroscopy to obtain the state of health of the energy storage battery to be evaluated. This invention balances the speed of AC impedance spectroscopy testing with the interpretability of the assessment model; by using a few points within different frequency ranges and employing adversarial neural networks, it generates a full-spectrum AC impedance spectrum, solving the problems of long testing times and difficulty in engineering applications; after performing equivalent circuit analysis using the generated full spectrum, it assesses the state of health of the energy storage battery, solving the problem that data-driven models are difficult to match with degradation mechanisms, making the model more scientific. This invention overcomes the contradiction between the time-consuming full-spectrum impedance testing and the low accuracy of limited-frequency testing in traditional assessments. By generating an impedance spectrum combined with parameter screening and model evaluation, it leverages the advantage of impedance data to reflect the internal electrochemical reaction mechanism of the battery in situ and non-destructively, while reducing invalid data interference through characteristic parameter screening. This allows the SOH assessment to balance efficiency and accuracy, and also aligns the assessment process with the battery degradation mechanism, improving the scientific rigor and interpretability of the assessment model. This invention employs characteristic frequency testing combined with generative impedance to obtain the full spectrum, and uses the full spectrum for equivalent circuit fitting to establish an energy storage battery state assessment model. To shorten the AC impedance spectrum testing time while ensuring the accuracy of energy storage battery state assessment, this invention measures a limited number of frequencies, uses generative algorithms based on these frequencies to generate the full AC impedance spectrum, and then uses the full spectrum for energy storage battery state estimation.

[0016] Furthermore, this invention defines a test scheme for eight characteristic frequencies within the range of 1000Hz - 0.1Hz, and generates a 70×4 full-spectrum matrix from an 8×4 matrix using a generative model. This feature significantly shortens the impedance testing time, avoiding the problem of traditional full-spectrum testing where a single battery test takes over 20 minutes, making it suitable for engineering applications. Simultaneously, the eight frequencies cover various orders of magnitude and contain multi-dimensional data. These eight frequencies represent the kinetic reactions occurring inside the battery at different relaxation times, mainly including processes such as the double-layer interface, lithium-ion penetration through the SEI film, and lithium-ion transfer within the electrode material. This provides sufficient basic information for the generative model, ensuring the integrity and reliability of the generated full-spectrum impedance data, providing high-quality data support for subsequent equivalent circuit fitting, and avoiding the information loss problem caused by testing with fewer frequencies.

[0017] Furthermore, this invention allows the battery to be cycled until its capacity decays to 80% SOH, collecting datasets under different health conditions to adapt the model to the impedance generation requirements throughout the battery's entire lifecycle. The introduction of a 100-dimensional Gaussian noise vector enhances the model's adaptability to data fluctuations. The hybrid loss function of the generator and discriminator drives rapid model convergence, ensuring a high degree of consistency between the generated full-spectrum data and the real full-spectrum data, significantly improving the accuracy of impedance full-spectrum generation and laying a solid model foundation for subsequent evaluation.

[0018] Furthermore, the generative impedance model in this invention employs a conditional generative adversarial neural network and provides multiple optional structures for the generator and discriminator. This technical feature allows the model to be highly adaptable. For small-sized data such as 8×4 matrices, a simple structure such as a fully connected network can be used to reduce computational costs; if improved feature extraction capabilities are required, an encoder-decoder structure or LSTM can also be employed. The multiple discriminator structure options can accurately verify the authenticity of the generated data. The combination of these two elements ensures both model computational efficiency and the quality of the generated full spectrum, adapting to application scenarios with different hardware conditions and accuracy requirements.

[0019] Furthermore, this invention uses the least squares method to fit the equivalent circuit parameters and employs the grey relational analysis method to screen feature parameters. The least squares method can efficiently and accurately extract key component parameters such as inductance and various internal resistances from full-spectrum impedance data. These parameters are directly related to core reaction processes such as charge transfer and ohmic loss within the battery. The grey relational analysis method can effectively quantify the correlation between each parameter and SOH, accurately screen parameters sensitive to battery health status, eliminate irrelevant or low-correlation parameters, reduce the interference of redundant data on the evaluation model, improve the efficiency and accuracy of subsequent SOH evaluation, and simultaneously link feature parameters closely with the battery health degradation pattern, enhancing the logic of the evaluation.

