Battery manufacturing quality detection and grade utilization method and device based on ultrasonic characteristics

Abstract

The application discloses a battery manufacturing quality detection and grade utilization method and device based on ultrasonic characteristics, relates to the technical field of battery manufacturing process monitoring and retired battery sorting detection, and performs characteristic enhancement on an ultrasonic frequency domain response spectrum of a battery to be detected, combines battery capacity data, and constructs a multi-dimensional acoustic characteristic data set; based on amplitude and phase decomposition of the ultrasonic frequency domain response spectrum of the multi-dimensional acoustic characteristic data set, equivalent acoustic impedance of the battery is defined by analogy to the concept of electrochemical complex impedance, an equivalent physical model of the ultrasonic frequency domain response spectrum of the battery is constructed; based on the equivalent physical model of the ultrasonic frequency domain response spectrum of the battery and the ultrasonic frequency domain response spectrum, optimal equivalent acoustic parameters are identified through parameter optimization; battery defects are judged based on the distribution characteristics of the optimal equivalent acoustic parameters, and the battery is graded according to a battery capacity prediction value. The application realizes accurate identification of battery manufacturing quality defects, rapid evaluation and grade utilization sorting of residual performance of retired batteries.

Description

Technical Field

[0001] This invention relates to the field of battery manufacturing process monitoring and retired battery sorting and testing technology, and in particular to a method and apparatus for battery manufacturing quality testing and cascade utilization based on ultrasonic features. Background Technology

[0002] The statements in this section are merely background information relating to this disclosure and do not necessarily constitute prior art.

[0003] As the core power source for electric vehicles, large-scale energy storage systems, and consumer electronics, the consistency of lithium-ion battery manufacturing quality and the accuracy of sorting for reuse after retirement are crucial for ensuring system operation safety and improving the economic benefits of the industry chain. In the manufacturing process, the internal physical structure of the battery (such as electrolyte wetting degree, electrode stacking tightness, and packaging pressure) directly determines its factory performance. However, in the retirement process, due to differences in service conditions, the nonlinear evolution of the battery's internal physical structure leads to highly discrete residual performance, posing a significant challenge to efficient sorting.

[0004] Traditional manufacturing inspection and sorting methods mainly include electrical parameter-based methods, X-ray or computed tomography (CT) methods, and static disassembly sampling methods, but all have certain limitations. Electrical parameter-based methods mainly use capacity, internal resistance, and voltage curves for quality control or sorting. However, electrical parameters are not sensitive to minor physical defects in the early stages of battery manufacturing (such as poor local wetting), and during the sorting process, it is difficult to distinguish retired batteries with similar capacities but different internal physical damage mechanisms based solely on electrical indicators, making it difficult to guarantee the safety of cascade utilization. Although X-ray or CT scans can directly observe the internal structure of batteries, the equipment is extremely expensive, the detection speed is slow, and there are radiation risks, making it difficult to integrate into high-speed production lines for full-volume quality inspection, and it cannot meet the rapid sorting needs of large-scale retired batteries. Static disassembly sampling methods assess manufacturing quality by destructively disassembling batches of batteries, but this method is not only costly and has limited representativeness, but also, due to its destructive nature, it cannot be applied to retired batteries awaiting sorting, making it difficult to achieve closed-loop management of the entire battery life cycle.

