Fault diagnosis method, system and equipment for lithium ion battery and medium

Abstract

The invention discloses a lithium ion battery fault diagnosis method, system and device and a medium, and the method is characterized in that the method comprises the steps: obtaining the voltage data, temperature data and impedance data of a lithium ion battery; performing feature extraction on the voltage data, the temperature data and the impedance data to obtain a voltage feature, a temperature feature and an impedance feature, calculating a first coupling feature based on the voltage feature and the temperature feature, and calculating a second coupling feature based on the voltage feature and the impedance feature, corresponding abnormal deviation degrees are calculated based on the voltage features, the temperature features, the impedance features, the first coupling features and the second coupling features, corresponding feature weights are determined based on the abnormal deviation degrees, and the voltage features, the temperature features, the impedance features, the first coupling features and the second coupling features are fused according to the feature weights; obtaining a target fusion feature; and determining a fault diagnosis result of the lithium ion battery based on the target fusion feature. According to the invention, the lithium ion battery fault diagnosis precision can be improved.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis, and in particular to a fault diagnosis method, system, device and medium for lithium-ion batteries. Background Technology

[0002] Lithium-ion batteries are widely used in electric vehicles, energy storage systems, and portable electronic devices due to their high energy density, long cycle life, and high charge / discharge efficiency. However, in actual operation, lithium-ion batteries are susceptible to factors such as the working environment, operating conditions, and their own aging, which can lead to faults such as abnormal internal resistance, abnormal temperature, and voltage fluctuations. Failure to identify the battery's operating status and potential faults in a timely and accurate manner can not only result in decreased battery performance and shortened lifespan but may also trigger safety risks such as thermal runaway. Therefore, real-time and accurate fault diagnosis of lithium-ion batteries is a crucial technical means to ensure the safe and stable operation of battery systems.

[0003] Existing lithium-ion battery fault diagnosis methods are mostly based on single-mode data or fixed-weight multi-parameter threshold judgment methods. For example, they rely solely on voltage changes, temperature rise rate, or internal resistance changes for fault identification, or use pre-set static weights for comprehensive judgment in multi-parameter joint analysis. However, under complex battery operating conditions and fault characteristics with multi-source coupling and dynamic changes, the above methods are difficult to accurately reflect the actual contribution of different modal parameters in different operating stages and different fault evolution processes. This can easily lead to misjudgment or omission, resulting in insufficient fault diagnosis accuracy and failing to meet the actual needs for early identification and refined diagnosis of lithium-ion battery faults. Summary of the Invention

[0004] This invention provides a method, system, device, and medium for fault diagnosis of lithium-ion batteries, which can improve the accuracy of fault diagnosis of lithium-ion batteries.

[0005] In a first aspect, embodiments of the present invention provide a fault diagnosis method for lithium-ion batteries, comprising:

[0006] Obtain voltage, temperature, and impedance data of lithium-ion batteries;

[0007] Feature extraction is performed on the voltage data, temperature data, and impedance data respectively to obtain voltage features, temperature features, and impedance features. Based on the voltage features, temperature features, and impedance features, corresponding abnormal deviation degrees are determined. Based on each abnormal deviation degree, corresponding feature weights are determined. The voltage features, temperature features, impedance features, first coupling features, and second coupling features are fused according to each feature weight to obtain target fused features. The first coupling feature is calculated from the voltage features and temperature features, and the second coupling feature is calculated from the voltage features and impedance features.

[0008] Based on the target fusion features, the fault diagnosis results of the lithium-ion battery are determined.

[0009] This invention provides a more comprehensive and reliable data foundation for subsequent fault characteristic analysis by acquiring multi-source operational data reflecting the electrical state, thermal state, and internal equivalent characteristics of lithium-ion batteries, thereby improving the accuracy of lithium-ion battery fault diagnosis. By extracting features from raw voltage, temperature, and impedance data, the high-dimensional, noisy raw data is transformed into feature representations that stably reflect changes in battery operating state, thus improving the accuracy of lithium-ion battery fault diagnosis. Furthermore, by characterizing the correlation between battery electrical and thermal behavior, the ability to represent complex fault characteristics such as thermal runaway and abnormal temperature rise is enhanced, contributing to improved accuracy of lithium-ion battery fault diagnosis. Finally, by coupling voltage and impedance features into a model, the invention comprehensively reflects both the battery's external output characteristics and internal characteristics. The relationship between state changes is analyzed to improve the ability to identify early faults and potential anomalies, thereby further improving the accuracy of lithium-ion battery fault diagnosis. By evaluating the degree of anomalousness of various features relative to the normal operating state and adaptively determining the weight of each feature at the current moment based on the degree of anomalousness, a feature weight allocation that is more consistent with the actual operating state is achieved, effectively improving the accuracy of lithium-ion battery fault diagnosis. By weighted fusion of multi-source features and coupled features based on feature weights, the expressive ability of fused features for complex fault modes is improved, thereby improving the accuracy of lithium-ion battery fault diagnosis. By using target fused features containing multi-source information and their coupling relationships for fault judgment, accurate identification of lithium-ion battery fault states is achieved, ultimately improving the accuracy of lithium-ion battery fault diagnosis.

[0010] Furthermore, the step of extracting features from the voltage data, temperature data, and impedance data to obtain voltage features, temperature features, and impedance features includes:

[0011] Calculate a first variance value of the voltage data to determine the voltage characteristics based on the first variance value;

[0012] Calculate the first absolute difference between the temperature data and the historical temperature data, and determine the temperature feature based on the first absolute difference;

[0013] Calculate the second absolute difference between the impedance data and the preset reference impedance value, and determine the impedance characteristics based on the second absolute difference.

[0014] This invention extracts differentiated features from voltage, temperature, and impedance data, enabling voltage features to reflect voltage fluctuations, temperature features to reflect temperature change trends, and impedance features to reflect impedance shifts. This allows for a more accurate characterization of abnormal changes in the electrical, thermal, and performance aspects of lithium-ion batteries, providing a reliable foundation for subsequent anomaly detection and feature fusion, thereby improving the accuracy of lithium-ion battery fault diagnosis.

