CN116400244B - Abnormal detection method and device for energy storage batteries - Google Patents

Abstract

The present invention proposes an abnormality detection method and device for energy storage batteries, which relates to the technical field of energy storage batteries. The method includes: performing signal decomposition on the initial voltage signals of each battery cell in the energy storage battery to obtain multiple modal signals. ; Based on the similarity measure of the time series corresponding to each modal signal, determine the first characteristic signal that is inconsistent with the operating status of the battery cell; Based on the initial voltage signal corresponding to the size of the moving window of the display platform, conduct characteristic signals on multiple modal signals Extract to obtain the second characteristic signal of each battery cell; perform clustering processing on the first characteristic signal and the second characteristic signal based on the preset signal strength threshold to obtain multiple signal clusters with different signal strengths; and then determine Whether there are abnormal battery cells in the energy storage battery. Based on the signal cluster corresponding to the energy storage battery, various faults of the energy storage battery can be accurately identified to achieve early warning and improve the safety of the actual operation of the energy storage battery.

Description

Abnormal detection method and device for energy storage battery

Technical field

The present invention relates to the technical field of energy storage batteries, and in particular, to an abnormality detection method, device, electronic equipment and storage medium for energy storage batteries.

Background technique

In recent years, energy storage batteries are a typical type of energy storage device involving complex electrochemical reactions/transmission mechanisms. Energy storage batteries themselves have high safety risks. On the one hand, mechanical and electrical problems may occur during actual operation of energy storage batteries. , Thermal abuse, such as overcharging, over-discharging, overheating, etc., can easily cause rapid decline in battery performance, or even internal short circuit, causing safety issues. On the other hand, when applied in the field of large-scale energy storage, a large number of battery cells in the energy storage battery form battery packs, battery packs and even battery clusters. There will be a large number of connecting components, which greatly increases the complexity of the system and will lead to The probability of various types of failures increases and safety hazards increase.

In related technologies, most existing battery fault detection methods require accurate battery modeling, require high calculations, have poor adaptive capabilities for abnormality detection thresholds, have low diagnostic reliability, and are not suitable for actual battery management systems. (Battery Management System, BMS) application, therefore there is an urgent need for a lightweight energy storage battery anomaly detection method.

Contents of the invention

The present invention aims to solve one of the technical problems in the related art, at least to a certain extent.

To this end, the first purpose of the present invention is to propose an abnormality detection method for energy storage batteries to accurately identify various faults of energy storage batteries, achieve early warning, and improve the safety of actual operation of energy storage batteries.

The second object of the present invention is to provide an abnormality detection device for an energy storage battery.

The third object of the present invention is to provide an electronic device.

The fourth object of the present invention is to provide a non-transitory computer-readable storage medium storing computer instructions.

To achieve the above object, a first embodiment of the present invention proposes an abnormality detection method for an energy storage battery, wherein the energy storage battery includes at least one battery cell, and the method includes:

Obtain the initial voltage signal of each of the battery cells, and perform signal decomposition on the initial voltage signal to obtain multiple modal signals corresponding to each of the initial voltage signals;

According to the similarity measure of the time series corresponding to each of the modal signals, determine the first characteristic signal among the plurality of modal signals that is inconsistent with the operating state of the battery cell;

Based on the size of the moving window of the display platform corresponding to the initial voltage signal, perform feature signal extraction on the plurality of modal signals to obtain the second feature signal of each of the battery cells;

Based on a preset signal strength threshold, cluster the first characteristic signal and the second characteristic signal to obtain multiple signal clusters with different signal strengths;

According to the signal cluster, it is determined whether there is an abnormal battery cell in the energy storage battery.

To achieve the above object, a second embodiment of the present invention proposes an abnormality detection device for an energy storage battery, wherein the energy storage battery includes at least one battery cell, and the device includes:

An acquisition module, configured to acquire the initial voltage signal of each of the battery cells, and perform signal decomposition on the initial voltage signal to obtain multiple modal signals corresponding to each of the initial voltage signals;

A first determination module, configured to determine the first characteristic signal among the plurality of modal signals that is inconsistent with the operating status of the battery cell based on the similarity measure of the time series corresponding to each of the modal signals;

An extraction module, configured to perform feature signal extraction on the plurality of modal signals based on the size of the moving window of the display platform corresponding to the initial voltage signal, so as to obtain the second feature signal of each of the battery cells;

A clustering module configured to cluster the first characteristic signal and the second characteristic signal based on a preset signal strength threshold to obtain multiple signal clusters with different signal strengths;

The second determination module is used to determine whether there are abnormal battery cells in the energy storage battery according to the signal cluster.