[0005] In recent years, ultrasonic testing technology, as a non-destructive testing method, has been increasingly applied to battery condition assessment due to its high sensitivity to battery material density, elastic modulus, and interface contact state. However, existing ultrasonic testing technologies still have some technical shortcomings in manufacturing inspection and graded sorting: Insufficient time-domain feature resolution: Existing methods mostly focus on analyzing the time-of-flight (ToF) and signal amplitude of ultrasonic waves. These indicators are easily affected by environmental coupling fluctuations and lack sufficient ability to identify micron-level structural defects and the complex physical evolution of retired batteries during battery manufacturing; Lack of deep feature mining: Existing ultrasonic monitoring methods fail to fully utilize the frequency domain information of the interaction between ultrasonic waves and the multi-layered, multi-phase structure of the battery at different frequencies, resulting in the inability to extract core feature vectors that can characterize manufacturing quality or sorting grade from complex acoustic signals; Lack of physical mapping relationship: Due to the lack of effective acoustic forward models and parameter inversion algorithms, it is difficult to transform the collected acoustic features into essential parameters characterizing the internal physical quality of the battery in real time, limiting its application in online quality inspection on production lines and automated sorting of retired batteries. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, this invention provides a method and apparatus for battery manufacturing quality inspection and secondary utilization based on ultrasonic features. It can deeply characterize the evolution of battery physical structure, has a rigorous physical logic support and high precision, and realizes accurate inspection of battery manufacturing quality and efficient sorting for secondary utilization.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: In a first aspect, the present invention provides a method for battery manufacturing quality inspection and secondary utilization based on ultrasonic features, comprising: The ultrasonic response signals of the battery under test under ultrasonic excitation signals of different frequencies were obtained and frequency domain analysis was performed to construct the ultrasonic frequency domain response spectrum of the battery under test. The ultrasonic frequency response spectrum is enhanced with features, and combined with battery capacity data, a multidimensional acoustic feature dataset is constructed. Based on the amplitude-phase decomposition of the ultrasonic frequency domain response spectrum using the multidimensional acoustic feature dataset, the equivalent acoustic impedance of the battery is defined by analogy with the complex impedance concept in electrochemistry, and an equivalent physical model of the ultrasonic frequency domain response spectrum of the battery is constructed. Based on the equivalent physical model and ultrasonic frequency domain response spectrum of the battery, the optimal equivalent acoustic parameters are identified through parameter optimization. Battery defects are determined based on the distribution characteristics of the optimal equivalent acoustic parameters, and battery capacity is predicted based on the optimal equivalent acoustic parameters. The batteries are then classified according to the predicted battery capacity values.

[0008] A further technical solution, wherein the construction of the ultrasonic frequency domain response spectrum of the battery under test specifically includes: The ultrasonic excitation signal and ultrasonic response signal are transformed in the frequency domain to obtain the spectrum of the excitation signal and the spectrum of the response signal, respectively. The ultrasonic frequency domain response function of the battery is calculated based on the spectrum of the excitation signal and the spectrum of the response signal. The amplitude and phase of the ultrasonic frequency domain response function are extracted, and then the propagation characteristic parameters are extracted. Based on the amplitude, phase, and propagation characteristic parameters extracted at each frequency point, a multidimensional ultrasonic frequency domain response spectrum feature vector of the battery under test is constructed.

[0009] A further technical solution involves equivalently modeling the internal structural parameters of the battery based on the equivalent acoustic impedance, and constructing an equivalent physical model of the battery's ultrasonic frequency domain response spectrum composed of acoustic impedance, acoustic capacitance, and acoustic inertia.

[0010] A further technical solution is that the equivalent physical model of the battery's ultrasonic frequency domain response spectrum is expressed as follows:

[0011] in, For equivalent acoustic impedance, For equivalent acoustic resistance, The imaginary unit, The ultrasonic angular frequency, For equivalent acoustic inertia, For equivalent sound and appearance.

[0012] A further technical solution, based on the equivalent physical model of the battery's ultrasonic frequency domain response spectrum, establishes a quantitative relationship between propagation characteristic parameters and equivalent acoustic parameters, expressed as follows:

[0013]

[0014] in, For group delay, For equivalent acoustic inertia, For equivalent sound and appearance, For equivalent speed of sound, This is the material's equivalent density.

[0015] A further technical solution, wherein the identification of optimal equivalent acoustic parameters through parameter optimization specifically includes: Based on the equivalent physical model of the ultrasonic frequency domain response spectrum of the battery and the ultrasonic frequency domain response spectrum, a functional relationship between the model prediction response and the equivalent acoustic parameters is established. Construct an equivalent physical model of the battery's ultrasonic frequency domain response spectrum to predict the error function between the predicted response and the measured ultrasonic frequency domain response spectrum; The optimal equivalent acoustic parameters are obtained by minimizing the error function using a parameter optimization algorithm.

[0016] A further technical solution employs a capacity prediction function to predict battery capacity based on optimal equivalent acoustic parameters; the capacity prediction function is determined using an adaptive two-layer fitting method that integrates local weighted regression and residual compensation mechanisms.