[0020] Furthermore, this invention sets the correlation threshold to 0.8, providing a clear and scientific standard for feature parameter screening. This threshold can accurately screen out core parameters strongly correlated with SOH, avoiding the problems of increased model complexity and evaluation error due to the introduction of weakly correlated parameters due to an excessively low threshold, or the loss of key information due to an excessively high threshold. This keeps the relative error of model validation within 2%, significantly improving the stability and reliability of SOH evaluation, ensuring consistent evaluation standards for batteries of different batches and states, and enhancing the versatility of the method.

[0021] Furthermore, in this invention, the SOH evaluation model takes feature parameters as input and SOH as output, and is constructed using a neural network algorithm. The neural network algorithm possesses powerful nonlinear fitting capabilities, accurately capturing the complex mapping relationship between feature parameters and battery health status. By using the selected highly correlated feature parameters as input, compared to directly using the original impedance data, the input dimensionality and noise interference are reduced, lowering the difficulty of model training and accelerating the training speed. Attached Figure Description

[0022] The accompanying drawings, which form part of this specification, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 Equivalent circuit diagram; Figure 2 A schematic diagram of the impedance full spectrum generation process; Figure 3 This is a flowchart illustrating a method for assessing the state of an energy storage battery based on generated impedance spectroscopy, according to an embodiment of the present invention. Figure 4 This is a schematic diagram of the structure of an energy storage battery state assessment device based on generated impedance spectrum according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0023] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0024] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this invention is for describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0025] This invention provides a method for assessing the state of energy storage batteries based on generated impedance spectra, comprising: (1) Impedance test at characteristic frequency 1) Energy storage battery cycle The energy storage battery is cycled under certain operating conditions (rate of 0.3-1.0C or 0.3-1.0P), and its capacity is calibrated once every certain number of cycles (100 or 200 cycles).

[0026] 2) Energy storage battery capacity calibration Under operating conditions of 0.5C or 0.5P, with a voltage range of 2.5~3.65V, the energy storage battery is charged and discharged three times, and the discharge capacity of the last discharge is selected as the rated capacity (initial battery capacity).

[0027] 3) Determine the testing frequency After calibration, two frequencies were tested at each order of magnitude within the range of 1000Hz to 0.1Hz, for a total of eight frequencies. The impedance at these eight frequencies was independent of the ambient temperature and the state of charge (SOC) of the battery.

[0028] In one specific implementation, eight frequencies were selected for different number of cycles: 601 and 186 Hz (1000~100Hz), 73 and 23 (100~10Hz), 7 and 2 (10~1Hz), and 0.6 and 0.2 (1~0.1Hz).

[0029] 4) Impedance under different health states (SOH) The impedance of the energy storage battery was tested at eight frequencies for different cycle counts until the battery's rated capacity decayed to below 80% SOH. The formula for the state of health is as follows:

[0030] Where Qr is the current battery capacity and Qi is the initial battery capacity.

[0031] 5) Full-spectrum impedance testing To verify the accuracy of the subsequent generative impedance and to serve as a label for the model, a full-spectrum AC impedance spectrum was simultaneously acquired.

[0032] (2) Construction of generative impedance model 1) Data format Impedance data at 8 frequencies includes frequency, real part, imaginary part, and phase angle, forming an 8×4 matrix. Impedance data across the entire spectrum forms a 70×4 matrix.

[0033] 2) Model An 8×4 matrix with different SOH values ​​and a 100-dimensional Gaussian distributed noise vector are used as inputs, and a 70×4 matrix is ​​used as the output. The generative impedance model architecture adopts a conditional generative adversarial neural network (cGAN), which mainly consists of a generator and a discriminator. After the generator generates the target matrix, the discriminator compares it with the real data (a 70×4 matrix formed from the measured full spectrum) until the discriminator can no longer distinguish between real and fake data, that is, the loss function meets the preset conditions, and the training ends.

[0034] The generator employs an encoder-decoder structure, such as a long short neural network (LSTM) or a Transformer. However, considering the small matrix size, a fully connected network (Dense) or a one-dimensional convolutional network (Conv1D) can also be used. The discriminator can be either a fully connected network or an LSTM.