To achieve the above object, a third embodiment of the present invention provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores information that can be used by the Instructions executed by at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the method described in the first aspect.

In order to achieve the above object, the fourth embodiment of the present invention provides a non-transient computer-readable storage medium storing computer instructions, and the computer instructions are used to cause the computer to execute the method described in the first aspect.

The abnormality detection method, device, electronic equipment and storage medium for energy storage batteries provided by embodiments of the present invention perform signal decomposition on the initial voltage signal of each battery cell in the energy storage battery to obtain multiple modal signals; according to each The similarity measure of the modal signal corresponding to the time series determines the first characteristic signal that is inconsistent with the operating status of the battery cell; based on the initial voltage signal corresponding to the size of the moving window of the display platform, feature signal extraction is performed on multiple modal signals to Obtain the second characteristic signal of each battery cell; cluster the first characteristic signal and the second characteristic signal based on the preset signal strength threshold to obtain multiple signal clusters with different signal strengths; and then determine the number of signal clusters in the energy storage battery. Whether there are abnormal battery cells, thus, based on the signal cluster corresponding to the energy storage battery, various faults of the energy storage battery can be accurately identified, early warning can be achieved, and the actual operation safety of the energy storage battery can be improved.

Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

Description of the drawings

The above and/or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

Figure 1 is a schematic flow chart of an abnormality detection method for energy storage batteries provided by an embodiment of the present invention;

Figure 2 is a schematic flow chart of another abnormality detection method for energy storage batteries provided by an embodiment of the present invention;

Figure 3 is a schematic structural diagram of an abnormality detection device for an energy storage battery provided by an embodiment of the present invention.

Detailed ways

Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present invention and are not to be construed as limiting the present invention.

Among them, it should be noted that the acquisition, storage, use, processing, etc. of data in the technical solution of the present invention all comply with the relevant provisions of national laws and regulations.

The following describes the abnormality detection method, device, electronic equipment and storage medium of the energy storage battery according to the embodiment of the present invention with reference to the accompanying drawings.

FIG. 1 is a schematic flowchart of an abnormality detection method for an energy storage battery provided by an embodiment of the present invention, where the energy storage battery includes at least one battery cell.

As shown in Figure 1, the method includes the following steps:

Step 101: Obtain the initial voltage signal of each battery cell, and perform signal decomposition on the initial voltage signal to obtain multiple modal signals corresponding to each initial voltage signal.

Alternatively, the energy storage battery may be a lithium-ion battery, but is not limited thereto, and is not specifically limited in this embodiment.

In some embodiments, an implementation method of obtaining the initial voltage signal of each battery cell and performing signal decomposition on the initial voltage signal to obtain multiple modal signals corresponding to each initial voltage signal may be: obtaining the initial voltage signal of each battery cell. The initial voltage signal of the body is decomposed to obtain multiple intrinsic mode functions corresponding to each initial voltage signal. When each intrinsic mode function becomes monotonic or meets the preset function termination criteria, Each intrinsic modal function is used as a modal signal, thereby achieving accurate decomposition of the initial voltage signal of each battery cell.

Specifically, when the energy storage battery is a lithium-ion battery, the initial voltage signal of each battery cell in the lithium-ion battery can be obtained through a voltage detection device, and then the initial voltage signal can be combined with empirical mode decomposition (EMD). The voltage signal is decomposed to obtain multiple intrinsic mode functions (IMF) corresponding to each initial voltage signal. If the frequency of the suspected IMF has a local maximum, repeat the above steps; if the suspected IMF becomes monotonic or If certain termination criteria are met, this IMF is regarded as the final modal signal.

Among them, the variational mode decomposition (VMD) algorithm can be used for the intrinsic mode function (IMF), which is related to the response to the corresponding external excitation or internal state in different frequency bands. VMD is used to evaluate the bandwidth of k components, and they are independently distributed around k center frequencies, that is, k IMFs. Since the number of IMFs is no less than 2, the IMFs in different main frequency bands are roughly divided into dynamic parts and static parts. As the transient response to the input current caused by the energy storage battery's participation in peak and frequency modulation, the dynamic component in the higher frequency band is considered to carry key information that characterizes specific anomalies. On the contrary, the difference in IMF in the lowest frequency band can be regarded as the main manifestation of inconsistent battery cell status.