[0017] Secondly, the present invention provides a battery manufacturing quality inspection and cascade utilization device based on ultrasonic features, comprising: The measured ultrasonic response unit is used to acquire the ultrasonic response signal of the battery under test under ultrasonic excitation signal at different frequencies, and to perform frequency domain analysis on it to construct the ultrasonic frequency domain response spectrum of the battery under test. The feature dataset construction unit is connected to the measured ultrasonic response unit and is used to enhance the features of the ultrasonic frequency domain response spectrum and construct a multidimensional acoustic feature dataset by combining battery capacity data. The equivalent physical model construction unit is connected to the measured ultrasonic response unit and the feature dataset construction unit, respectively. It is used to decompose the amplitude and phase of the ultrasonic frequency domain response spectrum based on the multidimensional acoustic feature dataset, define the equivalent acoustic impedance of the battery by analogy with the complex impedance concept of electrochemistry, and construct the equivalent physical model of the ultrasonic frequency domain response spectrum of the battery. The model prediction response unit is connected to the equivalent physical model construction unit and the measured ultrasonic response unit, respectively, and is used to identify the optimal equivalent acoustic parameters based on the equivalent physical model and ultrasonic frequency domain response spectrum of the battery through parameter optimization. The quality inspection and grading unit, connected to the model prediction response unit, is used to determine battery defects based on the distribution characteristics of the optimal equivalent acoustic parameters, predict battery capacity based on the optimal equivalent acoustic parameters, and grade the battery according to the predicted battery capacity value.

[0018] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features as described in the first aspect.

[0019] Fourthly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the method for battery manufacturing quality inspection and secondary utilization based on ultrasonic features as described in the first aspect.

[0020] The above one or more technical solutions have the following beneficial effects: This invention applies broadband ultrasonic excitation to the battery under test and extracts multi-dimensional frequency domain features to construct an ultrasonic frequency domain response equivalent physical model (UIS) analogous to electrochemical impedance spectroscopy (EIS). By using parameter inversion to identify the internal physical structure characteristics of the battery, it can achieve accurate identification of battery manufacturing quality defects and rapid evaluation and tiered utilization sorting of the residual performance of retired batteries.

[0021] This invention is applicable to the quality inspection and grading of various power and energy storage batteries in fields such as quality control of battery production lines, recycling of retired power batteries, and large-scale energy storage systems.

[0022] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0023] The accompanying drawings, which form part of this invention, 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 improper limitation of the invention.

[0024] Figure 1 This is a flowchart of a battery manufacturing quality inspection and cascade utilization method based on ultrasonic features according to an embodiment of the present invention. Detailed Implementation

[0025] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0026] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0027] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0028] Example 1 like Figure 1 As shown, this embodiment discloses a method for battery manufacturing quality inspection and secondary utilization based on ultrasonic features. The method includes the following steps: S1: Obtain the ultrasonic response signal of the battery under test under ultrasonic excitation signal at different frequencies, and perform frequency domain analysis on it to construct the ultrasonic frequency domain response spectrum of the battery under test; In this embodiment, an excitation signal is applied to the battery under test using a broadband ultrasonic excitation device, and ultrasonic response signals of the battery at different frequencies are collected. Frequency domain analysis is performed on the response signals to extract the amplitude, phase and propagation characteristic parameters corresponding to each frequency point, and the ultrasonic frequency domain response spectrum (UFRS) of the battery is constructed.

[0029] Under normal or controlled operating conditions, a swept-frequency continuous wave or multi-frequency superimposed ultrasonic excitation signal is applied to the battery under test using a broadband ultrasonic excitation device. The excitation signal covers a preset frequency range. Within the frequency range, ultrasonic transducers deployed on the battery surface are used to collect the battery's response signals under different frequency excitations. The response signal includes transmitted waves, reflected waves, or a combination thereof.

[0030] The acquired time-domain signal and By performing frequency domain transformation, the spectra of the excitation signals are obtained respectively. With response signal spectrum And calculate the ultrasonic frequency domain response function of the battery:

[0031] in, For the actual response, The imaginary unit, The ultrasonic angular frequency, This is the imaginary part of the response.

[0032] Further extraction yields the ultrasonic frequency domain response function. Amplitude and phase information:

[0033]

[0034] in, Let be the amplitude of the ultrasonic frequency domain response function. Let be the phase of the ultrasonic frequency domain response function.

[0035] Further extraction of propagation characteristic parameters, expressed as:

[0036]

[0037] in, For group delay, For equivalent speed of sound, The length of the ultrasonic propagation path can be obtained by experimentally measuring the battery thickness.

[0038] Based on the amplitude, phase, and propagation characteristic parameters extracted at each frequency point, a multidimensional ultrasonic frequency domain response spectrum feature vector of the battery is constructed:

[0039] in, The total number of frequency points. For batteries The multidimensional ultrasonic frequency domain response spectrum eigenvector. For the first The amplitude of the response function at each frequency point For the first Phase of the response function at each frequency point For the first Group delay at each frequency point For the first The equivalent speed of sound at each frequency point. The total number of frequency points can be selected according to the experimental accuracy requirements.

