Lithium battery test analysis method and system based on deep learning and feature analysis

Abstract

The invention relates to the technical field of battery performance test and analysis, and discloses a lithium battery test analysis method based on deep learning and feature analysis, which comprises the following steps: S1, establishing a dynamic environment model, and obtaining dynamic distribution of temperature and humidity in a battery use environment, the dynamic distribution of the temperature and the humidity is based on the combined influence of diffusion, convection and external disturbance; s2, dynamic environment features are extracted, and main amplitudes and frequencies of temperature and humidity serve as environment feature vectors; and S3, based on the lithium ion diffusion model, a coupling relationship between lithium ion concentration distribution and environmental characteristics is established, and the lithium ion concentration distribution is influenced by temperature and humidity changes in a dynamic environment. By adopting the technical scheme based on dynamic environment modeling and feature extraction, dynamic distribution of temperature and humidity is described through a partial differential equation, and environment feature vectors are extracted, so that the technical effect of accurately simulating and quantifying the influence of complex environment change on battery performance is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of battery performance testing and analysis, and specifically to a lithium battery testing and analysis method and system based on deep learning and feature analysis. Background Art

[0002] As a high energy density and high safety energy storage device, polymer lithium-ion batteries have been widely used in electric vehicles, portable devices, energy storage systems and other fields. However, battery performance is significantly affected by environmental factors, especially in actual use. Dynamic changes in temperature and humidity may lead to significant fluctuations in the internal reaction rate of the battery, the lithium ion diffusion process, and the overall discharge performance. Therefore, how to accurately evaluate battery performance under dynamic environment has become an important issue in current research.

[0003] In general, the existing technology usually tests the performance of lithium-ion batteries under constant temperature and humidity conditions. This technical solution can indeed obtain stable test results in a laboratory environment, but it ignores the complexity of environmental conditions under real working conditions. For example, during the actual driving of electric vehicles, the ambient temperature may fluctuate significantly with seasons and regions, and the humidity may also be affected by a variety of external conditions. The existing methods lack in-depth research on dynamic environment modeling and simulation, resulting in poor applicability of test results in actual scenarios.

[0004] In the prior art, some test methods attempt to introduce changes in environmental factors into the evaluation of battery performance. However, these methods often rely on simple point measurements and usually only test a few fixed discrete values ​​of temperature and humidity. Although this method can preliminarily reflect the impact of environmental changes on battery performance, it fails to simulate continuous and dynamic environmental changes, so the test results are difficult to accurately depict the performance change process under real working conditions. This defect greatly reduces the representativeness of the test results and the refinement of the analysis.

[0005] In terms of physical modeling of lithium-ion batteries, most existing research focuses on static diffusion and reaction processes, and the diffusion coefficient and reaction rate are usually assumed to be fixed values. Although this assumption simplifies the modeling process, in a dynamic environment, this method cannot reflect the regulatory effects of temperature and humidity on the internal diffusion process and reaction kinetics of the battery. For example, when the ambient temperature rises, the diffusion rate of lithium ions will change significantly, and the static model cannot capture this change process. The lack of a coupling mechanism between the dynamic environment and the internal reaction process makes the existing model less adaptable to complex working conditions. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention provides a lithium battery testing and analysis method and system based on deep learning and feature analysis, which solves the technical problem in the prior art that the impact of temperature and humidity changes in a dynamic environment on the discharge performance of polymer lithium-ion batteries cannot be accurately evaluated.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A lithium battery testing and analysis method based on deep learning and feature analysis, comprising: S1. Establish a dynamic environment model to obtain the dynamic distribution of temperature and humidity in the battery use environment, where the dynamic distribution of temperature and humidity is based on the combined influence of diffusion, convection and external disturbance; S2, extract dynamic environmental features, and use the main amplitude and frequency of temperature and humidity as environmental feature vectors; S3. Based on the lithium ion diffusion model, a coupling relationship between lithium ion concentration distribution and environmental characteristics is established, wherein the lithium ion concentration distribution is affected by changes in temperature and humidity in a dynamic environment; S4, collecting the discharge signal of the polymer lithium-ion battery, and extracting the voltage, current and power signals as the discharge feature vector; S5, interactively analyzing the environment feature vector and the discharge feature vector to generate an interactive feature vector; S6. Based on the interactive feature vector, determine whether the discharge performance of the polymer lithium-ion battery is normal.

[0008] Preferably, the dynamic environment model in step S1 describes the dynamic changes of temperature and humidity through the following partial differential equation: The change of temperature is determined by diffusion, convection and external disturbance terms; The change of humidity is determined by diffusion, convection and external disturbance.

[0009] Preferably, extracting the dynamic environment features in step S2 includes: performing time-frequency domain analysis on the dynamic distribution of temperature and humidity to obtain their main amplitudes and frequencies to form a dynamic environment feature vector.

[0010] Preferably, the lithium ion diffusion model in step S3 describes the influence of temperature and humidity on the diffusion coefficient of lithium ions, and adjusts the diffusion rate and reaction rate through a dynamic environment characteristic vector.