Using the Hilbert transform in the frequency domain, the collected initial voltage signal time series can be converted into a constrained variational optimization problem, as shown in Equation (1):

Among them: u _k represents all modes of the initial voltage signal; ω _k represents its center frequency; f represents the original signal, δ represents the Dirac distribution; t represents the sampling time series; * represents the convolution operator. After introducing the quadratic penalty term and Lagrange multiplier, the expression of the constrained variational optimization problem is as shown in Equation (2):

Among them: α represents the hyperparameter of data fidelity constraints. It can be solved using the alternating direction method of multipliers. In the spectrum, the generated IMF is shown in equation (3):

Among them: f(ω), u _i (ω), λ(ω) and Respectively represent y(t), y _i (t), λ(t) and/> The Fourier transform of . The time domain component can be obtained from the initial voltage signal as the real part of the inverse Fourier transform with a Wiener filter structure.

Step 102: Based on the similarity measure of the time series corresponding to each modal signal, determine the first characteristic signal among the plurality of modal signals that is inconsistent with the operating status of the battery cell.

It can be understood that after decomposing the initial voltage signal and obtaining the corresponding multiple modal signals IMF, the initial voltage signal can be effectively decomposed through VMD to meet different analysis needs, and the offset caused by the inconsistent status of the battery cells The impact is smaller. Then the features of the IMF are extracted and selected, and distance-based evaluation of state inconsistency is performed.

Specifically, when the energy storage battery is applied in an industrial battery management system (BATTERY MANAGEMENT SYSTEM, BMS), since the computing power of the industrial BMS hardware is temporarily insufficient to support the online implementation of the model-based battery state estimation method, time can be used Sequence similarity measures are used to describe battery inconsistencies, such as Euclidean Distance (ED), Dynamic Time Warping (DTW) and Singular Value Decomposition (SVD). This embodiment is useful for This is not specifically limited.

Specifically, Euclidean distance (ED), dynamic time warping (DTW), and singular value decomposition (SVD). However, SVD has the risk of misjudgment when the time series is not multivariate, so it is more suitable for processing data from multiple sensors. Since the present invention only uses the initial voltage signal, ED and DTW are used. It is assumed that a certain voltage segment observed with N discrete sampling moments of M battery cells is expressed as Equation (4):

And the average voltage sequence Used as redundancy in distance calculations. In order to evaluate the difference between the jth battery cell voltage signal (j=1, 2,...,M) and the average voltage sequence, the sum of the absolute values of the differences between each two corresponding points in equation (5) is calculated as a whole ED:

The DTW distance with the minimum warp path W is calculated as shown in Equation (6):

Among them: w _k = (a, b) represents the point a and point a in v _j calculated using ED The detailed rules for the distance between points b in can be expressed as (7):

w ₁ = (1, 1) w _k = (N, N)

w _k = (a, b) w _k-1 = (a′, b′) (7)

Among them: 0≤a-a'≤1,0≤b-b'≤1

The similarity measure of the time series corresponding to each modal signal can be divided into segments under charging or discharging scenarios based on the timestamp, current and speed in the uploaded record. By using the VMD algorithm to obtain the components of each segment to evaluate the state inconsistency of the battery cells before feature extraction, the static components in the lowest frequency band can be effectively used for battery state inconsistency estimation.

In addition, in addition to incremental capacity analysis (ICA) through curve features related to aging mechanisms, DTW distance can also replace filtering algorithms and is more discriminable than ED.

Step 103: Extract characteristic signals from multiple modal signals based on the initial voltage signal corresponding to the size of the moving window of the display platform to obtain the second characteristic signal of each battery cell.

In some embodiments, since a battery that experiences internal short circuit or thermal runaway is not necessarily the battery with the worst consistency, it is also necessary to diagnose the dynamic component of the battery cell. Therefore, based on the initial voltage signal, the corresponding display platform By moving the size of the window, feature signals are extracted from multiple modal signals to obtain the second characteristic signal of each battery cell and realize dynamic component diagnosis of each battery cell.