[0040] S2: Enhance the features of the ultrasonic frequency response spectrum and construct a multidimensional acoustic feature dataset by combining it with battery capacity data; In this embodiment, the ultrasonic frequency domain response spectrum of the battery under test is enhanced and standardized, battery capacity data is collected to obtain battery capacity labels, and a multidimensional acoustic feature dataset is constructed.

[0041] Cross-band statistical feature extraction is performed on the multidimensional ultrasonic frequency domain response spectrum feature vector of the battery to enhance its ability to characterize changes in battery capacity, resulting in an enhanced feature vector:

[0042]

[0043] in, For batteries exist The mean of the amplitude of the response function at each frequency point For the first The amplitude of the response function at each frequency point For batteries exist The standard deviation of the response function amplitude at each frequency point For batteries The multidimensional ultrasonic frequency domain response spectrum enhancement feature vector.

[0044] The enhanced feature vectors are standardized to eliminate dimensional differences. The standardization formula is as follows:

[0045] in, The standardized feature vector, The feature mean vector of the training set. is the standard deviation vector of the features in the training set.

[0046] The standardized feature vectors are matched with battery capacity labels to construct a multidimensional acoustic feature dataset:

[0047]

[0048]

[0049] in, For multidimensional acoustic feature datasets, For training set features, For training set labels, For the first Multidimensional ultrasonic frequency domain response spectrum enhancement feature vector of each battery sample For the corresponding number Battery capacity data for individual battery samples. This is a transposition. Battery capacity data sources include experimental measurement data, battery management system estimation results, or standard operating condition calibration data.

[0050] S3: Based on the amplitude-phase decomposition of the ultrasonic frequency domain response spectrum using a multidimensional acoustic feature dataset, the equivalent acoustic impedance of the battery is defined by analogy with the complex impedance concept in electrochemistry, and an equivalent physical model of the ultrasonic frequency domain response spectrum of the battery is constructed. In this embodiment, by analogy with the electrochemical impedance spectroscopy (EIS) theory, an equivalent physical model (UIS) of the ultrasonic frequency domain response spectrum of the battery is constructed to establish the analytical relationship between acoustic propagation characteristics (propagation characteristic parameters) and physical structure parameters (equivalent acoustic parameters).

[0051] S301: Based on a multidimensional acoustic feature dataset For the ultrasonic frequency domain response function Perform amplitude-phase decomposition to obtain its polar coordinate expression:

[0052] S302: Analogous to the concept of complex impedance in electrochemistry, the equivalent acoustic impedance of a battery is defined as:

[0053] in, This is the equivalent acoustic impedance of the battery. Equivalent acoustic impedance is used to describe the dissipation characteristics of sound energy. The imaginary unit, This is the equivalent acoustic impedance, used to describe energy storage characteristics.

[0054] S303: Equivalent modeling of the battery's internal structural parameters. The expansion behavior of active materials, electrolyte wetting characteristics, and pore structure evolution are respectively equated as distributed parameter elements, constructing a model based on acoustic resistance. voice and appearance and acoustic inertia The equivalent physical model (UIS) of the battery's ultrasonic frequency domain response spectrum:

[0055] in, Equivalent acoustic impedance characterizes the energy dissipation caused by material viscoelasticity and interfacial friction; Equivalent acoustic inertia characterizes inertial effects (such as vibration of a solid skeleton). This is the equivalent acoustic-capacitance, characterizing compressibility and porosity energy storage effect. Within the frequency band, , , It is approximately a constant.

[0056] S304: Establish a quantitative relationship between propagation characteristic parameters and UIS model parameters (equivalent acoustic parameters):

[0057]

[0058] in, The equivalent density of the material is determined by the properties of the battery material.

[0059] S4: Based on the equivalent physical model of the ultrasonic frequency domain response spectrum of the battery and the ultrasonic frequency domain response spectrum, the optimal equivalent acoustic parameters are identified through parameter optimization. In this embodiment, based on the equivalent physical model of the battery's ultrasonic frequency domain response spectrum and the ultrasonic frequency domain response spectrum, the optimal equivalent acoustic parameters are identified through parameter optimization, so as to minimize the error between the model's predicted response and the actual response.