[0015] Furthermore, the calculation of the corresponding abnormal deviation based on the voltage characteristic, the temperature characteristic, the impedance characteristic, the first coupling characteristic, and the second coupling characteristic includes:

[0016] A first coupling characteristic is calculated based on the voltage characteristic and the temperature characteristic, and a second coupling characteristic is calculated based on the voltage characteristic and the impedance characteristic;

[0017] The corresponding characteristic values ​​are determined based on the voltage characteristic, the temperature characteristic, the impedance characteristic, the first coupling characteristic, and the second coupling characteristic;

[0018] Calculate the moving average and moving standard deviation for each of the aforementioned eigenvalues;

[0019] Calculate a third absolute difference between each of the aforementioned feature values ​​and each of the aforementioned moving average values, and determine the corresponding degree of abnormal deviation based on the third absolute difference.

[0020] This invention compares voltage, temperature, impedance characteristics and their coupling characteristics with their respective dynamic statistical benchmarks to quantify the degree of deviation of each characteristic from its historical state at the current moment, thereby achieving a unified characterization of abnormal changes in multimodal characteristics. This method can effectively suppress the influence of normal operating condition fluctuations on characteristic judgment, highlight abnormal characteristic changes caused by fault evolution, and provide a reliable abnormal measurement basis for subsequent feature weighted fusion and fault diagnosis.

[0021] Furthermore, the calculation of the first coupling feature based on the voltage feature and the temperature feature includes:

[0022] The voltage sequence and temperature sequence are determined based on the voltage characteristics and the temperature characteristics, respectively.

[0023] Calculate the correlation coefficient between the voltage sequence and the temperature sequence to determine the first coupling feature based on the correlation coefficient.

[0024] This invention, through the calculation of the correlation coefficient between voltage and temperature sequences, couples and models the battery's electrical and thermal characteristics. This quantitatively characterizes the intrinsic relationship between voltage and temperature changes, thus forming a first coupled feature reflecting the battery's operating state. This approach is beneficial for characterizing multi-source coupled fault features that are difficult to reflect with a single modal feature, improving the accuracy and reliability of identifying abnormal operating conditions and early faults.

[0025] Furthermore, the calculation of the second coupling characteristic based on the voltage characteristic and the impedance characteristic includes:

[0026] Calculate the difference between the impedance characteristic and the historical impedance characteristic at the current moment, and determine the amount of impedance change based on the difference;

[0027] Calculate the second variance value corresponding to the voltage characteristic and the third variance value corresponding to the historical voltage characteristic, and determine the voltage fluctuation amount based on the second variance value and the third variance value;

[0028] The ratio of the impedance change to the voltage fluctuation is calculated to determine the second coupling characteristic based on the ratio.

[0029] This invention employs a ratio-based coupling analysis of impedance changes and voltage fluctuations to model the correlation between internal battery aging characteristics and external voltage stability changes. This highlights the impact of impedance anomalies on voltage fluctuations, thus forming a physically meaningful second coupling feature. This approach helps distinguish between faults caused by internal resistance degradation and voltage changes caused by operating condition fluctuations, reducing the risk of misdiagnosis and improving the sensitivity and accuracy of diagnosing impedance-related faults.

[0030] Furthermore, the step of calculating the moving average and moving standard deviation corresponding to each of the aforementioned feature values ​​includes:

[0031] Based on a preset attenuation factor, each of the aforementioned feature values ​​and the historical moving average are weighted and summed to obtain the moving average corresponding to each of the aforementioned feature values;

[0032] Based on the attenuation factor, each of the eigenvalues, each of the moving averages, and the historical moving variances corresponding to each of the eigenvalues, the moving standard deviation corresponding to each of the eigenvalues ​​is determined.

[0033] This invention integrates the current feature value with the historical moving average by introducing a weighting method with an attenuation factor. This allows the moving average to better reflect recent feature change trends. Furthermore, it dynamically calculates the moving standard deviation by combining the historical moving variance, thereby constructing a feature statistical baseline that is adaptively updated over time. This avoids the problem of fixed mean and variance lagging in response to changes in operating status, improves the accuracy of abnormal deviation calculation, and thus enhances the accuracy of lithium-ion battery fault diagnosis.

[0034] Furthermore, determining the corresponding feature weights based on each of the abnormal deviations includes: normalizing the abnormal deviations corresponding to the voltage feature, the temperature feature, the impedance feature, the first coupling feature, and the second coupling feature to obtain the feature weights corresponding to each of the abnormal deviations.

[0035] This invention normalizes the abnormal deviations corresponding to voltage features, temperature features, impedance features, first coupling features, and second coupling features, enabling features with different dimensions and varying amplitudes to be mapped to a unified weight scale. This also allows features with higher abnormality levels to gain greater influence during the fusion process, thereby achieving adaptive adjustment of feature contribution, enhancing the sensitivity of multi-feature fusion results to abnormal states, and improving the accuracy and stability of lithium-ion battery fault diagnosis.

[0036] Secondly, embodiments of the present invention provide a lithium-ion battery fault diagnosis system, the system comprising: an acquisition module, a fusion module, and a comparison module;

[0037] The acquisition module is used to acquire voltage data, temperature data, and impedance data of the lithium-ion battery.

[0038] The fusion module is used to extract features from the voltage data, temperature data, and impedance data respectively to obtain voltage features, temperature features, and impedance features; determine the corresponding abnormal deviation degree based on the voltage features, temperature features, and impedance features; determine the corresponding feature weight based on each abnormal deviation degree; and fuse the voltage features, temperature features, impedance features, first coupling features, and second coupling features according to each feature weight to obtain the target fused features. The first coupling feature is calculated from the voltage features and temperature features, and the second coupling feature is calculated from the voltage features and impedance features.

[0039] The comparison module is used to determine the fault diagnosis result of the lithium-ion battery based on the target fusion features.

[0040] This invention introduces multi-source feature extraction of voltage, temperature, and impedance features at the system level, and further constructs voltage-temperature coupling features and voltage-impedance coupling features. This enables the system to simultaneously characterize the single-mode abnormal behavior and multi-physical quantity coupling abnormal behavior of lithium-ion batteries. At the same time, based on the abnormal deviation degree corresponding to each feature, the feature weights are dynamically determined and weighted fusion is performed, so that features with higher abnormality degrees dominate in the fusion result. This achieves adaptive and refined discrimination of lithium-ion battery fault states, improving the accuracy and stability of fault diagnosis and the ability to identify early faults.