Specifically, feature signals can be extracted from multiple modal signals based on a Graphics Device Interface (GDI) to obtain the second feature signal of each battery cell. For example, the GDI corresponding display platform can be determined. The size of the moving window N prevents small fault signals from being ignored when extracting signal features from longer similar voltage sequences. The dimensionless indicator (Device Interface, DI) is sensitive to abnormal signals rather than operating conditions, so it is used in industrial diagnosis. widely used. In battery fault diagnosis research based on current signals, correlation coefficient and information entropy are called key signal features, but the DI formula cannot correctly express these two. Since they are actually dimensionless in nature, there is a lack of generalized construction formulas for character construction. Therefore, in order to construct dimensionless signal characteristics, the generalized dimensionless index GDI is constructed according to equation (8):

Among them, z _i (v), s _j (v) represent functions for simple processing of input signal vectors v (t) collected from single or multiple sources; p (·) represents the probability density function of signal values; l _i , m _j represents different positive constants, n _z , n _s represents the order of GDI. The integral in equation (8) can be calculated by moving the discrete voltage signal in the window. The steps for GDI extraction are as follows:

a) In order to enhance the flexibility of the formula, use z _i (·), s _j (·) to process the components decomposed from the input voltage vector. z _i (·), s _j (·) can be various simple functions, such as Take logarithm, center deviation, absolute value, data normalization, etc.

b) Assign various positive constants to l _i and m _j to ensure diversity;

c) The order of GDI can be adjusted by the order constant n _z , n _s , because no matter what constants and signals are assigned, the numerator and denominator of GDI should have the same dimensions, and n _z , n _s are set to different values respectively to ensure The GDI numerator and denominator have the same dimensions;

The main parameter values n, l, and m in the present invention are positive odd numbers, positive odd numbers, and positive even numbers respectively (for example, n=3, l=3, m=2), and the GDI derived from the diversified parameter allocation can be further used for subsequent fusion of features.

Step 104: Perform clustering processing on the first characteristic signal and the second characteristic signal based on a preset signal strength threshold to obtain multiple signal clusters with different signal strengths.

In some embodiments, an implementation method of performing clustering processing on the first characteristic signal and the second characteristic signal based on a preset signal strength threshold to obtain multiple signal clusters with different signal strengths may be to first cluster the first characteristic signal and the second characteristic signal. The first characteristic signal and the second characteristic signal are subjected to dimensionality reduction processing of the data volume of the time characteristic, and the dimensionally reduced signals are clustered, thereby speeding up the efficiency of clustering processing of the first characteristic signal and the second characteristic signal, And achieve accurate classification of signal clusters.

The preset signal strength threshold may be determined based on the application scenario of the energy storage battery, or may be set by a technician, which is not specifically limited in this embodiment.

Step 105: Determine whether there are abnormal battery cells in the energy storage battery based on the signal cluster.

In some embodiments, an implementation method of determining whether there are abnormal battery cells in the energy storage battery based on the signal clusters may be to select the abnormal clusters in the signal clusters based on the cluster characteristics of each signal cluster, and then based on the abnormal clusters , determine the battery cell abnormalities corresponding to the abnormal clusters, thereby achieving early warning in the presence of abnormal cells and improving the safety, stability and reliability of the actual operation of the battery system.

The abnormality detection method of the energy storage battery according to the embodiment of the present invention performs signal decomposition on the initial voltage signal of each battery cell in the energy storage battery to obtain multiple modal signals; based on the similarity of the corresponding time series of each modal signal Measure and determine the first characteristic signal that is inconsistent with the operating status of the battery cell; based on the initial voltage signal corresponding to the size of the moving window of the display platform, perform feature signal extraction on multiple modal signals to obtain the second characteristics of each battery cell signal; based on the preset signal strength threshold, perform clustering processing on the first characteristic signal and the second characteristic signal to obtain multiple signal clusters with different signal strengths; and then determine whether there are abnormal battery cells in the energy storage battery, by Therefore, based on the signal cluster corresponding to the energy storage battery, various faults of the energy storage battery can be accurately identified, early warning can be achieved, and the actual operation safety of the energy storage battery can be improved.

In order to clearly illustrate the previous embodiment, FIG. 2 is a schematic flowchart of another abnormality detection method for energy storage batteries provided by an embodiment of the present invention.