[0060] S401: Based on the equivalent physical model of the ultrasonic frequency domain response spectrum of a battery, utilizing the obtained measured ultrasonic frequency domain response function. Establish the functional relationship between the model's predicted response and the equivalent acoustic parameters:

[0061]

[0062] S402: Construct the error function between the model's predicted response and the measured response (the measured ultrasonic frequency domain response function), and define it as the objective function for parameter inversion:

[0063] in, Let be the objective function. The total number of frequency points. For the first The objective function is a formula for parameter optimization, aiming to minimize the error between the measured ultrasonic frequency domain response function and the model predicted response to obtain the optimal equivalent acoustic parameters.

[0064] S403: The objective function (error function) is minimized using a parameter optimization algorithm to obtain the optimal equivalent acoustic parameters.

[0065] in, This is the optimal equivalent acoustic impedance. For optimal equivalent acoustic inertia, This is the optimal equivalent acoustic volume.

[0066] The parameter optimization algorithm adopts an improved sparrow search algorithm that integrates chaotic mapping and adaptive mutation. This method can effectively overcome the problem of search blind zone caused by strong randomness in the initial population distribution and severe sample clustering in the traditional sparrow search algorithm. At the same time, combined with the adaptive mutation mechanism of iterative decay, it solves the defects of existing algorithms that are prone to getting trapped in local optima and have insufficient convergence accuracy in the later stages.

[0067] The improved sparrow search algorithm includes the following steps: S4031: Population initialization. Set the population size to... The maximum number of iterations is The position of each individual sparrow is represented as:

[0068] in, For the first The position of each individual sparrow.

[0069] The initial sequence is generated using the Logistic chaotic mapping:

[0070] in, For the first A sequence of values, For the first A sequence of values.

[0071] This is then mapped to the parameter search range to improve the initial population diversity and global traversal capability.

[0072] in, Let the lower bound vector be the search vector for the parameters to be optimized. For the first The initial position vector of each sparrow after mapping Let be the upper bound vector for the search of the parameters to be optimized.

[0073] S4032: Position Update and Adaptive Mutation.

[0074] Calculate the fitness value for each individual:

[0075]

[0076] in, For the first The fitness value of an individual sparrow. This is the fitness function, used to evaluate the quality of a sparrow's location.

[0077] The positions of the discoverer and followers are updated according to the sparrow search mechanism to obtain the new individual. Then, an adaptive mutation perturbation is applied to the current best individual:

[0078]

[0079] in, For the first The mutation perturbation of the next iteration. This represents the current iteration number. For the first The optimal individual in the next iteration. For the first The next best individual. The vector is random. In the early stage, the perturbation range is increased to escape local optima, and in the later stage, the perturbation amplitude is decreased to improve convergence accuracy.

[0080] S4033: Iterative optimization and result output.

[0081] Repeat step S4032, and apply boundary constraints to the update results for each round:

[0082] The iteration terminates when the following condition is met:

[0083] in, For the first The fitness function for the next iteration. For the first The fitness function of the order of iterations. A predefined iteration stopping threshold.

[0084] Output the global optimal solution:

[0085] S404: Based on the optimal equivalent acoustic parameters, and combining the quantitative relationship between propagation characteristic parameters and UIS model parameters, calculate the relevant propagation characteristic parameters:

[0086]

[0087] in, For optimal group delay, This is the optimal equivalent speed of sound.

[0088] S5: Determine battery defects based on the distribution characteristics of the optimal equivalent acoustic parameters, predict battery capacity based on the optimal equivalent acoustic parameters, and classify the battery according to the predicted battery capacity value.

[0089] In this embodiment, by combining the identified optimal equivalent acoustic parameters with the predicted battery capacity, manufacturing quality inspection and defect identification are achieved through parameter consistency analysis, while the predicted battery capacity is used for tiered and graded utilization.

[0090] S501: Multidimensional acoustic feature dataset For each battery sample of the battery under test, the parameter identification process in step S4 is performed to obtain the corresponding optimal equivalent acoustic parameters. and the corresponding capacity data label Perform a match.

[0091] S502: Based on optimal equivalent acoustic parameters With capacity data tags Establish the actual battery capacity Functional relationship with equivalent acoustic parameters:

[0092] in, This is the capacity prediction function.

[0093] The capacity prediction function is determined using an adaptive two-level fitting method that integrates locally weighted regression and residual compensation mechanisms. This method uses a main regression model to characterize the overall trend of change between capacity and equivalent acoustic parameters, and utilizes a locally weighted residual compensation mechanism to correct nonlinear errors and individual sample differences that are difficult to represent by traditional single fitting models. This solves the problems of insufficient generalization ability and large local prediction bias in existing capacity prediction methods. Specifically, it includes: Using equivalent acoustic parameters as input features, a main regression model is established:

[0094] in, For the first The initial prediction capacity of the group of samples, , , , The main model parameters are obtained by solving using the least squares method.