[0011] Preferably, the discharge signal collected in step S4 includes voltage, discharge current and power value in a time series, and the discharge signal is preprocessed to obtain a discharge feature vector.

[0012] Preferably, in step S5, a feature interaction analysis method is adopted to fuse the environment feature vector and the discharge feature vector through an attention mechanism to generate an interaction feature vector.

[0013] Preferably, the attention mechanism is implemented by the following steps: Mapping the environment feature vector and the discharge feature vector into a query vector, a key vector and a value vector respectively; Calculate the similarity score between the query vector and the key vector; The value vectors are weighted averaged using the similarity scores to generate an interaction feature vector.

[0014] Preferably, in step S6, whether the discharge performance of the polymer lithium-ion battery is normal is judged by a classification model based on an interactive feature vector, and the classification model is a deep neural network classifier.

[0015] Preferably, the classification model determines the battery performance through the probability distribution of the interactive feature vector, and outputs the classification result of the normal state or the abnormal state.

[0016] Lithium battery test and analysis system based on deep learning and feature analysis, including: Signal acquisition module, used to collect discharge signals of polymer lithium-ion batteries and temperature and humidity data of the battery working environment; Dynamic environment modeling module, used to establish dynamic distribution models of temperature and humidity; A feature extraction module, used to extract the dynamic features of temperature and humidity and the features of the discharge signal; A feature interaction analysis module is used to interactively analyze dynamic environment features and discharge features to generate interactive feature vectors; The performance evaluation module is used to determine whether the discharge performance of the polymer lithium-ion battery is normal according to the interactive feature vector and output the performance evaluation result.

[0017] The present invention provides a lithium battery testing and analysis method and system based on deep learning and feature analysis. It has the following beneficial effects: 1. The present invention adopts a technical solution based on dynamic environment modeling and feature extraction, describes the dynamic distribution of temperature and humidity through partial differential equations, and extracts environmental feature vectors, achieving the technical effect of accurately simulating and quantifying the impact of complex environmental changes on battery performance. Compared with the technical solution in the prior art that simply relies on constant temperature and humidity testing, the present invention solves the problem that traditional methods cannot adapt to actual dynamic environments, and significantly improves the authenticity and applicability of test results.

[0018] 2. The present invention achieves an accurate description of the physical process inside the battery under the influence of environmental variables by coupling the lithium ion diffusion model with dynamic environmental characteristics and using environmentally sensitive diffusion coefficient and reaction rate expressions. This technical solution can dynamically reflect the regulatory effect of temperature and humidity changes on the lithium ion concentration distribution. Compared with the static diffusion model in the prior art that ignores environmental factors, it effectively solves the problem of insufficient accuracy in analyzing the physical process inside the battery under complex working conditions.

[0019] 3. The present invention adopts a feature interaction analysis method based on the attention mechanism to combine dynamic environmental features with battery discharge features to generate interactive feature vectors, thereby fully exploring the nonlinear relationship between environmental factors and battery performance. Compared with the prior art solution that relies on simple linear analysis, the present invention overcomes the defect that the complex coupling between environmental features and battery performance is difficult to characterize, and provides a more reliable and comprehensive data basis for performance classification.

[0020] 4. The present invention constructs a deep learning classification model, analyzes and judges the interactive feature vectors, and outputs the classification results of the battery discharge performance, thereby achieving a rapid and accurate evaluation of the battery performance status under the influence of a dynamic environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Example: Please see attached Figure 1 , an embodiment of the present invention provides a lithium battery testing and analysis method based on deep learning and feature analysis, including: S1. Establish a dynamic environment model to obtain the dynamic distribution of temperature and humidity in the battery use environment, where the dynamic distribution of temperature and humidity is based on the combined influence of diffusion, convection and external disturbance; During the actual use of polymer lithium-ion batteries, their working environment is usually affected by the combined influence of dynamic temperature and humidity. This dynamic environmental change may have a significant impact on the performance and life of the battery. Therefore, in order to more comprehensively evaluate the battery performance, it is necessary to first establish a dynamic environmental model to simulate the temporal and spatial distribution of temperature and humidity. As mentioned in the above technical content, the dynamic distribution of temperature and humidity is mainly affected by the combined influence of diffusion, convection and external disturbances. This step describes the changing characteristics of temperature and humidity in detail by constructing a mathematical model based on partial differential equations, providing the necessary basic data and analysis support for subsequent steps.

[0024] In this embodiment, the dynamic environment model is established based on a two-dimensional partial differential equation to describe the distribution and variation of temperature and humidity in time and space.

[0025] Specifically, the dynamic change of temperature can be expressed as the following partial differential equation: ; The dynamic change of humidity can be expressed as: ; in: Indicates a moment Next, coordinates Temperature at K ).

[0026] Indicates a moment Next, coordinates Humidity at % ).