Step 201: Obtain the initial voltage signal of each battery cell, and perform signal decomposition on the initial voltage signal to obtain multiple modal signals corresponding to each initial voltage signal.

Step 202: Based on the similarity measure of the time series corresponding to each modal signal, determine the first characteristic signal among the plurality of modal signals that is inconsistent with the operating status of the battery cell.

Step 203: Based on the initial voltage signal corresponding to the size of the moving window of the display platform, feature signal extraction is performed on multiple modal signals to obtain the second feature signal of each battery cell.

It should be noted that, regarding the specific implementation of steps 201 to 202, please refer to the relevant descriptions in the above embodiments.

Step 204: Perform dimensionality reduction processing on the data volume of time characteristics on the first characteristic signal and the second characteristic signal to obtain the reduced dimensionality of the first target characteristic signal and the second target characteristic signal.

In some embodiments, when feature signal extraction is performed on multiple modal signals based on a graphics device interface (GDI), since a large number of feature sequences are generated using GDI, many feature points in the original feature space are likely to be Identified as outliers. However, this situation can be effectively alleviated by implementing density-based clustering on the reduced feature space. Therefore, a manifold learning method can be used as an example to reduce the feature dimension to comprehensively utilize the key information in each feature dimension. .

Specifically, assuming that the data points (the first feature signal and the second feature signal) are uniformly sampled from a low-dimensional manifold in a high-dimensional Euclidean space, manifold learning is used to recover the low-dimensional manifold from the high-dimensional sampled data. structure. Dimensionality reduction of temporal features can be achieved by finding low-dimensional manifolds with corresponding embedding maps in high-dimensional space. In manifold learning, Laplacian eigenmap (LE) is used to achieve dimensionality reduction. The original similarity between points in each local region is calculated and expected to persist in low-dimensional space. The LE method generates dimensionality reduction features as follows:

a) Use the best weight to determine the nearest neighbor point

At a certain moment when the feature sequence F is fused, the weighted similarity calculated by the following equation (9) can be used To establish the k nearest neighbors of its element _fi :

Among them: k represents the dimension of the feature at this time, so the similarity matrix can be formulated as Equation (10):

When W is not a symmetric matrix, W can be further processed as (11):

b) Build the objective function

Construct an objective function to maintain the similarity in low-dimensional space, such as equation (12):

Let D=diag(d ₁ , d ₂ ,...d _n ), where i-1,2,...,n, so DW is a positive semidefinite matrix, and the objective function is expressed as Equation (13):

Therefore, the optimization problem becomes equation (14):

c) Solve the optimization problem Let M=D-W, so one eigenvalue of M is 0, and all values of the corresponding eigenvector are 1. After eigenvalue decomposition, the corresponding eigenvectors consisting of eigenvalues from the second minimum value to the (d+1)th minimum value are embedded into the d-dimensional matrix Y as the output.

Step 205: Cluster the first target characteristic signal and the second target characteristic signal based on a preset signal strength threshold to obtain multiple signal clusters with different signal strengths.

In some embodiments, an implementation method of clustering the first target characteristic signal and the second target characteristic signal based on a preset signal strength threshold to obtain multiple signal clusters with different signal strengths may be based on The density clustering method, combined with the preset signal strength threshold, clusters the first target characteristic signal and the second target characteristic signal to achieve accurate clustering of signal clusters.

Specifically, the clustering rules of the density-based clustering method can be used to collect coherent features characterizing similar responses of battery cells by forming various clusters. To obtain clustering of signal clusters with various characteristics to clearly detect and localize anomalies, the clusters are grouped by the following assumptions:

a) Normal instances (normal signal strength values) belong to one cluster in the signal cluster, and abnormal instances (abnormal signal strength values) do not belong to any signal cluster;

b) Normal instances are close to their nearest signal cluster centroid, and abnormal instances are far away from their nearest signal cluster centroid;

c) Normal instances belong to large and dense signal clusters, while abnormal instances belong to small or sparse signal clusters.