[0035] The loss function is a hybrid loss function that combines generator loss and discriminator loss.

[0036] (3) Health status estimation model 1) Feature extraction The impedance data generated by the generative impedance model, i.e., a 70×4 matrix, is used to fit an equivalent circuit using the least squares method to obtain the inductance L and the internal resistance R in ohms. s Charge transfer internal resistance R ct The parameters Y0 and n of the constant phase angle element Q (describing the degree of deviation from the capacitive element), and the parameter Y of the Weber impedance W. 0w .

[0037] Using methods such as grey relational analysis, the correlation between equivalent component parameters and battery state of health (SOH) is calculated, and parameters with a correlation of ≥0.8 are selected as feature parameters for battery state of health estimation.

[0038] 2) Model building The selected feature parameters are used as input parameters for the health status estimation model, and the health status is used as the output parameter. A systemic impedance spectroscopy (SOH) evaluation model for energy storage batteries is constructed using neural network algorithms.

[0039] (4) Health status estimation The impedance of the energy storage battery to be evaluated at eight frequencies was obtained using the method in step (1); Using the method in step (2), the impedance of the energy storage battery to be evaluated at 8 frequencies is input into the generative impedance model to generate full-spectrum impedance data of the energy storage battery to be evaluated. Using the method in step (3), equivalent circuit fitting is performed based on the full-spectrum impedance data of the energy storage battery to be evaluated to obtain equivalent component parameters; the correlation between the equivalent component parameters and the battery health status is calculated, and parameters with a correlation greater than the threshold are selected as feature parameters. The characteristic parameters are input into the energy storage battery SOH evaluation model based on generative impedance spectrum established in step (3) to obtain the health status of the energy storage battery to be evaluated.

[0040] This invention provides a method for assessing the state of energy storage batteries based on generated impedance spectra, comprising: 1) Impedance test at characteristic frequency A brand new 60Ah lithium iron phosphate battery was taken and calibrated. Three charge-discharge tests were conducted at a constant power of 0.5P (96W), and the initial actual capacity was found to be 59.33Ah.

[0041] The battery was charged and discharged at a constant power of 0.3P (64W) for 100 cycles. After the cycle, the battery was calibrated at a power of 0.5P (96W). After calibration, an AC impedance test was performed every 5% of SOC (SOC is the state of charge). This cycle was repeated until the battery degraded to 80% SOH.

[0042] The frequencies, real parts, imaginary parts, and phase angles of 601 and 186 Hz (1000~100Hz), 73 and 23 (100~10Hz), 7 and 2 (10~1Hz), and 0.6 and 0.2 (1~0.1Hz) at different cycle numbers are used to construct 4800 sets of 8×4 matrices. The frequencies, real parts, imaginary parts, and phase angles of the entire spectrum are used to construct 4800 sets of 70×4 matrices.

[0043] 2) Construction of Generative Impedance Model The model uses 4800 sets of 8×4 matrices as input parameters and 4800 sets of 70×4 matrices as output parameters. The model architecture adopts a conditional generative adversarial neural network (cGAN), which mainly consists of a generator and a discriminator. After the generator generates the target matrix, the discriminator compares it with the real data (70×4 matrices formed by the measured full spectrum) until the discriminator can no longer distinguish between real and fake data, that is, the loss function meets the preset conditions, and the training ends.

[0044] The generator adopts a Long Short Memory Neural Network (LSTM) encoder-decoder structure, which mainly includes a bidirectional LSTM encoder: processing the condition matrix to generate a context vector, a noise fusion layer: splicing the context vector with Gaussian noise, and a unidirectional LSTM decoder: outputting the target matrix.

[0045] The discriminator also uses an LSTM structure, which mainly includes a matrix concatenation module: merging the condition matrix and the target matrix along the feature dimension, and a spectral normalized bidirectional LSTM feature extractor.

[0046] The loss function adopts a hybrid loss function, which mainly includes adversarial loss function, content loss function (L1 norm) and dynamic weight adjuster (λ∈[5,20]).