[0027] is the temperature diffusion coefficient (unit: / s), which represents the diffusion rate of temperature in space.

[0028] is the humidity diffusion coefficient (unit: / s), which represents the diffusion rate of humidity in space.

[0029] is the temperature convection coefficient (unit: m / s), which indicates the influence of convection on temperature.

[0030] is the humidity convection coefficient (unit: m / s), which indicates the influence of convection on humidity.

[0031] is the rate of temperature change caused by external heat sources (unit: 𝐾 / 𝑠, mainly caused by external disturbances, such as rapid fluctuations in ambient temperature.

[0032] The rate of change of humidity (in % / s) caused by external humidity sources, such as periodic fluctuations in air humidity.

[0033] In the above partial differential equation, the diffusion term and Describes the diffusion of temperature and humidity in space, and is usually used to characterize the diffusion of heat and humidity in a uniform environment. and It is used to describe the process of temperature and humidity transfer along a specific direction in the presence of wind speed, convection, etc. and It can be set according to specific environmental conditions, such as the impact of rapid fluctuations in outdoor temperature and humidity on the local environment of the battery.

[0034] As an option, the dynamic environment model in this embodiment can be implemented by a numerical solution technique, such as a finite difference method or a finite element method, to obtain the distribution of temperature and humidity in space and time.

[0035] Specifically, in some embodiments, the two-dimensional environment field can be divided into a number of finite grid nodes. By discretizing the partial differential equation, the following numerical update formula is obtained: ; ; in: and Respectively represent Time step time node ( ) temperature and humidity.

[0036] is the time step, which indicates the time interval between each calculation (unit: s) and are the second-order spatial partial derivatives of temperature and humidity, respectively, calculated using a discrete format.

[0037] In some specific implementations, the accuracy of the environmental simulation can also be controlled by setting appropriate initial conditions and boundary conditions. For example, the initial conditions can be set to the ambient temperature and humidity distribution when the battery starts running: ; The boundary condition can be set to be an adiabatic boundary, which means that there is no outflow of heat or moisture across the boundary: ; in Represents the boundary normal vector.

[0038] Specifically, the external disturbance term and It can be fitted by actual measured data or set according to specific application scenarios, such as: During the operation of electric vehicles, external disturbances may be caused by engine heat or temperature differences in the external environment; In energy storage systems, humidity disturbances can be caused by seasonal climate changes.

[0039] In some embodiments, such disturbances can be simulated by introducing a periodic function, for example: ; in , are the amplitude of the external disturbance, , is the frequency of the disturbance.

[0040] This embodiment constructs the above dynamic environment model to obtain the accurate distribution of temperature and humidity around the battery over time and space, providing a reliable basis for the subsequent extraction of dynamic environment features. This model can be flexibly adapted to different usage scenarios and has high applicability and versatility.

[0041] S2, extract dynamic environmental features, and use the main amplitude and frequency of temperature and humidity as environmental feature vectors; After completing the establishment of the dynamic environment model in step S1 and obtaining the dynamic distribution of temperature and humidity, in order to further quantify the impact of the dynamic environment on the discharge performance of the polymer lithium-ion battery, it is necessary to extract the dynamic environment features. The extraction of dynamic environment features is based on the main change characteristics of temperature and humidity. Through the time-frequency domain analysis method, the main amplitude and frequency of temperature and humidity changing over time are obtained. The extracted feature vector will serve as important input data for feature interaction analysis in subsequent steps.

[0042] Generally speaking, temperature and humidity signals in a dynamic environment contain both stable components (such as long-term change trends) and volatile components (such as periodic changes). In order to more effectively characterize the characteristics of these signals, this step uses Fourier transform to transform the time series signal into the frequency domain and extract the main frequency and amplitude information of the signal. In this way, not only can the interference of noise be eliminated, but also the amount of data can be greatly compressed, improving the efficiency of subsequent analysis.

[0043] In this embodiment, the dynamic environment feature vector is extracted by analyzing the time-frequency domain of the dynamic distribution of temperature and humidity. The specific implementation method is as follows: Specifically, the temperature signal and humidity signal The dynamic changes of are first transformed from the time domain to the frequency domain through Fourier transform. The mathematical expression of Fourier transform is as follows: ; in: Indicates signal Amplitude in the frequency domain; is the frequency (unit: Hz); is the time (unit: s).

[0044] In the Fourier transform process of temperature and humidity signals, the temperature and humidity time series are discretized into sample data of finite length in actual operation, and the fast Fourier transform (FFT) algorithm is used to improve the calculation efficiency.

[0045] In some embodiments, the Fourier transform of the temperature and humidity signal can further decompose its main frequency , and the main amplitude , .in: , Represent the frequencies of the main periodic components of the temperature signal and the humidity signal respectively; and They represent the amplitudes of the corresponding main frequency components, reflecting the intensity of changes in temperature and humidity.

[0046] For example, in one possible implementation, the main frequency of the temperature signal in the low frequency band is It can be used to describe the periodic impact of the temperature difference between day and night on the battery environment and the high-frequency amplitude of the humidity signal. This may reflect the potential interference of rapid fluctuations in local air humidity on battery performance.