[0041] Thirdly, embodiments of the present invention provide a terminal device, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus;

[0042] The memory is used to store at least one executable instruction that causes the processor to perform the operation of a fault diagnosis method for a lithium-ion battery as described in this application.

[0043] Fourthly, embodiments of the present invention provide a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device or system where the computer-readable storage medium is located to perform a fault diagnosis method for a lithium-ion battery as described in this application.

[0044] The above description is merely an overview of the technical solutions of the embodiments of the present invention. In order to better understand the technical means of the embodiments of the present invention and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0045] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0046] Figure 1 This is a schematic flowchart of an embodiment of a fault diagnosis method for lithium-ion batteries provided in this application;

[0047] Figure 2 This is a flowchart illustrating steps S201 to S203 provided in this application;

[0048] Figure 3This is a flowchart illustrating steps S301 to S304 provided in this application;

[0049] Figure 4 This is a flowchart illustrating steps S401 to S403 provided in this application;

[0050] Figure 5 This is a schematic diagram of an embodiment of a fault diagnosis method for lithium-ion batteries provided in this application. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0052] Unless otherwise defined, 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 application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0053] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0054] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0055] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0056] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0057] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0058] Lithium-ion batteries are widely used in electric vehicles, energy storage systems, and portable electronic devices due to their high energy density, long cycle life, and high charge / discharge efficiency. However, in actual operation, lithium-ion batteries are susceptible to changes in the working environment, fluctuations in operating conditions, and battery aging, leading to fault phenomena such as abnormal internal resistance, abnormal temperature, and voltage fluctuations. If the battery's operating status and potential faults cannot be identified in a timely and accurate manner, it will not only cause battery performance degradation and shorten its lifespan but may also trigger safety hazards such as thermal runaway. Existing lithium-ion battery fault diagnosis methods mostly rely on single-mode data or use fixed-weight threshold judgment methods in multi-parameter analysis, such as judging based solely on changes in voltage, temperature, or internal resistance, or fusing multiple parameters through static weights. However, given the complex operating conditions of batteries and the multi-source coupling and dynamic changes of fault characteristics over time, it is difficult to accurately characterize the actual contribution of different parameters at different operating stages and in the fault evolution process. This easily leads to misjudgments or omissions, resulting in insufficient fault diagnosis accuracy and failing to meet the application requirements for early fault identification and refined diagnosis of lithium-ion batteries.

[0059] See Figure 1 To improve the accuracy of fault diagnosis of lithium-ion batteries, an embodiment of the present invention provides a fault diagnosis method for lithium-ion batteries, including steps S101 to S103.

[0060] Step S101: Obtain the voltage data, temperature data, and impedance data of the lithium-ion battery;

[0061] In some embodiments, a multimodal data acquisition device is used to synchronously acquire voltage, temperature, and impedance data of the lithium-ion battery under test at the current moment and at multiple historical moments adjacent to the current moment. First, voltage data can be acquired using a voltage acquisition device connected to the electrodes of the lithium-ion battery under test. This voltage acquisition device can be a voltage sensor or a voltage detection unit integrated in a battery management system (BMS), used to acquire the terminal voltage value of the lithium-ion battery in real time according to a preset sampling period. For example, the sampling period for voltage data can be 1 second or shorter to capture small voltage fluctuations caused by early internal short-circuit faults. Second, temperature data can be acquired using a temperature acquisition device placed on the surface of the lithium-ion battery or near the cell. This temperature acquisition device can be a thermistor, thermocouple, or other temperature sensor, used to acquire the operating temperature of the lithium-ion battery in real time. Considering that temperature changes typically have a certain lag, the sampling period for temperature data can be the same as or slightly longer than that for voltage data, for example, 1 to 5 seconds. Finally, impedance data can be acquired using an impedance acquisition device. In this embodiment, the impedance acquisition device employs a single-frequency AC excitation method. A small-amplitude AC current signal of a preset frequency is applied to the lithium-ion battery under test, and the corresponding AC voltage response is measured to calculate the impedance data of the lithium-ion battery. For example, the frequency of the AC excitation signal can be 1 kHz, and the real part of the complex impedance is selected as the impedance data to reduce computational complexity and improve real-time performance.

[0062] In some embodiments, the collected voltage, temperature, and impedance data are aligned according to a unified timestamp and stored as multimodal time series data to ensure the consistency of different modal data in the time dimension, thereby providing a reliable data foundation for subsequent multimodal data feature extraction based on time windows.

[0063] Through the above steps, highly time-consistent acquisition and alignment processing of multi-modal operating data such as voltage, temperature, and impedance of lithium-ion batteries are achieved, enabling joint analysis of different physical quantities under the same time reference. On the one hand, high-frequency voltage sampling is beneficial for capturing subtle electrical changes caused by early battery faults; on the other hand, the synchronous acquisition of temperature and impedance data can reflect the battery's thermal behavior and internal state evolution characteristics, thus providing a complete and reliable data foundation for subsequent multi-modal feature extraction and coupled analysis, improving the timeliness and accuracy of fault diagnosis.

[0064] Step S102: Feature extraction is performed on the voltage data, temperature data, and impedance data respectively to obtain voltage features, temperature features, and impedance features. Based on the voltage features, temperature features, and impedance features, the corresponding abnormal deviation degree is determined. Based on each abnormal deviation degree, the corresponding feature weight is determined. The voltage features, temperature features, impedance features, first coupling features, and second coupling features are fused according to each feature weight to obtain the target fused features. The first coupling feature is calculated from the voltage features and temperature features, and the second coupling feature is calculated from the voltage features and impedance features.

[0065] Please refer to Figure 2 In some embodiments, the step of extracting features from the voltage data, the temperature data, and the impedance data to obtain voltage features, temperature features, and impedance features includes steps S201 to S203.