[0047] 2) Health status estimation model Full-spectrum data of the battery under different health states were obtained using a generative impedance model, and an equivalent circuit was constructed for all data. Figure 1 By fitting the data, the inductance L and the internal resistance R in ohms are obtained. s Charge transfer internal resistance R ct The parameters Y0 and n of the constant phase angle element Q, and the parameter Y of the Weber impedance W. 0w Using the grey relational analysis method, the correlation between the four parameters Rs, Rct, Y0, and Y and the battery's state of health (SOH) was found to be greater than 0.8. Therefore, these four parameters were used as feature parameters and input parameters, with a BP neural network model as the core algorithm and SOH as the output parameter, to establish a battery health state estimation model. The model was validated using 10 sets of data that were not used in training, and the relative errors were all within 2%.

[0048] Please see Figure 3 This invention provides a method for assessing the state of energy storage batteries based on generated impedance spectra, comprising: S100. Obtain the impedance full spectrum data of the energy storage battery to be evaluated; S200. Based on the impedance full spectrum data of the energy storage battery to be evaluated, perform equivalent circuit fitting to obtain equivalent component parameters; S300: Calculate the correlation between equivalent element parameters and battery health status, and select parameters with a correlation greater than the threshold as feature parameters. S400. Input the characteristic parameters into a pre-established energy storage battery SOH evaluation model based on generative impedance spectrum to obtain the health status of the energy storage battery to be evaluated. Step S100 specifically includes: Within the range of 1000Hz to 0.1Hz, the impedance at two frequencies is tested at each order of magnitude to obtain the impedance of the energy storage battery under evaluation at eight frequencies. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The 8×4 matrix is ​​input into a pre-trained generative impedance model to obtain impedance data of a 70×4 matrix representing the full spectrum of the energy storage battery to be evaluated.

[0049] In one specific embodiment, the step of obtaining impedance data of the energy storage battery to be evaluated in S100 specifically includes: Obtain the initial battery capacity of the energy storage battery to be evaluated; Within the range of 1000Hz to 0.1Hz, the impedance at two frequencies is tested at each order of magnitude to obtain the impedance of the energy storage battery under evaluation at eight frequencies. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The 8×4 matrix is ​​input into a pre-trained generative impedance model to obtain impedance data of a 70×4 matrix representing the full spectrum of the energy storage battery to be evaluated.

[0050] In one specific implementation, the pre-trained generative impedance model described in S100 is obtained through the following steps: The energy storage battery is charged and discharged at a preset rate to obtain the rated capacity; After calibration, impedance was tested at two frequencies per order of magnitude within the range of 1000Hz to 0.1Hz to obtain the impedance of the energy storage battery at eight frequencies for different cycle counts until the calibrated capacity of the energy storage battery decayed to below 80% SOH. The full-spectrum AC impedance spectrum was collected simultaneously. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The impedance data of the full spectrum forms a 70×4 matrix. The 8×4 matrix with different SOH and the 100-dimensional Gaussian distributed noise vector are used as input to the generative impedance model, and the 70×4 matrix is ​​used as output to train the generative impedance model until the loss function converges, thus obtaining the pre-trained generative impedance model. The loss function is a hybrid loss function that combines generator loss and discriminator loss.

[0051] In one specific implementation, the pre-trained generative impedance model in S100 adopts a conditional generative adversarial neural network, which consists of a generator and a discriminator; the generator adopts an encoder-decoder structure, a fully connected network, or a one-dimensional convolution; the discriminator adopts a fully connected network or LSTM.

[0052] In one specific embodiment, the step S200 of performing equivalent circuit fitting based on the impedance data of the energy storage battery to be evaluated to obtain equivalent component parameters specifically includes: The impedance data are fitted to an equivalent circuit using the least squares method to obtain the inductance L and the internal resistance R in ohms. s Charge transfer internal resistance R ct The parameters Y0 and n of the constant phase angle element Q, and the parameter Y of the Weber impedance W. 0w ; The grey relational analysis method is used to calculate the correlation between equivalent component parameters and battery health status, and parameters with a correlation greater than a threshold are selected as feature parameters.

[0053] In one specific embodiment, in step S300, which calculates the correlation between equivalent element parameters and battery health status and selects parameters with a correlation greater than a threshold as feature parameters, the threshold is 0.8.