[0047] Generally, there are multiple peak points in the spectrum of the signal amplitude after Fourier transformation, and each peak point corresponds to a main frequency component of the temperature and humidity signal. In order to further simplify the feature extraction process, only the main frequency with the largest amplitude and its corresponding amplitude value can be retained, and other minor components can be ignored.

[0048] In one possible implementation, the main frequency and amplitude values ​​can be extracted by the following formula: ; ; in: and are the absolute values ​​of the amplitudes of the temperature signal and the humidity signal in the frequency domain respectively; Indicates the frequency corresponding to the maximum amplitude point.

[0049] Through the above extraction process, the dynamic environment feature vector can be obtained: ; The characteristic vector quantifies the main amplitude and frequency information of temperature and humidity in a dynamic environment, which facilitates interactive analysis with the battery discharge characteristics in subsequent steps.

[0050] As an option, the Fourier transform process can also be combined with a filtering algorithm to further improve the accuracy of feature extraction. For example, in some embodiments, a bandpass filter is used to limit the main frequency range of the temperature and humidity signal to the actual interval of interest, thereby removing high-frequency noise or low-frequency interference. The transfer function of the bandpass filter can be expressed as: ; in and are the lower and upper frequency limits of the filter respectively. By performing Fourier transform on the filtered signal, interference can be significantly reduced and the extraction accuracy of the main frequency and amplitude can be improved.

[0051] In this embodiment, the feature vector may also be expanded according to the actual application scenario.

[0052] Specifically, in some embodiments, in addition to the main frequency and main amplitude, statistical features of the temperature and humidity signal can also be extracted, such as the mean, standard deviation, etc. For example: Mean temperature and humidity mean Respectively represent long-term trends; The standard deviation of temperature σT\sigma_TσT and the standard deviation of humidity σH\sigma_HσH represent the volatility of the signal.

[0053] The addition of the above extended features can make the dynamic environment feature vector from ; ; Through the dynamic environmental feature extraction in this step, complex temperature and humidity signals can be converted into quantitative feature vectors. These feature vectors can not only effectively reflect the intensity and periodicity of environmental changes, but also provide sufficient basic data support for the interactive analysis of the environment and battery discharge performance in subsequent steps.

[0054] S3. Based on the lithium ion diffusion model, a coupling relationship between lithium ion concentration distribution and environmental characteristics is established, wherein the lithium ion concentration distribution is affected by changes in temperature and humidity in a dynamic environment; After completing the extraction of dynamic environmental features in step S2, the main characteristic vectors of ambient temperature and humidity, such as amplitude and frequency, have been obtained. Changes in dynamic environmental features directly affect the diffusion rate of lithium ions in polymer lithium-ion batteries and the electrode reaction rate. Therefore, it is necessary to couple the dynamic environmental features with the lithium ion concentration distribution through a lithium ion diffusion model to form an accurate description of the physical process inside the battery. The goal of this step is to establish a coupling relationship between the dynamic environment and the electrochemical behavior inside the battery, and to provide theoretical support and data basis for the subsequent analysis of the interaction between the discharge signal and the dynamic environment.

[0055] In general, the lithium ion diffusion model is the main tool to describe the diffusion behavior of lithium ions in battery electrode materials. Its core is to use the diffusion coefficient and reaction rate The environmental dependence of the diffusion equation is characterized by the modulation effect of environmental factors on the lithium ion diffusion process. As an option, the diffusion model in the present invention is based on the Newman model, and the environmentally sensitive form of the diffusion equation is obtained by further coupling the dynamic changes of temperature and humidity by modifying the dynamic characteristic dependence of the diffusion coefficient and the reaction rate.

[0056] In this embodiment, the lithium ion diffusion model is described by the following partial differential equation: ; in: Represents the lithium ion concentration in units of 𝑚𝑜l ; is the time, the unit is ; is the spatial coordinate in the electrode thickness direction, in units of ; is the lithium ion diffusion coefficient, in s; is the current density, in units of ; is the Faraday constant, which is 96485 .

[0057] In the diffusion equation, the diffusion coefficient is a function that changes with the characteristics of the dynamic environment and is defined as: in: is the base diffusion coefficient of lithium ions, in units of s; is the diffusion activation energy, in units ofJ / ; is the gas constant, and its value is 8.314 J ; is the temperature in units of K ; is the humidity, unit is %; is the humidity sensitivity coefficient, which is used to quantify the effect of humidity on the diffusion coefficient and its unit is dimensionless.

[0058] Specifically, the diffusion coefficient The first item Describing the exponential effect of temperature on the diffusion rate, the second term Then it describes the linear correction of humidity to the diffusion coefficient. In one possible implementation, the degree of correction of humidity to the diffusion coefficient can be calibrated experimentally. This can be achieved by fitting the diffusion rate measurements in environments with humidity ranging from 20% to 80%.