For the clustering algorithm that satisfies a), the clustering algorithm obtains clusters from the feature sequence with all signal strength thresholds set in advance. Furthermore, if anomalies in the data themselves form clusters, then an outlier may itself become the centroid and only member of the cluster, and methods satisfying b) cannot detect such anomalies. Therefore, c) is used for clustering, that is, if the second characteristic signal corresponding to the initial voltage signal of the battery cell belongs to a cluster whose size and/or density is lower than the signal strength threshold, the initial battery voltage signal is considered abnormal. Because actual energy storage battery failures are likely to be caused by a very small number of battery cells, without timely warning or leading to catastrophic accidents, density-based clustering is used to find abnormal signals in the second characteristic signal. Utilizing density-based clustering via minimization and optimization of centroid distance, clusters containing very few points (less than 2 points at the sampling moment) in the time feature series may suggest potential anomalies of the battery cells, which reduces Differences in judgment processes using diverse features.

In summary, supplementary correction can be achieved through the traditional signal mild threshold-based method to effectively eliminate misjudgments. Due to some inevitable severe fluctuations in operating conditions or measurement noise, moments with rather discrete distribution characteristics tend to be subject to false alarms at the alarm threshold. However, the corresponding second characteristic signal anomalies of battery cells that are not correctly detected at a certain moment are corrected by a density-based clustering method.

Step 206: Determine whether there are abnormal battery cells in the energy storage battery based on the signal cluster.

The abnormality detection method of the energy storage battery according to the embodiment of the present invention performs signal decomposition on the initial voltage signal of each battery cell in the energy storage battery to obtain multiple modal signals; based on the similarity of the corresponding time series of each modal signal Measure and determine the first characteristic signal that is inconsistent with the operating status of the battery cell; based on the initial voltage signal corresponding to the size of the moving window of the display platform, perform feature signal extraction on multiple modal signals to obtain the second characteristics of each battery cell Signal; perform dimensionality reduction processing on the data volume of time characteristics on the first abnormal signal and the second abnormal signal to obtain the first target characteristic signal and the second target characteristic signal after dimensionality reduction; based on the preset signal strength threshold, The first target characteristic signal and the second target characteristic signal are clustered to obtain a plurality of signal clusters with different signal strengths. ; and then determine whether there are abnormal battery cells in the energy storage battery. Therefore, based on the signal cluster corresponding to the energy storage battery, various types of faults such as gradual and sudden faults can be accurately and effectively identified to achieve early warning and improve the actual operation of the battery system. security, stability and reliability.

In order to implement the above embodiments, the present invention also proposes an abnormality detection device for an energy storage battery.

Figure 3 is a schematic structural diagram of an abnormality detection device for an energy storage battery provided by an embodiment of the present invention.

As shown in FIG. 3 , the energy storage battery anomaly detection device 30 includes: an acquisition module 31 , a first determination module 32 , an extraction module 33 , a clustering module 34 and a second determination module 35 .

The acquisition module 31 is used to acquire the initial voltage signal of each of the battery cells, and perform signal decomposition on the initial voltage signal to obtain multiple modal signals corresponding to each of the initial voltage signals;

The first determination module 32 is configured to determine the first characteristic signal among the plurality of modal signals that is inconsistent with the operating state of the battery cell based on the similarity measure of the time series corresponding to each of the modal signals;

The extraction module 33 is configured to perform feature signal extraction on the plurality of modal signals based on the size of the moving window of the display platform corresponding to the initial voltage signal, so as to obtain the second feature signal of each of the battery cells;

The clustering module 34 is configured to perform clustering processing on the first characteristic signal and the second characteristic signal based on a preset signal strength threshold to obtain multiple signal clusters with different signal strengths;

The second determination module 35 is used to determine whether there are abnormal battery cells in the energy storage battery according to the signal cluster.

Further, in a possible implementation of the embodiment of the present invention, the acquisition module 31 is specifically used to:

Obtain the initial voltage signal of each of the battery cells, and perform signal decomposition on the initial voltage signal to obtain multiple intrinsic mode functions corresponding to each of the initial voltage signals;

When each of the intrinsic mode functions becomes monotonic or meets a preset function termination criterion, each of the intrinsic mode functions is used as a modal signal.

Further, in a possible implementation of the embodiment of the present invention, the clustering module 34 is specifically used to:

Perform dimensionality reduction processing on the data volume of time characteristics on the first characteristic signal and the second characteristic signal to obtain the first target characteristic signal and the second target characteristic signal after dimensionality reduction;

Based on a preset signal strength threshold, the first target characteristic signal and the second target characteristic signal are clustered to obtain multiple signal clusters with different signal strengths.