[0054] In one specific implementation, the pre-established energy storage battery SOH evaluation model based on generative impedance spectroscopy described in S400 is established through the following steps: using feature parameters as input parameters of the health state estimation model and health state as output parameters, and employing a neural network algorithm to construct the energy storage battery SOH evaluation model based on generative impedance spectroscopy. Please see Figure 4 The present invention provides a battery state assessment device based on generated impedance spectrum, comprising: The acquisition module is used to acquire the impedance full spectrum data of the energy storage battery to be evaluated; The fitting module is used to perform equivalent circuit fitting based on the impedance full spectrum data of the energy storage battery to be evaluated, and obtain the equivalent component parameters. The feature module is used to calculate the correlation between equivalent component parameters and battery health status, and selects parameters with a correlation greater than a threshold as feature parameters. The evaluation module is used to input the characteristic parameters into a pre-established energy storage battery SOH evaluation model based on generative impedance spectroscopy to obtain the health status of the energy storage battery to be evaluated. The acquisition module is specifically configured as follows: Within the range of 1000Hz to 0.1Hz, the impedance at two frequencies is tested at each order of magnitude to obtain the impedance of the energy storage battery under evaluation at eight frequencies. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The 8×4 matrix is ​​input into a pre-trained generative impedance model to obtain impedance data of a 70×4 matrix representing the full spectrum of the energy storage battery to be evaluated.

[0055] Please see Figure 5 As shown, this embodiment of the invention provides an electronic device 100 for implementing a method for assessing the state of a storage battery based on generated impedance spectrum; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0056] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the energy storage battery state assessment method based on generated impedance spectrum as described in the embodiment by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0057] The at least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor. The processor 102 is the control center of the electronic device 100, connecting various parts of the electronic device 100 via various interfaces and lines.

[0058] The memory 101 in the electronic device 100 stores multiple instructions to implement a method for assessing the state of an energy storage battery based on generated impedance spectra, and the processor 102 can execute the multiple instructions to achieve the following: Obtain the impedance full spectrum data of the energy storage battery to be evaluated; Equivalent circuit fitting is performed based on the impedance full spectrum data of the energy storage battery to be evaluated to obtain the equivalent component parameters; Calculate the correlation between equivalent component parameters and battery health status, and select parameters with a correlation greater than a threshold as feature parameters; The characteristic parameters are input into a pre-established energy storage battery SOH evaluation model based on generative impedance spectroscopy to obtain the health status of the energy storage battery to be evaluated. The step of obtaining the impedance full spectrum data of the energy storage battery to be evaluated specifically includes: Within the range of 1000Hz to 0.1Hz, the impedance at two frequencies is tested at each order of magnitude to obtain the impedance of the energy storage battery under evaluation at eight frequencies. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The 8×4 matrix is ​​input into a pre-trained generative impedance model to obtain impedance data of a 70×4 matrix representing the full spectrum of the energy storage battery to be evaluated.

[0059] If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0060] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0061] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0062] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0063] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for assessing the state of energy storage batteries based on generated impedance spectroscopy, characterized in that, include: Obtain the impedance full spectrum data of the energy storage battery to be evaluated; Equivalent circuit fitting is performed based on the impedance full spectrum data of the energy storage battery to be evaluated to obtain the equivalent component parameters; Calculate the correlation between equivalent component parameters and battery health status, and select parameters with a correlation greater than a threshold as feature parameters; The characteristic parameters are input into a pre-established energy storage battery SOH evaluation model based on generative impedance spectroscopy to obtain the health status of the energy storage battery to be evaluated. The step of obtaining the impedance full spectrum data of the energy storage battery to be evaluated specifically includes: Within the range of 1000Hz to 0.1Hz, the impedance at two frequencies is tested at each order of magnitude to obtain the impedance of the energy storage battery under evaluation at eight frequencies. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The 8×4 matrix is ​​input into a pre-trained generative impedance model to obtain impedance data of a 70×4 matrix representing the full spectrum of the energy storage battery to be evaluated.