[0095] Calculate the master model residuals:

[0096] in, For the first The residuals of the group samples For the first The size data labels of the group samples.

[0097] The weighting coefficients are constructed based on the distance between the sample to be predicted and historical samples in the parameter space:

[0098]

[0099] in, For the first The weighting coefficients of the group samples, For the sample to be predicted and the first The Euclidean distance between a group of historical samples in the parameter space. This is the distance attenuation coefficient, used to control the rate at which the weighting coefficient decreases as the distance increases; The equivalent acoustic impedance for historical samples. Let be the equivalent acoustic impedance of the sample to be predicted. The equivalent acoustic inertia of historical samples, The equivalent acoustic inertia of the sample to be predicted, The equivalent acoustic volume for historical samples. The equivalent acoustic volume of the sample to be predicted.

[0100] The local residual compensation amount is established using a weighted average method:

[0101] in, This is the amount of local residual compensation.

[0102] The output of the master model is superimposed with the residual compensation result to obtain the final capacity prediction function:

[0103] Right now:

[0104] S503: During battery manufacturing inspection or secondary utilization sorting, the ultrasonic response signal is acquired in real time and step S4 is executed to obtain the current optimal equivalent acoustic parameters. .

[0105] S504: Based on the distribution characteristics of the current optimal equivalent acoustic parameters, quality inspection is performed on the battery manufacturing process, specifically including: Optimal equivalent acoustic parameters of a batch of battery samples for the test battery Perform statistical analysis and calculate the mean of each parameter. with standard deviation :

[0106]

[0107] in, Total number of battery samples For the first For each battery sample, corresponding parameter values ​​are used to establish a baseline for the distribution of normal product parameters.

[0108] Construct parameter offset vector:

[0109] in, For the first The parameter offset vector of each battery sample. This is the equivalent acoustic impedance offset. For the first The optimal equivalent acoustic impedance for each battery sample. This is the equivalent acoustic impedance mean. The standard deviation of equivalent acoustic impedance. For equivalent acoustic inertial offset, For the first The optimal equivalent acoustic inertia of a single battery sample. The equivalent mean acoustic inertia, For the equivalent acoustic inertia standard deviation, This is the equivalent acoustic-capacitance shift. For the first The optimal equivalent acoustic capacitance of a single battery sample. The equivalent acoustic volume mean, The standard deviation is the equivalent acoustic-volume standard deviation.

[0110] Based on the joint change direction, amplitude ratio, and coupling relationship of the three parameters, a manufacturing defect mapping rule base is established: Different parameter combinations correspond to different defect types, including: when , , At that time, it was determined that the electrode sheet was not compacted sufficiently; when , , At that time, it was determined to be electrode misalignment or interlayer wrinkles; when , , If the electrolyte is excessive or the wetting is abnormal, it is determined that there is an excess of electrolyte. when , , If the weld is faulty, it is determined to be a loose weld or an abnormal contact of the current collector. when , , At that time, it was determined to be an internal foreign object or local structural damage.

[0111] S505: Employs a capacity prediction function to predict battery capacity based on the current optimal equivalent acoustic parameters, and classifies batteries according to the predicted battery capacity values.

[0112] When sorting batteries for reuse, in addition to quality inspection to remove defective samples, they should also be graded according to their predicted capacity, specifically including: Calculate capacity retention rate :

[0113] in, This is a predicted value for battery capacity. This is the initial rated capacity.

[0114] Based on capacity retention rate Battery grading: When When it is determined to be a first-level secondary battery, it is used for energy storage or high-performance applications; when When it is determined to be a secondary battery, it is used for low-power backup scenarios; when When a battery is deemed obsolete, it enters the dismantling and recycling process.

[0115] In summary, this invention constructs an equivalent physical model (UIS) of the ultrasonic frequency domain response of a battery, analogous to electrochemical impedance spectroscopy, and combines it with multidimensional acoustic feature enhancement and parameter inversion identification methods to achieve accurate identification of physical structural defects in the battery manufacturing process, as well as quantitative evaluation and tiered utilization sorting of the residual performance of retired batteries.