[0066] Step S201: Calculate the first variance value of the voltage data to determine the voltage characteristics based on the first variance value;

[0067] In some embodiments, voltage data of the lithium-ion battery under test at the current time t and several historical times are acquired. Using the current time t as the time endpoint, a preset voltage time window is constructed by retrospectively counting backwards for a preset duration. Within the preset voltage time window, the number of voltage sampling points is N, and the corresponding voltage data are V. t-N+1 V t-N+2 ,…,V t First, calculate the average voltage within the preset voltage time window, which can be expressed as:

[0068]

[0069] in, V represents the average voltage value within a voltage time window ending at the current time t. t-i Let represent the voltage data collected at time ti; N represents the number of voltage sampling points included in the voltage time window. Based on this, calculate the first variance of the voltage data within the preset voltage time window, which can be expressed as:

[0070]

[0071] Among them, Var V (t) represents the voltage variance value corresponding to the current time t, i.e., the first variance value. In this embodiment, the first variance value Var V(t) is determined as the voltage characteristic corresponding to the current moment. Since the stability of the terminal voltage of a lithium-ion battery will decrease significantly when an internal short circuit fault occurs, even under normal current fluctuation conditions, the amplitude of small voltage fluctuations will increase significantly. Therefore, characterizing the voltage characteristics by using the variance of the voltage data can sensitively capture abnormal voltage fluctuations related to internal short circuit faults.

[0072] Step S202: Calculate the first absolute difference between the temperature data and the historical temperature data, and determine the temperature feature based on the first absolute difference;

[0073] In some embodiments, temperature data of the lithium-ion battery under test at the current time t and several historical times are acquired. Using the current time t as the end point, a preset temperature time window is constructed by tracing back a preset duration. Within the preset temperature time window, the number of temperature sampling points is M, and the corresponding temperature data are as follows: T t-M+1 ,T t-M+2 ,…,T t Next, the absolute difference between the temperature data at two adjacent moments within the preset temperature time window is calculated. This absolute difference can be expressed as:

[0074] ΔT t-j =|T t-j -T t-j-1 |;

[0075] Among them, T t-j and T t-j-1 These represent the temperature data at two adjacent sampling times; ΔT t-j This represents the corresponding temperature change range. Based on this, the average of multiple absolute temperature differences within the preset temperature time window is calculated to obtain the first absolute difference, which can be expressed as:

[0076]

[0077] Among them, D T (t) represents the temperature dynamics corresponding to the current time t. In this embodiment, the first absolute difference D is... T (t) is determined as the temperature characteristic corresponding to the current moment. Since internal short circuit faults in lithium-ion batteries continuously generate heat, causing the battery temperature to show a continuous and slow upward trend, calculating the average change of temperature data over a short period of time can effectively reflect the abnormal heat generation state of the battery, thereby improving the accuracy of internal short circuit fault diagnosis.

[0078] Step S203: Calculate the second absolute difference between the impedance data and the preset reference impedance value, so as to determine the impedance characteristics based on the second absolute difference.

[0079] In some embodiments, impedance data Z of the lithium-ion battery under test is acquired at the current time t. t The impedance data can be obtained using a single-frequency AC excitation method, and in this embodiment of the invention, the real part of the complex impedance is taken as the impedance data. A reference impedance value Z is predetermined. ref The reference impedance value can be: the initial impedance value of the lithium-ion battery under test in a healthy state; or the average impedance value of lithium-ion batteries of the same model and batch as the lithium-ion battery under test, in a healthy state. Based on this, a second absolute difference between the impedance data at the current moment and the reference impedance value is calculated. The second absolute difference can be expressed as:

[0080] D Z (t)=|Z t -Z ref |;

[0081] Among them, D Z (t) represents the impedance change at the current time t. In this embodiment, the second absolute difference D is... Z (t) is determined as the impedance characteristic at the current moment. Because the internal electrochemical structure of a lithium-ion battery changes in the early stages of an internal short-circuit fault, resulting in a continuous decrease or abnormal change in the equivalent impedance, comparing the current impedance with the reference impedance can effectively reflect the abnormal state of the battery's internal structure.

[0082] Please refer to Figure 3 In some embodiments, the step of calculating the corresponding abnormal deviation based on the voltage characteristic, the temperature characteristic, the impedance characteristic, the first coupling characteristic, and the second coupling characteristic includes: steps S301 to S304;

[0083] Step S301: Calculate the first coupling feature based on the voltage feature and the temperature feature, and calculate the second coupling feature based on the voltage feature and the impedance feature;

[0084] In some embodiments, taking the current sampling time t as the time endpoint, voltage and temperature feature sequences covering multiple consecutive sampling times are selected to characterize the correlation between voltage and temperature changes in the time dimension. Correlation analysis is performed on the voltage and temperature feature sequences to calculate the correlation coefficient between them, which is then used as the first coupling feature to characterize the electro-thermal coupling degree of the lithium-ion battery under its current operating state. When abnormal discharge or internal short circuit tends to occur inside the battery, voltage fluctuations and temperature changes often exhibit strong synchronicity, significantly increasing the value of the first coupling feature. Simultaneously, the impedance feature corresponding to the current time and the impedance feature corresponding to a preset historical time are selected, and the change between them is calculated to reflect the evolution trend of impedance over a short period. The voltage feature changes corresponding to the current time and historical times are calculated respectively to obtain the voltage fluctuation. Based on this, the ratio of the impedance change to the voltage fluctuation is calculated to obtain the second coupling feature, which characterizes the sensitivity of impedance changes to voltage fluctuations. When an abnormality occurs in the internal structure of the battery but the overall voltage fluctuation is not yet significant, the second coupling feature can amplify the expression of abnormal impedance changes. Step S302: Determine the corresponding feature value based on the voltage feature, the temperature feature, the impedance feature, the first coupling feature, and the second coupling feature;

[0085] In some embodiments, at the current sampling time t, the voltage characteristic, temperature characteristic, impedance characteristic, first coupling characteristic, and second coupling characteristic are uniformly represented as the set of characteristic values ​​corresponding to the current time:

[0086]

[0087] Wherein, F1(t) represents the voltage characteristic value at the current moment; F2(t) represents the temperature characteristic value at the current moment; F3(t) represents the impedance characteristic value at the current moment; F4(t) represents the first coupling characteristic value (voltage-temperature coupling characteristic) at the current moment; and F5(t) represents the second coupling characteristic value (voltage-impedance coupling characteristic) at the current moment. All of the above characteristic values ​​are obtained from the corresponding modal data or coupling characteristic calculation steps, and are used to uniformly enter the subsequent anomaly analysis process, thereby achieving multi-feature collaborative evaluation.