[0059] As an option, this embodiment further considers the impact of the dynamic environment on the battery electrode reaction rate and defines the environment-sensitive reaction rate : in: is the base value of the reaction rate, in units of

[0060] The above reaction rate expression shows that the reaction rate is mainly affected by temperature and increases exponentially with increasing temperature. In some embodiments, the effect of humidity on the reaction interface can be extended by For example, high humidity may reduce the interfacial impedance and thus indirectly increase the reaction rate.

[0061] In this embodiment, the dynamic distribution of lithium ion concentration can be accurately described by coupling the dynamic environmental characteristics with the diffusion model.

[0062] Specifically, the dynamic environment feature vector The main frequency in , and amplitude , Will act on the diffusion coefficient and reaction rate In one possible implementation, the direct association between the dynamic features and the diffusion parameters can be achieved in the following way: Temperature amplitude By amendment Activation Energy , indirectly affecting the diffusion rate; Humidity frequency Characterizes the rapidity of humidity fluctuations, thus affecting the humidity correction factor Dynamic response.

[0063] For example, when the temperature changes the main frequency When it is higher, the diffusion process will show highly nonlinear dynamic characteristics. At this time, the diffusion model needs to be solved with a finer time step to capture the transient behavior of the lithium ion concentration change.

[0064] In a possible implementation, the diffusion model may also be optimized by a variational method to minimize the error of the diffusion concentration distribution.

[0065] The optimization objective of the variational method can be expressed as: in: represents the optimization objective function, which is the square integral of the concentration distribution error; is the integration area in the electrode thickness direction.

[0066] By minimizing , the optimized diffusion coefficient and reaction rate parameters can be obtained, thereby further improving the simulation accuracy of lithium ion concentration distribution.

[0067] By coupling the lithium ion diffusion model of this step with the dynamic environment characteristics, the dynamic environment has been successfully used to accurately describe the physical processes inside the battery. This model can fully capture the dynamic regulation of temperature and humidity on the diffusion rate and reaction rate, laying a solid theoretical foundation for the interactive analysis of discharge performance in subsequent steps.

[0068] S4, collecting the discharge signal of the polymer lithium-ion battery, and extracting the voltage, current and power signals as the discharge feature vector; After completing the coupling of the dynamic environment and the lithium ion diffusion model in step S3, in order to comprehensively evaluate the actual impact of the environment on the performance of the polymer lithium-ion battery, it is also necessary to collect the signal data generated by the battery during the discharge process and extract the characteristic vector representing the battery performance. The collection of discharge signals mainly includes time series data such as voltage, current and power, which can truly reflect the discharge behavior of the battery under different environmental conditions and its dynamic change characteristics.

[0069] In general, the volatility and characteristic changes of discharge signals are closely related to the electrochemical behavior inside the battery. By preprocessing and extracting features from these signals, the battery performance characteristics can be effectively quantified, providing reliable input data for subsequent interactive analysis and performance classification.

[0070] In this embodiment, the discharge signal of the battery, including voltage, current and power signals, is acquired in real time through a signal acquisition device. The specific implementation method is as follows: Specifically, the discharge voltage of the battery , discharge current and power The data are collected through high-precision sensors, and the sampling frequency and accuracy are set according to actual needs. The calculation formula is: in: For time The discharge power at the time is in W; For time The battery terminal voltage at ; For time The discharge current at this time is in A.

[0071] Generally, the signal acquisition device needs to meet the requirements of high resolution and low noise to ensure the accuracy and authenticity of the collected data. For example, in one possible implementation, the sampling frequency can be set to 10kHz to ensure the capture of the rapid fluctuation characteristics of voltage and current.

[0072] As an option, the collected discharge signals can be preprocessed to eliminate the influence of noise and abnormal data on feature extraction.

[0073] Specifically, in some embodiments, signal preprocessing includes the following steps: Filtering: A low-pass filter is used to remove high-frequency noise, and the filter cutoff frequency is set according to the frequency range of the battery discharge signal. For example, the main frequency of the discharge signal is usually below 1kHz, and the filter cutoff frequency can be set to 1.2kHz.

[0074] Outlier detection: Detect and remove outliers through statistical analysis. For example, the standard deviation of voltage and current can be calculated, and outliers that are outside the range of two times the standard deviation of the mean can be removed.

[0075] Normalization: Normalize the signal amplitude so that the characteristic data of different dimensions can be processed uniformly. The normalization formula is: in: is the normalized signal value; is the original signal value; , are the minimum and maximum values ​​of the signal, respectively.

[0076] Through the above preprocessing steps, clean and easy-to-analyze signal data can be obtained, providing a reliable foundation for subsequent feature extraction.

[0077] In this embodiment, the extraction of discharge signal features is based on a time series analysis method, which converts the time domain and frequency domain features of the signal into feature vectors.

[0078] Specifically, the voltage , Current and power The time series characteristics include: Mean: The average value of the signal is used to characterize the overall level of the signal: in is the total time length of the signal, Indicates the signal value.