2. The method for assessing the state of a storage battery based on generated impedance spectroscopy according to claim 1, characterized in that, The pre-trained generative impedance model is obtained through the following steps: The energy storage battery is charged and discharged at a preset rate to obtain the rated capacity; After calibration, impedance was tested at two frequencies per order of magnitude within the range of 1000Hz to 0.1Hz to obtain the impedance of the energy storage battery at eight frequencies for different cycle counts until the calibrated capacity of the energy storage battery decayed to below 80% SOH. The full-spectrum AC impedance spectrum was collected simultaneously. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The impedance data of the full spectrum forms a 70×4 matrix. The 8×4 matrix with different SOH and the 100-dimensional Gaussian distributed noise vector are used as input to the generative impedance model, and the 70×4 matrix is ​​used as output to train the generative impedance model until the loss function converges, thus obtaining the pre-trained generative impedance model. The loss function is a hybrid loss function that combines generator loss and discriminator loss.

3. The method for assessing the state of a storage battery based on generated impedance spectroscopy according to claim 1, characterized in that, In the step of testing the impedance at two frequencies per order of magnitude within the range of 1000Hz to 0.1Hz to obtain the impedance of the energy storage battery to be evaluated at eight frequencies, the impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The eight frequencies are: 601 Hz, 186 Hz, 73 Hz, 23 Hz, 7 Hz, 2 Hz, 0.6 Hz and 0.2 Hz.

4. The method for assessing the state of a storage battery based on generated impedance spectroscopy according to claim 2, characterized in that, The pre-trained generative impedance model employs a conditional generative adversarial neural network, consisting of a generator and a discriminator. The generator uses an encoder-decoder structure, a fully connected network, or a one-dimensional convolution. The discriminator uses a fully connected network or an LSTM.

5. The method for assessing the state of a storage battery based on generated impedance spectroscopy according to claim 1, characterized in that, The step of performing equivalent circuit fitting based on the impedance data of the energy storage battery to be evaluated to obtain equivalent component parameters specifically includes: The impedance data are fitted to an equivalent circuit using the least squares method to obtain the inductance L and the internal resistance R in ohms. s Charge transfer internal resistance R ct The parameters Y0 and n of the constant phase angle element Q, and the parameter Y of the Weber impedance W. 0w ; The grey relational analysis method is used to calculate the correlation between equivalent component parameters and battery health status, and parameters with a correlation greater than a threshold are selected as feature parameters.

6. The method for assessing the state of a storage battery based on generated impedance spectroscopy according to claim 1, characterized in that, In the step of calculating the correlation between equivalent element parameters and battery health status, and selecting parameters with a correlation greater than a threshold as feature parameters, the threshold is 0.

8.

7. The method for assessing the state of a storage battery based on generated impedance spectroscopy according to claim 1, characterized in that, The pre-established SOH evaluation model for energy storage batteries based on generative impedance spectroscopy is established through the following steps: By using characteristic parameters as input parameters and health status as output parameters in the health status estimation model, a neural network algorithm is employed to construct a state of health (SOH) assessment model for energy storage batteries based on generative impedance spectroscopy.

8. A state assessment device for energy storage batteries based on generated impedance spectroscopy, characterized in that, include: The acquisition module is used to acquire the impedance full spectrum data of the energy storage battery to be evaluated; The fitting module is used to perform equivalent circuit fitting based on the impedance full spectrum data of the energy storage battery to be evaluated, and obtain the equivalent component parameters. The feature module is used to calculate the correlation between equivalent component parameters and battery health status, and selects parameters with a correlation greater than a threshold as feature parameters. The evaluation module is used to input the characteristic parameters into a pre-established energy storage battery SOH evaluation model based on generative impedance spectroscopy to obtain the health status of the energy storage battery to be evaluated. The step of acquiring impedance full-spectrum data of the energy storage battery to be evaluated specifically includes: Within the range of 1000Hz to 0.1Hz, the impedance at two frequencies is tested at each order of magnitude to obtain the impedance of the energy storage battery under evaluation at eight frequencies. The impedance at the eight frequencies includes the frequency, real part, imaginary part and phase angle, forming an 8×4 matrix. The 8×4 matrix is ​​input into a pre-trained generative impedance model to obtain impedance data of a 70×4 matrix representing the full spectrum of the energy storage battery to be evaluated.

9. An electronic device, characterized in that, It includes a processor and a memory, the processor being used to execute a computer program stored in the memory to implement a method for assessing the state of a storage battery based on generated impedance spectroscopy as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction, which, when executed by a processor, implements a method for assessing the state of an energy storage battery based on a generated impedance spectrum as described in any one of claims 1 to 7.

Images