[0088] Step S303: Calculate the moving average and moving standard deviation corresponding to each of the aforementioned feature values;

[0089] In some embodiments, for each feature value F i (t), constructing a corresponding feature time window with the current time t as the endpoint, and determining the corresponding decay factor α based on the length of the time window. i ,in:

[0090]

[0091] Where, α i N represents the attenuation factor corresponding to the i-th feature; i This represents the number of sampling points within the feature time window corresponding to the i-th feature. Based on this, the moving average μ corresponding to the i-th feature at the current time... i (t) is calculated recursively as follows:

[0092] μ i (t)=α i ×F i (t)+(1-α i )×μ i (t-1);

[0093] Where, μ i (t) represents the moving average of the i-th feature at the current sampling time; F i (t) represents the feature value of the i-th feature at the current sampling time; μ i (t-1) represents the moving average of this feature at the previous sampling time. Further, based on the current moving average, the moving variance var corresponding to the i-th feature is... i (t) is calculated as follows:

[0094] var i (t)=α i ×[F i (t)-μ i (t)] 2 +(1-α i )×var i (t-1);

[0095] Where, var i (t) represents the moving variance of the i-th feature at the current sampling time; F i (t) represents the feature value at the current sampling time; μ i (t) represents the moving average value at the current sampling time; var i (t-1) represents the historical moving variance corresponding to the previous sampling time. Therefore, the moving standard deviation of the i-th feature at the current time is obtained:

[0096]

[0097] Where, σ i (t) represents the moving standard deviation of the i-th feature at the current sampling time. Using the above recursive method, the statistical benchmarks of each feature can be continuously updated without relying on complete historical data.

[0098] Step S304: Calculate the third absolute difference between each of the said feature values ​​and each of the said moving average values, so as to determine the corresponding abnormal deviation based on the third absolute difference.

[0099] In some embodiments, for each feature F i (t), calculate the third absolute difference D between it and the corresponding moving average. i (t), specifically:

[0100] D i (t)=|F i (t)-μ i (t)|;

[0101] Among them, D i (t) represents the third absolute difference of the i-th feature at the current time; F i (t) represents the characteristic value at the current time; μ i (t) represents the corresponding moving average. Based on this, the corresponding moving standard deviation σ is used... i (t), determine the abnormal deviation A of the i-th feature at the current time. i (t), for example, can be represented as:

[0102]

[0103] Among them, A i (t) represents the abnormal deviation degree corresponding to the i-th feature; ε is a preset minimum positive number used to avoid the denominator being zero. By normalizing the change of feature values ​​relative to their dynamic statistical benchmark, features with different dimensions and different rates of change can be compared on the same scale, providing a unified basis for subsequent feature weight determination and fusion processing.

[0104] In some embodiments, calculating the first coupling feature based on the voltage feature and the temperature feature includes: determining a voltage sequence and a temperature sequence based on the voltage feature and the temperature feature, respectively; calculating a correlation coefficient between the voltage sequence and the temperature sequence, so as to determine the first coupling feature based on the correlation coefficient.

[0105] In some embodiments, the voltage sequence and temperature sequence are determined based on the voltage feature and the temperature feature, respectively. Specifically, to calculate the coupling relationship between the voltage feature and the temperature feature, a time window for voltage-temperature coupling analysis is constructed by taking the current time t as the time endpoint and looking back for a preset duration. This time window is called the voltage-temperature time window. The voltage-temperature time window contains both voltage data and temperature data, and covers the main time period from the onset of abnormal voltage changes to the occurrence of a temperature response. For example, the duration of the voltage-temperature time window can be set to 2 to 5 minutes. Within the voltage-temperature time window, the voltage features at corresponding times are extracted in chronological order to form a voltage sequence; and the temperature features at corresponding times are extracted to form a temperature sequence. Specifically, if the voltage-temperature time window contains L consecutive sampling times, the voltage sequence can be represented as:

[0106] V=[V(t-L+1),V(t-L+2),…,V(t)];

[0107] Where V represents the voltage sequence; V(ti) represents the voltage characteristic at time ti; and L represents the number of sampling points within the voltage-temperature time window. Correspondingly, the temperature sequence can be represented as:

[0108] T=[T(t-L+1),T(t-L+2),…,T(t)];

[0109] Where T represents the temperature sequence; T(ti) represents the temperature feature at time ti. It should be noted that the voltage and temperature features mentioned above are not the original voltage and temperature data, but rather voltage micro-variable features and temperature dynamic features extracted by the methods defined in the aforementioned embodiments, thereby effectively suppressing noise interference and highlighting dynamic change information related to internal short-circuit faults.

[0110] In some embodiments, a correlation coefficient between the voltage sequence and the temperature sequence is calculated to determine the first coupling feature based on the correlation coefficient. Specifically, after obtaining the voltage sequence V and the temperature sequence T, a correlation coefficient between them is calculated to characterize the degree of synchronization between voltage changes and temperature changes in the time dimension, thereby depicting the electro-thermal coupling relationship. In this embodiment, the Pearson correlation coefficient is used to calculate the correlation coefficient. Specifically, the correlation coefficient between the voltage sequence and the temperature sequence can be expressed as:

[0111]

[0112] Where, ρ VT(t) represents the voltage-temperature correlation coefficient at the current time t, i.e., the first coupling characteristic; V(ti) represents the voltage characteristic at time ti; T(ti) represents the temperature characteristic at time ti. This represents the mean of the voltage sequence V over a voltage-temperature time window; The value of temperature sequence T within the voltage-temperature time window is represented by ρ; L represents the number of sampling points within the voltage-temperature time window. In this embodiment, the correlation coefficient ρ is... VT (t) is directly identified as the first coupling characteristic (voltage-temperature coupling characteristic). When an internal short circuit occurs in a lithium-ion battery, the abnormal current will simultaneously cause voltage fluctuations and local heating, resulting in a clear synchronous relationship between voltage changes and temperature changes in time. Therefore, by calculating the correlation coefficient between the voltage characteristic and the temperature characteristic, it is possible to effectively distinguish between the electro-thermal coupling anomaly caused by the internal short circuit and the non-coupling change caused by changes in ambient temperature or fluctuations in normal operating conditions.