[0079] Standard Deviation: The volatility of the signal is used to quantify the stability of the signal: Frequency domain main frequency and the main amplitude : Extract the main frequencies and amplitudes of the signal through Fourier transform: in is the amplitude of the signal in the frequency domain.

[0080] In a possible implementation, the extracted discharge features can be further composed into feature vectors as input for subsequent analysis.

[0081] The discharge eigenvector is defined as: in: Respectively represent the mean and standard deviation of the voltage signal; Respectively represent the main frequency and main amplitude of the voltage signal; Respectively represent the mean and standard deviation of the current signal; Respectively represent the main frequency and main amplitude of the current signal; Represent the mean and standard deviation of the power signal respectively.

[0082] In some embodiments, additional signal features, such as instantaneous maximum value, minimum value, etc., may be added by expanding the dimension of the feature vector.

[0083] Through the discharge signal collection and feature extraction in this step, complex time series signals can be converted into feature vectors that are easy to analyze, laying a data foundation for the subsequent interactive analysis of dynamic environmental characteristics and discharge characteristics. This feature extraction method fully considers the time domain and frequency domain characteristics of the signal, improving the accuracy of analysis and the efficiency of data processing.

[0084] S5, interactively analyzing the environment feature vector and the discharge feature vector to generate an interactive feature vector; After completing the collection and feature extraction of the discharge signal in step S4, the dynamic environment feature vector and the discharge feature vector have been obtained. These two sets of features describe the changing laws of temperature and humidity in the dynamic environment and the battery discharge behavior. Since the dynamic environment directly affects the electrochemical reaction and diffusion process of the lithium-ion battery, it is necessary to interactively analyze the environmental features and the discharge features to form a unified interactive feature vector to provide complete data input for subsequent performance judgment.

[0085] In general, the relationship between dynamic environmental features and discharge features is nonlinear and time-series complex, and is difficult to be represented by a simple linear model. As an option, the present invention adopts a feature interaction analysis method to fuse environmental features with discharge features in the feature space, combines the attention mechanism to achieve weight allocation of key features, and generates an interactive feature vector.

[0086] In this embodiment, the dynamic environment feature is associated with the discharge feature through the feature interaction analysis method, and the specific implementation method is as follows: the dynamic environment feature vector and the discharge characteristic vector Respectively expressed as: That middle: are the amplitude and frequency of the temperature, respectively ； are the amplitude and frequency of humidity, respectively; is the mean and standard deviation of the voltage signal; is the mean and standard deviation of the current signal; , are the main frequency and main amplitude of voltage and current respectively; are the mean and standard deviation of the power signal respectively.

[0087] In one possible implementation, feature interaction adopts an attention-based approach.

[0088] Specifically, the dynamic environment feature vector and the discharge feature vector are mapped into the query, key, and value matrices so that the correlation between the two can be calculated through the attention weight. The calculation formula of the attention mechanism is as follows: in: They are query matrix, key matrix and value matrix respectively; ; is the weight matrix, which is used to map the feature vector to the high-dimensional feature space; is the dimension of the key vector, used to normalize the attention scores.

[0089] In the above formula The function is used to calculate the attention weight, which represents the correlation between the environmental features and the discharge features. Weighted average to generate interaction feature vector : As an option, feature interaction can also be achieved through simple vector concatenation or weighted fusion.

[0090] In some embodiments, the dynamic environment feature vector and the discharge feature vector are directly concatenated to form an initial interaction feature: In order to further improve the expressiveness of interactive features, the concatenated feature vectors can be transformed nonlinearly, for example, by using a fully connected neural network: in: is the weight matrix, is the bias term; is an activation function, such as the ReLU function or the sigmoid function.

[0091] In this embodiment, the interactive feature vector may also be expanded according to actual application scenarios.

[0092] Specifically, in some application scenarios, dynamic environmental features may have different effects on different dimensions of discharge features. For example, the frequency feature of temperature may be more sensitive to voltage fluctuations, while the amplitude feature of humidity may have a greater impact on current fluctuations. In this case, the multi-head attention mechanism can be introduced to calculate the weights of the impact of environmental features on different discharge features.

[0093] The calculation formula of the multi-head attention mechanism is:

[0094] in: Indicates Attention head; For the The weight matrix of the attention heads; is the output weight matrix, which is used to map the results of multi-head attention back to the output space.

[0095] Through the multi-head attention mechanism, the expressiveness of interactive features can be significantly improved, while more accurately capturing the complex impact of the dynamic environment on battery discharge performance.

[0096] The interactive feature vector generated by the feature interaction analysis in this step is It not only contains information about dynamic environmental characteristics, but also integrates the changing laws of discharge characteristics.

[0097] This interactive feature vector provides complete and rich data input for subsequent performance classification, and its results can better reflect the dynamic characteristics of battery performance as the environment changes. Whether through the attention mechanism or direct splicing and fusion, this step lays an important foundation for performance analysis and diagnosis.