[0116] The beneficial effects of this invention are: High physical interpretability: This invention constructs an equivalent physical model (UIS) of the ultrasonic frequency domain response of a battery by analogy with electrochemical impedance spectroscopy (EIS) theory, establishing an analytical mapping relationship between acoustic characteristics (such as acoustic impedance and attenuation coefficient) and internal physical structural parameters of the battery (such as effective elastic modulus, porosity, and wetting state). This provides a scientific basis for identifying structural defects in the manufacturing process and the internal degradation mechanism of retired batteries, overcoming the shortcomings of traditional machine learning "black box models" that lack mechanistic support in quality judgment and sorting logic.

[0117] High accuracy in multidimensional feature perception: This invention constructs an ultrasonic frequency domain response spectrum (UFRS) through broadband ultrasonic excitation, enabling the simultaneous extraction of multidimensional frequency domain parameters such as amplitude, phase, and propagation characteristics. Compared to traditional time-domain analysis methods that rely solely on time-of-flight (ToF) or amplitude changes, this invention can more sensitively capture subtle differences in the internal microstructure of batteries, significantly improving the resolution of manufacturing quality inspection and the feature dimensions for tiered utilization sorting.

[0118] In-depth parameter identification: This invention achieves quantitative identification of the equivalent acoustic parameters inside the battery through parameter inversion methods. This "from the surface to the core" deduction capability enables the system not only to monitor macroscopic electrical performance but also to quantify changes in the mechanical properties of electrode materials and the degree of interface contact degradation. This deep technical insight provides core technical support for accurately distinguishing the health level of retired batteries and screening high-consistency manufacturing batches.

[0119] High efficiency and practicality of detection and sorting: The acoustic detection method used in this invention is non-invasive and fast, and can be highly integrated with battery production lines and automated sorting equipment. Through the established evaluation and sorting model, online full inspection of battery manufacturing quality and efficient grading and sorting of retired batteries are realized, effectively solving the problems of low efficiency, high cost, and inability to cover the full sample volume of traditional detection methods (such as EIS or disassembly sampling inspection).

[0120] Example 2 This embodiment discloses a battery manufacturing quality inspection and secondary utilization device based on ultrasonic features, including: The measured ultrasonic response unit is used to acquire the ultrasonic response signal of the battery under test under ultrasonic excitation signal at different frequencies, and to perform frequency domain analysis on it to construct the ultrasonic frequency domain response spectrum of the battery under test. The feature dataset construction unit is connected to the measured ultrasonic response unit and is used to enhance the features of the ultrasonic frequency domain response spectrum and construct a multidimensional acoustic feature dataset by combining battery capacity data. The equivalent physical model construction unit is connected to the measured ultrasonic response unit and the feature dataset construction unit, respectively. It is used to decompose the amplitude and phase of the ultrasonic frequency domain response spectrum based on the multidimensional acoustic feature dataset, define the equivalent acoustic impedance of the battery by analogy with the complex impedance concept of electrochemistry, and construct the equivalent physical model of the ultrasonic frequency domain response spectrum of the battery. The model prediction response unit is connected to the equivalent physical model construction unit and the measured ultrasonic response unit, respectively, and is used to identify the optimal equivalent acoustic parameters based on the equivalent physical model and ultrasonic frequency domain response spectrum of the battery through parameter optimization. The quality inspection and grading unit, connected to the model prediction response unit, is used to determine battery defects based on the distribution characteristics of the optimal equivalent acoustic parameters, predict battery capacity based on the optimal equivalent acoustic parameters, and grade the battery according to the predicted battery capacity value.

[0121] Example 3 The purpose of this embodiment is to provide a computing device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method of Embodiment 1.

[0122] Example 4 The purpose of this embodiment is to provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps of the method of Embodiment 1.

[0123] The steps and methods involved in the apparatuses of Embodiments 3 and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0124] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0125] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0126] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features, characterized in that, include: The ultrasonic response signals of the battery under test under ultrasonic excitation signals of different frequencies were obtained and frequency domain analysis was performed to construct the ultrasonic frequency domain response spectrum of the battery under test. The ultrasonic frequency response spectrum is enhanced with features, and combined with battery capacity data, a multidimensional acoustic feature dataset is constructed. Based on the amplitude-phase decomposition of the ultrasonic frequency domain response spectrum using the multidimensional acoustic feature dataset, the equivalent acoustic impedance of the battery is defined by analogy with the complex impedance concept in electrochemistry, and an equivalent physical model of the ultrasonic frequency domain response spectrum of the battery is constructed. Based on the equivalent physical model and ultrasonic frequency domain response spectrum of the battery, the optimal equivalent acoustic parameters are identified through parameter optimization. Battery defects are determined based on the distribution characteristics of the optimal equivalent acoustic parameters, and battery capacity is predicted based on the optimal equivalent acoustic parameters. The batteries are then classified according to the predicted battery capacity values.