[0113] Please refer to Figure 4 In some embodiments, the calculation of the second coupling feature based on the voltage feature and the impedance feature includes: steps S401 to S403;

[0114] Step S401: Calculate the difference between the impedance characteristic and the historical impedance characteristic at the current moment, so as to determine the impedance change based on the difference;

[0115] In some embodiments, the impedance feature is the impedance change feature extracted at the current time, and the historical impedance feature is the impedance change feature corresponding to a set historical time. The set historical time is earlier than the current time, and the time interval between the set historical time and the current time is used to characterize the short-term evolution process of the impedance change. Specifically, the impedance change can be obtained by subtracting the impedance feature at the current time t from the impedance feature at the set historical time t0, and the calculation formula is as follows:

[0116] ΔZ(t) = |Z(t) - Z(t0)|;

[0117] Where ΔZ(t) represents the impedance change at the current moment; Z(t) represents the impedance characteristic at the current moment t; Z(t0) represents the impedance characteristic at the set historical moment t0; and |·| represents the absolute value operation. In this embodiment, the set historical moment can be selected as a historical sampling moment within a preset time interval before the current moment, for example, a historical moment 1 to 3 minutes away from the current moment, thereby suppressing instantaneous noise interference while highlighting the continuous change trend of impedance caused by the evolution of internal short-circuit faults.

[0118] Step S402: Calculate the second variance value corresponding to the voltage characteristic and the third variance value corresponding to the historical voltage characteristic, and determine the voltage fluctuation amount based on the second variance value and the third variance value;

[0119] In some embodiments, the voltage characteristic is the voltage fluctuation characteristic corresponding to the current moment, and the historical voltage characteristic is the voltage fluctuation characteristic corresponding to a set historical moment. Both are characterized by the variance of voltage data within the corresponding time window. Specifically, within a preset voltage time window ending at the current moment t, the voltage variance value corresponding to the current moment is calculated based on the voltage data at each moment within the window, serving as the second variance value; simultaneously, within the corresponding voltage time window ending at a set historical moment t0, the voltage variance value corresponding to the historical voltage characteristic is calculated, serving as the third variance value. The voltage fluctuation can be expressed as:

[0120] ΔV(t)=|Var V (t)-Var V (t0)|;

[0121] Where ΔV(t) represents the voltage fluctuation at the current moment; Var V (t) represents the second variance value corresponding to the voltage characteristic at the current time t; Var V (t0) represents the third-party difference value corresponding to the voltage characteristics at a set historical time t0. In this embodiment, since the impact of an internal short-circuit fault on the voltage in the early stage is usually manifested as a small and gradually increasing fluctuation, by calculating the difference in voltage variance values ​​at different times, the instantaneous impact of current condition fluctuations can be effectively weakened, thereby more accurately reflecting the voltage stability changes caused by fault evolution.

[0122] Step S403: Calculate the ratio of the impedance change to the voltage fluctuation to determine the second coupling characteristic based on the ratio.

[0123] In some embodiments, the second coupling feature is used to characterize the coupling relationship between impedance change and voltage fluctuation, reflecting the sensitivity of impedance change under unit voltage fluctuation, thereby enhancing the amplification capability of early internal short-circuit characteristics. Specifically, the second coupling feature can be obtained by calculating the ratio of the impedance change obtained in step S301 to the voltage fluctuation obtained in step S302, as shown in the following formula:

[0124]

[0125] Among them, C ZVΔZ(t) represents the second coupling feature (impedance-voltage coupling feature) at the current moment; ΔZ(t) represents the impedance change at the current moment; ΔV(t) represents the voltage fluctuation at the current moment; ε represents a preset minimum positive number to avoid numerical instability caused by a zero denominator. In this embodiment, when the internal short-circuit fault is in its early stage, the impedance may have already undergone continuous changes, while the overall voltage fluctuation is not yet obvious. At this time, the value of the second coupling feature will be significantly increased, thereby amplifying the expression of small impedance changes. This allows the feature to obtain a higher abnormal response capability during the fusion process, thereby improving the early diagnosis accuracy of internal short-circuit faults in lithium-ion batteries.

[0126] In some embodiments, calculating the moving average and moving standard deviation corresponding to each of the eigenvalues ​​includes: weighting and summing each of the eigenvalues ​​and the historical moving average based on a preset decay factor to obtain the moving average corresponding to each of the eigenvalues; and determining the moving standard deviation corresponding to each of the eigenvalues ​​based on the decay factor, each of the eigenvalues, each of the moving averages, and the historical moving variance corresponding to each of the eigenvalues.

[0127] In some embodiments, the moving average value corresponding to each feature value is obtained by weighted summation of each feature value and the historical moving average value based on a preset attenuation factor. Specifically, for the i-th feature, at the current sampling time t, the attenuation factor α is determined based on the feature time window length corresponding to that feature. i ,in:

[0128]

[0129] Where, α i N represents the attenuation factor corresponding to the i-th feature; i This represents the number of sampling points within the feature time window corresponding to the i-th feature. After obtaining the attenuation factor, the feature value F at the current sampling time is... i (t) and the moving average μ of the previous sampling time i (t-1) is weighted and summed to obtain the moving average μ corresponding to the current sampling time. i (t), its calculation formula is:

[0130] μ i (t)=α i ×F i (t)+(1-α i )×μ i (t-1);

[0131] Where, μ i (t) represents the moving average of the i-th feature at the current sampling time; F i(t) represents the feature value of the i-th feature at the current sampling time; μ i (t-1) represents the moving average value of this feature at the previous sampling time. This method allows the moving average value to reflect the current trend of the feature while maintaining a continuous description of historical states.