Further, in a possible implementation of the embodiment of the present invention, the second determination module 35 is specifically used to:

Based on the cluster characteristics of each signal cluster, select abnormal clusters in the signal cluster;

According to the abnormality cluster, it is determined that the battery cell abnormality corresponding to the abnormality cluster is abnormal.

It should be noted that the foregoing explanation of the method embodiment also applies to the device of this embodiment, and will not be described again here.

The anomaly detection device for an energy storage battery according to an embodiment of the present invention performs signal decomposition on the initial voltage signal of each battery cell in the energy storage battery to obtain multiple modal signals; based on the similarity of the corresponding time series of each modal signal Measure and determine the first characteristic signal that is inconsistent with the operating status of the battery cell; based on the initial voltage signal corresponding to the size of the moving window of the display platform, perform feature signal extraction on multiple modal signals to obtain the second characteristics of each battery cell signal; based on the preset signal strength threshold, perform clustering processing on the first characteristic signal and the second characteristic signal to obtain multiple signal clusters with different signal strengths; and then determine whether there are abnormal battery cells in the energy storage battery, by Therefore, based on the signal cluster corresponding to the energy storage battery, various faults of the energy storage battery can be accurately identified, early warning can be achieved, and the actual operation safety of the energy storage battery can be improved.

In order to implement the above embodiments, the present invention also provides an electronic device, including:

at least one processor; and

a memory communicatively connected to the at least one processor; wherein,

The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the foregoing method.

In order to implement the above embodiments, the present invention also proposes a non-transitory computer-readable storage medium storing computer instructions, and the computer instructions are used to cause the computer to execute the aforementioned method.

In the description of this specification, reference to the terms "one embodiment," "some embodiments," "an example," "specific examples," or "some examples" or the like means that specific features are described in connection with the embodiment or example. , structures, materials or features are included in at least one embodiment or example of the invention. In this specification, the schematic expressions of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the specific features, structures, materials or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, those skilled in the art may combine and combine different embodiments or examples and features of different embodiments or examples described in this specification unless they are inconsistent with each other.

In addition, the terms “first” and “second” are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of indicated technical features. Therefore, features defined as "first" and "second" may explicitly or implicitly include at least one of these features. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise expressly and specifically limited.

Any process or method descriptions in flowcharts or otherwise described herein may be understood to represent modules, segments, or portions of code that include one or more executable instructions for implementing customized logical functions or steps of the process. , and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed out of the order shown or discussed, including in a substantially simultaneous manner or in the reverse order, depending on the functionality involved, which shall It should be understood by those skilled in the art to which embodiments of the present invention belong.

The logic and/or steps represented in the flowcharts or otherwise described herein, for example, may be considered a sequenced list of executable instructions for implementing the logical functions, and may be embodied in any computer-readable medium, For use by, or in combination with, instruction execution systems, devices or devices (such as computer-based systems, systems including processors or other systems that can fetch instructions from and execute instructions from the instruction execution system, device or device) or equipment. For the purposes of this specification, a "computer-readable medium" may be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer readable media include the following: electrical connections with one or more wires (electronic device), portable computer disk cartridges (magnetic device), random access memory (RAM), Read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium may even be paper or other suitable medium on which the program may be printed, as the paper or other medium may be optically scanned, for example, and subsequently edited, interpreted, or otherwise suitable as necessary. process to obtain the program electronically and then store it in computer memory.

It should be understood that various parts of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if it is implemented in hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: discrete logic gate circuits with logic functions for implementing data signals; Logic circuits, application specific integrated circuits with suitable combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

Those of ordinary skill in the art can understand that all or part of the steps involved in implementing the methods of the above embodiments can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable storage medium. The program can be stored in a computer-readable storage medium. When executed, one of the steps of the method embodiment or a combination thereof is included.

In addition, each functional unit in various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically alone, or two or more units can be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software function modules. If the integrated module is implemented in the form of a software function module and is sold or used as an independent product, it can also be stored in a computer-readable storage medium.

The storage media mentioned above can be read-only memory, magnetic disks or optical disks, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above-mentioned embodiments are illustrative and should not be construed as limitations of the present invention. Those of ordinary skill in the art can make modifications to the above-mentioned embodiments within the scope of the present invention. The embodiments are subject to changes, modifications, substitutions and variations.