2. The method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features as described in claim 1, characterized in that, The ultrasonic frequency domain response spectrum of the battery under test specifically includes: The ultrasonic excitation signal and ultrasonic response signal are transformed in the frequency domain to obtain the spectrum of the excitation signal and the spectrum of the response signal, respectively. The ultrasonic frequency domain response function of the battery is calculated based on the spectrum of the excitation signal and the spectrum of the response signal. The amplitude and phase of the ultrasonic frequency domain response function are extracted, and then the propagation characteristic parameters are extracted. Based on the amplitude, phase, and propagation characteristic parameters extracted at each frequency point, a multidimensional ultrasonic frequency domain response spectrum feature vector of the battery under test is constructed.

3. The method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features as described in claim 1, characterized in that, Based on the equivalent acoustic impedance, the internal structural parameters of the battery are modeled equivalently to construct an equivalent physical model of the battery's ultrasonic frequency domain response spectrum, which consists of acoustic impedance, acoustic capacitance, and acoustic inertia.

4. The method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features as described in claim 1, characterized in that, The equivalent physical model of the battery's ultrasonic frequency domain response spectrum is expressed as follows: in, For equivalent acoustic impedance, For equivalent acoustic resistance, The imaginary unit, The ultrasonic angular frequency, For equivalent acoustic inertia, For equivalent sound and appearance.

5. The method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features as described in claim 1, characterized in that, Based on the equivalent physical model of the battery's ultrasonic frequency domain response spectrum, a quantitative relationship between the propagation characteristic parameters and the equivalent acoustic parameters is established, expressed as follows: in, For group delay, For equivalent acoustic inertia, For equivalent sound and appearance, For equivalent speed of sound, This is the material's equivalent density.

6. The method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features as described in claim 1, characterized in that, The process of identifying the optimal equivalent acoustic parameters through parameter optimization specifically includes: Based on the equivalent physical model of the ultrasonic frequency domain response spectrum of the battery and the ultrasonic frequency domain response spectrum, a functional relationship between the model prediction response and the equivalent acoustic parameters is established. Construct an equivalent physical model of the battery's ultrasonic frequency domain response spectrum to predict the error function between the predicted response and the measured ultrasonic frequency domain response spectrum; The optimal equivalent acoustic parameters are obtained by minimizing the error function using a parameter optimization algorithm.

7. The method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features as described in claim 1, characterized in that, A capacity prediction function is used to predict the battery capacity based on the optimal equivalent acoustic parameters; the capacity prediction function is determined by an adaptive two-layer fitting method that integrates local weighted regression and residual compensation mechanism.

8. A battery manufacturing quality inspection and cascade utilization device based on ultrasonic characteristics, characterized in that, include: The measured ultrasonic response unit is used to acquire the ultrasonic response signal of the battery under test under ultrasonic excitation signal at different frequencies, and to perform frequency domain analysis on it to construct the ultrasonic frequency domain response spectrum of the battery under test. The feature dataset construction unit is connected to the measured ultrasonic response unit and is used to enhance the features of the ultrasonic frequency domain response spectrum and construct a multidimensional acoustic feature dataset by combining battery capacity data. The equivalent physical model construction unit is connected to the measured ultrasonic response unit and the feature dataset construction unit, respectively. It is used to decompose the amplitude and phase of the ultrasonic frequency domain response spectrum based on the multidimensional acoustic feature dataset, define the equivalent acoustic impedance of the battery by analogy with the complex impedance concept of electrochemistry, and construct the equivalent physical model of the ultrasonic frequency domain response spectrum of the battery. The model prediction response unit is connected to the equivalent physical model construction unit and the measured ultrasonic response unit, respectively, and is used to identify the optimal equivalent acoustic parameters based on the equivalent physical model and ultrasonic frequency domain response spectrum of the battery through parameter optimization. The quality inspection and grading unit, connected to the model prediction response unit, is used to determine battery defects based on the distribution characteristics of the optimal equivalent acoustic parameters, predict battery capacity based on the optimal equivalent acoustic parameters, and grade the battery according to the predicted battery capacity value.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the method for battery manufacturing quality inspection and secondary utilization based on ultrasonic features as described in any one of claims 1-7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the method for battery manufacturing quality inspection and cascade utilization based on ultrasonic features as described in any one of claims 1-7.

Images