[0098] S6. Based on the interactive feature vector, determine whether the discharge performance of the polymer lithium-ion battery is normal.

[0099] After completing the interactive analysis of the environmental features and the discharge features in step S5, an interactive feature vector has been generated, which integrates the complex relationship between the dynamic environment and the discharge signal. In order to further accurately judge the discharge performance of the battery, this step is based on the interactive feature vector, through a deep learning classification model or other appropriate analysis methods, to complete the classification and judgment of the discharge performance.

[0100] In general, the interactive feature vector contains rich dynamic information, such as the impact of temperature and humidity changes on voltage and current fluctuations. These features can reflect the potential abnormal state of battery performance, such as ion migration barriers inside the battery or voltage abnormalities caused by electrolyte decomposition. Therefore, accurate identification of these states requires processing the feature vector through a classifier or analysis model to determine whether the battery's discharge performance is normal.

[0101] In this embodiment, based on the interactive feature vector To judge the discharge performance of polymer lithium-ion batteries, the specific implementation method is as follows: Specifically, the interactive feature vector Dynamic environment features and discharge characteristics The fusion result is: in: Represents the main amplitude and frequency characteristics of the dynamic environment; Represents the statistical and frequency domain characteristics of the discharge signal.

[0102] Interaction feature vector It is input into the classification model to determine whether the battery discharge performance is normal.

[0103] In one possible implementation, the classification model uses a deep neural network (DNN) with the following structure: Input layer: receives the interaction feature vector , the feature dimension is ,in Hidden layer: consists of several fully connected layers, each containing neurons, and uses the activation function , such as the ReLU function: Output layer: Output the classification results of discharge performance , the probability of each category is normalized by the softmax function: in: is the output weight matrix; is the output of the last hidden layer; is the output layer bias term.

[0104] The training data of the model can be generated by the discharge performance data collected in the laboratory, and the training goal is to minimize the cross entropy loss: in: is the sample size; is the number of categories (e.g., two categories: “normal” and “abnormal”); For sample Belongs to category The true label of For the model to sample Belongs to category The predicted probability of .

[0105] As an option, a probability distribution analysis method can also be used to judge the discharge performance.

[0106] Specifically, the features of each dimension in the interaction feature vector are regarded as high-dimensional random variables, and the feature distribution under normal conditions is estimated by the Gaussian distribution model: in: is the feature mean vector; is the feature covariance matrix; is the feature dimension.

[0107] By calculating the probability density value of the interaction feature vector , to determine whether it deviates significantly from the normal distribution. If it is less than a certain threshold, it is considered to be an abnormal state.

[0108] In a possible implementation, the classification result may also be corrected in combination with rule-based diagnosis logic.

[0109] For example, when the temperature amplitude in the dynamic environment characteristic and humidity amplitude At the same time, when it exceeds a specific range, the battery performance can be directly judged as abnormal without relying on the results of the classifier. This method can effectively enhance the robustness of the model and is suitable for specific extreme environmental conditions.

[0110] In this embodiment, the classification result can be directly output as "normal" or "abnormal" and used to further optimize the design and use conditions of the battery.

[0111] By classifying the results, you can also analyze the specific environmental factors that affect battery performance. For example, if the judgment result shows that the humidity frequency Voltage fluctuation and main frequency If there is a significant correlation, it indicates that humidity fluctuation is the main cause of performance abnormalities. In this case, humidity adaptability can be improved by optimizing the sealing of the battery or adjusting the electrolyte formulation.

[0112] Through the classification analysis of this step, the discharge performance of polymer lithium-ion batteries can be accurately judged, providing an important basis for battery status monitoring and optimization. This step is closely combined with the aforementioned feature interaction analysis step, and the environmental and discharge features in the interactive feature vector are used to comprehensively analyze the changes in battery performance, thereby significantly improving the accuracy and efficiency of diagnosis.

[0113] The polymer lithium-ion battery test and analysis system based on deep learning and feature analysis includes: Signal acquisition module, used to collect discharge signals of polymer lithium-ion batteries and temperature and humidity data of the battery working environment; This module is used to collect real-time data on the working status of polymer lithium-ion batteries, including voltage, current and power signals generated during the discharge process, as well as temperature and humidity data in the battery working environment. The signal acquisition module collects raw data through high-precision sensors, and filters and denoises the data through the signal processing unit to ensure the accuracy and effectiveness of the data.

[0114] Implementation Use high-resolution voltage sensors and current sensors to collect discharge signals; Acquire environmental data through temperature and humidity sensors, and set the acquisition frequency according to dynamic change requirements, such as 10Hz or higher; The data preprocessing part includes low-pass filter to remove noise, as well as outlier detection and removal.

[0115] Dynamic environment modeling module, used to establish dynamic distribution models of temperature and humidity; The dynamic environment modeling module uses partial differential equations to model and simulate the dynamic distribution of ambient temperature and humidity based on the temperature and humidity data provided by the signal acquisition module. This module generates a dynamic model that describes the changes of temperature and humidity over time and space by simulating the diffusion, convection and external disturbances of temperature and humidity.