Claims (10)

1. An abnormality detection method for energy storage batteries, characterized in that, the energy storage battery includes at least one battery cell, and the method includes: Obtain the initial voltage signal of each of the battery cells, and perform signal decomposition on the initial voltage signal to obtain multiple modal signals corresponding to each of the initial voltage signals; According to the similarity measure of the time series corresponding to each of the modal signals, determine the first characteristic signal among the plurality of modal signals that is inconsistent with the operating state of the battery cell; Based on the size of the moving window of the display platform corresponding to the initial voltage signal, perform feature signal extraction on the plurality of modal signals to obtain the second feature signal of each of the battery cells; Based on a preset signal strength threshold, perform clustering processing on the first characteristic signal and the second characteristic signal to obtain multiple signal clusters with different signal strengths; According to the signal cluster, it is determined whether there is an abnormal battery cell in the energy storage battery. 2. The method according to claim 1, characterized in that: obtaining the initial voltage signal of each of the battery cells, and performing signal decomposition on the initial voltage signal to obtain the corresponding voltage signal of each of the initial voltage signals. Multiple modal signals, including: Obtain the initial voltage signal of each of the battery cells, and perform signal decomposition on the initial voltage signal to obtain multiple intrinsic mode functions corresponding to each of the initial voltage signals; When each of the intrinsic mode functions becomes monotonic or meets a preset function termination criterion, each of the intrinsic mode functions is used as a modal signal. 3. The method according to claim 1, characterized in that, based on a preset signal strength threshold, the first characteristic signal and the second characteristic signal are clustered to obtain a plurality of different signal strengths. Signal clusters, including: Perform dimensionality reduction processing on the data volume of time characteristics on the first characteristic signal and the second characteristic signal to obtain the first target characteristic signal and the second target characteristic signal after dimensionality reduction; Based on a preset signal strength threshold, the first target characteristic signal and the second target characteristic signal are clustered to obtain multiple signal clusters with different signal strengths. 4. The method according to claim 1, wherein determining whether there is an abnormal battery cell in the energy storage battery according to the signal cluster includes: Based on the cluster characteristics of each signal cluster, select abnormal clusters in the signal cluster; According to the abnormality cluster, it is determined that the battery cell abnormality corresponding to the abnormality cluster is abnormal. 5. An abnormality detection device for an energy storage battery, characterized in that, the energy storage battery includes at least one battery cell, and the device includes: An acquisition module, configured to acquire the initial voltage signal of each of the battery cells, and perform signal decomposition on the initial voltage signal to obtain multiple modal signals corresponding to each of the initial voltage signals; A first determination module, configured to determine the first characteristic signal among the plurality of modal signals that is inconsistent with the operating status of the battery cell based on the similarity measure of the time series corresponding to each of the modal signals; An extraction module, configured to perform feature signal extraction on the plurality of modal signals based on the size of the moving window of the display platform corresponding to the initial voltage signal, so as to obtain the second feature signal of each of the battery cells; A clustering module, configured to perform clustering processing on the first characteristic signal and the second characteristic signal based on a preset signal strength threshold to obtain multiple signal clusters with different signal strengths; The second determination module is used to determine whether there are abnormal battery cells in the energy storage battery according to the signal cluster. 6. The device according to claim 5, characterized in that the acquisition module is specifically used for: Obtain the initial voltage signal of each of the battery cells, and perform signal decomposition on the initial voltage signal to obtain multiple intrinsic mode functions corresponding to each of the initial voltage signals; When each of the intrinsic mode functions becomes monotonic or meets a preset function termination criterion, each of the intrinsic mode functions is used as a modal signal. 7. The device according to claim 5, characterized in that the clustering module is specifically used for: Perform dimensionality reduction processing on the data volume of time characteristics on the first characteristic signal and the second characteristic signal to obtain the first target characteristic signal and the second target characteristic signal after dimensionality reduction; Based on a preset signal strength threshold, the first target characteristic signal and the second target characteristic signal are clustered to obtain multiple signal clusters with different signal strengths. 8. The device according to claim 5, characterized in that the second determination module is specifically used to: Based on the cluster characteristics of each signal cluster, select abnormal clusters in the signal cluster; According to the abnormality cluster, it is determined that the battery cell abnormality corresponding to the abnormality cluster is abnormal. 9. An electronic device, characterized in that it includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can perform any one of claims 1-4. Methods. 10. A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the method according to any one of claims 1-4.