[0132] In some embodiments, the moving standard deviation corresponding to each feature value is determined based on the attenuation factor, each feature value, each moving average value, and the historical moving variance corresponding to each feature value. Specifically, this involves obtaining the moving average value μ corresponding to the current sampling time. i After (t), the moving variance var corresponding to the i-th feature is further calculated based on the deviation between the current feature value and the moving average. i (t), its calculation formula is:

[0133] var i (t)=α i ×[F i (t)-μ i (t)] 2 +(1-α i )×var i (t-1);

[0134] Where, var i (t) represents the moving variance of the i-th feature at the current sampling time; F i (t) represents the feature value at the current sampling time; μ i (t) represents the moving average value at the current sampling time; var i (t-1) represents the historical moving variance corresponding to the previous sampling time. Based on this, the square root operation is performed on the moving variance to obtain the moving standard deviation σ corresponding to the i-th feature at the current sampling time. i (t), specifically:

[0135]

[0136] Where, σ i (t) represents the moving standard deviation of the i-th feature at the current sampling time. By employing an exponentially decaying moving average and moving standard deviation calculation method, the statistics can be dynamically updated over time, thus providing a stable and real-time reference benchmark for subsequent analysis of feature anomalies.

[0137] In some embodiments, determining the corresponding feature weights based on each of the abnormal deviations includes: normalizing the abnormal deviations corresponding to the voltage feature, the temperature feature, the impedance feature, the first coupling feature, and the second coupling feature to obtain the feature weights corresponding to each of the abnormal deviations.

[0138] In some embodiments, the abnormal deviations corresponding to the voltage feature, temperature feature, impedance feature, first coupling feature, and second coupling feature are normalized to obtain the feature weights corresponding to each abnormal deviation. Specifically, a normalization method based on proportional constraints is used to process each abnormal deviation. Specifically, the ratio of each abnormal deviation to the sum of all abnormal deviations is calculated to obtain the corresponding normalized value, and the calculation formula is as follows:

[0139]

[0140] Among them, W i (t) represents the feature weight corresponding to the i-th feature at the current sampling time; D i (t) represents the abnormal deviation of the i-th feature at the current sampling time; This represents the sum of the abnormal deviations corresponding to the five features; ε is a preset minimum positive number used to avoid numerical instability issues when the denominator is zero. Through the above normalization process, the weights of each feature satisfy: This allows abnormal deviations of different dimensions and magnitudes to be uniformly mapped to the same weight space.

[0141] It should be noted that, under the above normalization rules, the greater the abnormal deviation of a certain feature, the more significant the deviation of that feature from its historical dynamic baseline at the current sampling time, and the greater its weight W in the fusion process. i The larger the corresponding value (t), the lower the proportion of features with smaller anomaly deviations in the fusion. By adaptively determining feature weights based on anomaly deviations, the influence of voltage features, temperature features, impedance features, and coupling features on the fusion results can be dynamically adjusted according to the real-time operating status, providing a reliable weight basis for the subsequent construction of target fusion features.

[0142] Through the above steps, an adaptive fusion modeling of the multimodal operating characteristics and coupling relationships of lithium-ion batteries is achieved. On the one hand, by extracting statistical features reflecting the dynamic fluctuation characteristics of voltage, temperature, and impedance data respectively, and further constructing voltage-temperature and voltage-impedance coupling features, the intrinsic correlation between the battery's electro-thermal and electrochemical behaviors can be explicitly characterized, thereby enhancing the characterization capability of multi-source synergistic features of early faults such as internal short circuits. On the other hand, by introducing an anomaly deviation calculation mechanism based on moving statistics, the degree of anomaly of each feature relative to its historical benchmark can be dynamically measured without relying on a fixed threshold, and the weight of each feature in the fusion process can be adaptively determined accordingly, so that features that contribute more significantly to the current fault state receive higher weights, thereby significantly improving the diagnostic accuracy and real-time performance of lithium-ion battery faults, especially early faults of internal short circuits.

[0143] Step S103: Based on the target fusion features, determine the fault diagnosis result of the lithium-ion battery.

[0144] In some embodiments, the target fusion feature F corresponding to the target sampling time is first obtained. fusion (t), where the target fusion feature is calculated based on voltage feature, temperature feature, impedance feature, first coupling feature, and second coupling feature, combined with the dynamic weights corresponding to each feature, and its expression is:

[0145]

[0146] Among them, F fusion (t) represents the target fusion feature corresponding to the target sampling time t; F i (t) represents the feature value of the i-th feature at the target sampling time; W i (t) represents the feature weight corresponding to the i-th feature at the target sampling time. The feature weight is obtained by normalizing the anomaly deviation corresponding to the feature, and satisfies the following conditions:

[0147] In some embodiments, the target fusion feature is compared with a preset fault determination threshold Θ to determine the operating state of the lithium-ion battery at the target sampling time, and the determination rule is as follows:

[0148]

[0149] Wherein, Θ represents the preset fault judgment threshold, which can be obtained from historical normal operating condition data; State(t) = 1 indicates that the lithium-ion battery is in an abnormal or fault-risk state at the target sampling time; State(t) = 0 indicates that the lithium-ion battery is in a normal state at the target sampling time. Through this method, the battery operating state at a single sampling time can be quickly determined.

[0150] Through the above steps, a unified quantitative characterization and rapid identification of the operating state of lithium-ion batteries is achieved. This method can comprehensively reflect the combined effects of voltage, temperature, impedance, and their coupling anomalies in the target fusion feature, avoiding the decision inconsistency problem caused by independent judgment of multiple indicators, thus achieving reliable identification of abnormal battery states under a single judgment rule. Simultaneously, based on the adaptively determined feature weights according to the anomaly deviation, features that contribute more significantly to the current fault evolution play a dominant role in the judgment, improving the sensitivity of fault judgment to early anomalies and the ability to suppress normal fluctuations, thereby enhancing the accuracy and real-time performance of lithium-ion battery fault diagnosis results.

[0151] like Figure 5As shown, based on the above method embodiments, corresponding apparatus embodiments are provided;

[0152] An embodiment of the present invention provides a schematic diagram of a lithium-ion battery fault diagnosis system, the system comprising: an acquisition module 100, a fusion module 200, and a comparison module 300;

[0153] The acquisition module 100 is used to acquire voltage data, temperature data, and impedance data of the lithium-ion battery.