[0116] A feature extraction module, used to extract the dynamic features of temperature and humidity and the features of the discharge signal; The feature extraction module extracts the main dynamic features of temperature and humidity, such as amplitude and frequency, as well as the mean, standard deviation and frequency domain features of the discharge signal based on the data provided by the dynamic environment modeling module and the signal acquisition module to form a feature vector.

[0117] A feature interaction analysis module is used to interactively analyze dynamic environment features and discharge features to generate interactive feature vectors; This module fuses dynamic environment features and discharge features through feature interaction analysis to generate interactive feature vectors. Interaction analysis can capture the complex nonlinear relationship between dynamic environment and discharge performance.

[0118] The performance evaluation module is used to determine whether the discharge performance of the polymer lithium-ion battery is normal according to the interactive feature vector and output the performance evaluation result.

[0119] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A lithium battery testing and analysis method based on deep learning and feature analysis, characterized in that: include: S1. Establish a dynamic environment model to obtain the dynamic distribution of temperature and humidity in the battery use environment, where the dynamic distribution of temperature and humidity is based on the combined influence of diffusion, convection and external disturbance; S2, extract dynamic environmental features, and use the main amplitude and frequency of temperature and humidity as environmental feature vectors; S3. Based on the lithium ion diffusion model, a coupling relationship between lithium ion concentration distribution and environmental characteristics is established, wherein the lithium ion concentration distribution is affected by changes in temperature and humidity in a dynamic environment; S4, collecting the discharge signal of the polymer lithium-ion battery, and extracting the voltage, current and power signals as the discharge feature vector; S5, interactively analyzing the environment feature vector and the discharge feature vector to generate an interactive feature vector; S6. Based on the interactive feature vector, determine whether the discharge performance of the polymer lithium-ion battery is normal.

2. The lithium battery testing and analysis method based on deep learning and feature analysis according to claim 1, characterized in that: The dynamic environment model in step S1 describes the dynamic changes of temperature and humidity through the following partial differential equation: The change of temperature is determined by diffusion, convection and external disturbance terms; The change of humidity is determined by diffusion, convection and external disturbance.

3. The lithium battery testing and analysis method based on deep learning and feature analysis according to claim 1, characterized in that: Extracting the dynamic environment features in step S2 includes: performing time-frequency domain analysis on the dynamic distribution of temperature and humidity, obtaining their main amplitudes and frequencies, and forming a dynamic environment feature vector.

4. The lithium battery testing and analysis method based on deep learning and feature analysis according to claim 1, characterized in that: The lithium ion diffusion model in step S3 describes the influence of temperature and humidity on the diffusion coefficient of lithium ions, and adjusts the diffusion rate and reaction rate through the dynamic environment characteristic vector.

5. The lithium battery testing and analysis method based on deep learning and feature analysis according to claim 1, characterized in that: The discharge signal collected in step S4 includes voltage, discharge current and power value in a time series, and the discharge signal is preprocessed to obtain a discharge feature vector.

6. The lithium battery testing and analysis method based on deep learning and feature analysis according to claim 1, characterized in that: In step S5, a feature interaction analysis method is adopted to fuse the environment feature vector and the discharge feature vector through an attention mechanism to generate an interaction feature vector.

7. The lithium battery testing and analysis method based on deep learning and feature analysis according to claim 6, characterized in that: The attention mechanism is implemented by the following steps: Mapping the environment feature vector and the discharge feature vector into a query vector, a key vector and a value vector respectively; Calculate the similarity score between the query vector and the key vector; The value vectors are weighted averaged using the similarity scores to generate an interaction feature vector.

8. The lithium battery testing and analysis method based on deep learning and feature analysis according to claim 1, characterized in that: In the step S6, whether the discharge performance of the polymer lithium-ion battery is normal is judged by a classification model based on the interactive feature vector, and the classification model is a deep neural network classifier.

9. The lithium battery testing and analysis method based on deep learning and feature analysis according to claim 8, characterized in that: The classification model determines the battery performance through the probability distribution of the interactive feature vector and outputs the classification result of the normal state or the abnormal state.

10. A lithium battery test and analysis system based on deep learning and feature analysis, based on the lithium battery test and analysis method based on deep learning and feature analysis according to any one of claims 1 to 9, characterized in that: include: Signal acquisition module, used to collect discharge signals of polymer lithium-ion batteries and temperature and humidity data of the battery working environment; Dynamic environment modeling module, used to establish dynamic distribution models of temperature and humidity; A feature extraction module, used to extract the dynamic features of temperature and humidity and the features of the discharge signal; A feature interaction analysis module is used to interactively analyze dynamic environment features and discharge features to generate interactive feature vectors; The performance evaluation module is used to determine whether the discharge performance of the polymer lithium-ion battery is normal according to the interactive feature vector and output the performance evaluation result.

Images