[0154] The fusion module 200 is used to extract features from the voltage data, temperature data, and impedance data respectively to obtain voltage features, temperature features, and impedance features; calculate a first coupling feature based on the voltage features and temperature features; calculate a second coupling feature based on the voltage features and impedance features; calculate corresponding abnormal deviation degrees based on the voltage features, temperature features, impedance features, first coupling features, and second coupling features respectively; determine corresponding feature weights based on each abnormal deviation degree; and fuse the voltage features, temperature features, impedance features, first coupling features, and second coupling features according to each feature weight to obtain a target fused feature.

[0155] The comparison module 300 is used to determine the fault diagnosis result of the lithium-ion battery based on the target fusion features.

[0156] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the fault diagnosis method for lithium-ion batteries provided by any of the above-described method embodiments of the present invention. More detailed workflows and principles of this system can be found, but are not limited to, the relevant descriptions of the above methods.

[0157] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0158] Based on the above-described embodiment of a fault diagnosis method for lithium-ion batteries, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a fault diagnosis method for lithium-ion batteries according to any embodiment of the present invention.

[0159] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0160] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0161] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0162] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute a fault diagnosis method for a lithium-ion battery as described in any of the above-described method embodiments of the present invention.

[0163] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0164] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that 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 for those skilled in the art.

Claims

1. A fault diagnosis method for lithium-ion batteries, characterized in that, include: Obtain voltage, temperature, and impedance data of lithium-ion batteries; Feature extraction is performed on the voltage data, temperature data, and impedance data respectively to obtain voltage features, temperature features, and impedance features. Based on the voltage features, temperature features, and impedance features, corresponding abnormal deviation degrees are determined. Based on each abnormal deviation degree, corresponding feature weights are determined. The voltage features, temperature features, impedance features, first coupling features, and second coupling features are fused according to each feature weight to obtain target fused features. The first coupling feature is calculated from the voltage features and temperature features, and the second coupling feature is calculated from the voltage features and impedance features. Based on the target fusion features, the fault diagnosis results of the lithium-ion battery are determined.

2. The fault diagnosis method for a lithium-ion battery as described in claim 1, characterized in that, The step of extracting features from the voltage data, temperature data, and impedance data to obtain voltage features, temperature features, and impedance features includes: Calculate a first variance value of the voltage data to determine the voltage characteristics based on the first variance value; Calculate the first absolute difference between the temperature data and the historical temperature data, and determine the temperature feature based on the first absolute difference; Calculate the second absolute difference between the impedance data and the preset reference impedance value, and determine the impedance characteristics based on the second absolute difference.

3. The fault diagnosis method for a lithium-ion battery as described in claim 1, characterized in that, The determination of the corresponding abnormal deviation based on the voltage characteristic, the temperature characteristic, and the impedance characteristic includes: A first coupling characteristic is calculated based on the voltage characteristic and the temperature characteristic, and a second coupling characteristic is calculated based on the voltage characteristic and the impedance characteristic; The corresponding characteristic values ​​are determined based on the voltage characteristic, the temperature characteristic, the impedance characteristic, the first coupling characteristic, and the second coupling characteristic; Calculate the moving average and moving standard deviation for each of the aforementioned eigenvalues; Calculate a third absolute difference between each of the aforementioned feature values ​​and each of the aforementioned moving average values, and determine the corresponding degree of abnormal deviation based on the third absolute difference.

4. The fault diagnosis method for a lithium-ion battery as described in claim 3, characterized in that, The calculation of the first coupling feature based on the voltage feature and the temperature feature includes: The voltage sequence and temperature sequence are determined based on the voltage characteristics and the temperature characteristics, respectively. Calculate the correlation coefficient between the voltage sequence and the temperature sequence to determine the first coupling feature based on the correlation coefficient.

5. A fault diagnosis method for a lithium-ion battery as described in claim 3, characterized in that, The calculation of the second coupling characteristic based on the voltage characteristic and the impedance characteristic includes: Calculate the difference between the impedance characteristic and the historical impedance characteristic at the current moment, and determine the amount of impedance change based on the difference; Calculate the second variance value corresponding to the voltage characteristic and the third variance value corresponding to the historical voltage characteristic, and determine the voltage fluctuation amount based on the second variance value and the third variance value; The ratio of the impedance change to the voltage fluctuation is calculated to determine the second coupling characteristic based on the ratio.

6. A fault diagnosis method for a lithium-ion battery as described in claim 3, characterized in that, The calculation of the moving average and moving standard deviation corresponding to each of the aforementioned feature values ​​includes: Based on a preset attenuation factor, each of the aforementioned feature values ​​and the historical moving average are weighted and summed to obtain the moving average corresponding to each of the aforementioned feature values; Based on the attenuation factor, each of the eigenvalues, each of the moving averages, and the historical moving variances corresponding to each of the eigenvalues, the moving standard deviation corresponding to each of the eigenvalues ​​is determined.

7. A fault diagnosis method for a lithium-ion battery as described in claim 1, characterized in that, The step of determining the corresponding feature weights based on each of the abnormal deviations includes: normalizing the abnormal deviations corresponding to the voltage feature, the temperature feature, the impedance feature, the first coupling feature, and the second coupling feature to obtain the feature weights corresponding to each of the abnormal deviations.

8. A lithium-ion battery fault diagnosis system, characterized in that, The system includes: an acquisition module, a fusion module, and a comparison module; The acquisition module is used to acquire voltage data, temperature data, and impedance data of the lithium-ion battery. The fusion module is used to extract features from the voltage data, temperature data, and impedance data respectively to obtain voltage features, temperature features, and impedance features; determine the corresponding abnormal deviation degree based on the voltage features, temperature features, and impedance features; determine the corresponding feature weight based on each abnormal deviation degree; and fuse the voltage features, temperature features, impedance features, first coupling features, and second coupling features according to each feature weight to obtain the target fused features. The first coupling feature is calculated from the voltage features and temperature features, and the second coupling feature is calculated from the voltage features and impedance features. The comparison module is used to determine the fault diagnosis result of the lithium-ion battery based on the target fusion features.

9. A terminal device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements a fault diagnosis method for a lithium-ion battery as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, include: A stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform a fault diagnosis method for a lithium-ion battery as described in any one of claims 1-7.

Images