Sparse data-oriented model and data fusion-driven energy storage lithium battery micro-short circuit and low-capacity fault diagnosis and classification method

Abstract

The invention discloses a sparse data-oriented model and data fusion-driven energy storage lithium battery micro-short circuit and low-capacity fault diagnosis and classification method, which belongs to the field of battery management systems and comprises the following steps of: estimating ohmic internal resistance of a battery by combining a first-order integer-order equivalent circuit model and a Rint model; and fault classification is carried out by combining the relative SOC and the relative capacity of the battery. Through application of the Rint model, effective classification of battery faults can be completed under the condition that the battery state is not completely known. The importance of the technical point lies in that the method can still maintain high fault diagnosis precision under the condition of low sampling density, and is of great significance to battery management and fault early warning in practical application.

Description

Technical Field

[0001] The present invention belongs to the field of battery management systems, and in particular relates to a sparse data-oriented model and data fusion-driven method for diagnosing and classifying micro-short circuit and low-capacity faults of energy storage lithium batteries. Background Art

[0002] In the field of fault diagnosis of photovoltaic energy storage systems and electric energy storage systems, some significant progress has been made in recent years. The existing technical framework mainly focuses on data analysis methods, using a large amount of historical operation data and real-time monitoring data for fault detection and diagnosis. Based on the data to form a sample set of equipment operation status, by calculating the distance between samples and establishing an abnormal sample judgment mechanism, the diagnostic efficiency and accuracy are effectively improved.

[0003] However, current methods generally rely on a large number of labeled samples and high-quality historical data, which not only increases labor costs, but also makes it impossible to obtain reliable diagnostic results when small samples or sparse data are present. Secondly, these methods are not adaptable enough to the dynamic changes of energy storage systems. When the structure or operating mode of the system changes, the model based on historical data often cannot be adjusted quickly, which affects the accuracy of fault detection. In addition, existing technologies are weak in detecting hidden faults, such as micro-short circuit faults, which may lead to serious consequences, but traditional methods have failed to effectively deal with their complexity. Finally, when processing large-scale complex data, many existing technologies also have problems with computational efficiency. This performance bottleneck often reduces the efficiency and real-time performance of overall fault detection when faced with real-time monitoring and diagnosis. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a sparse data-oriented model and data fusion-driven energy storage lithium battery micro-short circuit and low capacity fault diagnosis and classification method, including:

[0005] Obtaining the operating data uploaded by the battery management system of the industrial and commercial energy storage equipment, performing data cleaning on the operating data, and obtaining cleaned data;

[0006] Performing feature analysis and selection on the cleaned data to obtain a feature data set;

[0007] Calculating the characteristic data set based on the DBSCAN algorithm to obtain the faulty battery, performing data interpolation and data alignment processing on the battery cluster current and cluster SOC data of the faulty battery to obtain the cluster current and cluster SOC information of the interpolated BCU;

[0008] Calculate the average ohmic internal resistance of the faulty battery based on a first-order integer-order battery equivalent circuit model;

[0009] Interpolating the single cell voltage information collected in the BMU based on the interpolated cluster current and cluster SOC information of the BCU, and fitting the normal battery average OCV-SOC curve according to the Rint model and the average ohmic internal resistance;

[0010] Calculating a relative SOC change value based on the average OCV-SOC curve, calculating a relative capacity based on the relative SOC change value, and designing a fault diagnosis and classification threshold;

[0011] The relative capacity is compared with the fault diagnosis and classification threshold to obtain the diagnosis and classification results of the battery MSC and the low capacity fault.

[0012] Preferably, the process of calculating the characteristic data set based on the DBSCAN algorithm to obtain the faulty battery includes:

[0013] Count the number of points in the neighborhood of each data point;

[0014] Randomly select a core point and add it to the first cluster, and continue to add all the core points in the neighborhood of the core point to the cluster, repeating this process until all core points are included;

[0015] Select the core points that have not been added to the cluster, add them to the next cluster, and repeat the above steps until all core points are included;

[0016] The non-core points that are not added to any cluster are marked as noise points, and the faulty battery data is obtained.

[0017] Preferably, the expression of the first-order integer-order battery equivalent circuit model is:

[0018]

[0019] Among them, U k is the battery terminal voltage U t The discrete expression of I k is the battery terminal current I t Discrete expression of battery, OCV is the open circuit voltage of the battery, R0 is the DC internal resistance of the battery, R1, C1 are the polarization resistance and capacitance, and U1 is the polarization voltage on the RC pair.

[0020] Preferably, the expression for fitting the average OCV-SOC curve of the normal battery according to the Rint model and the average ohmic internal resistance is:

[0021] OCV=U t +I t R0;

[0022] Among them, U t is the open circuit voltage of the battery, I tis the battery terminal current.

[0023] Preferably, the process of fitting the average OCV-SOC curve of a normal battery according to the Rint model and the average ohmic internal resistance also includes: fitting the calculation results using a multi-order polynomial function, and judging the convergence of the fitting by the root mean square error, and screening the battery data that can be used.

[0024] Preferably, the relative SOC change value is calculated based on the average OCV-SOC curve, and the process of calculating the relative capacity based on the relative SOC change value includes:

[0025] Calculate the OCV of the faulty battery, and obtain the SOC change trend of the faulty battery in the normal battery SOC section by querying the OCV-SOC data table of the normal battery;

[0026] By calculating the variation range of the SOC of the faulty battery, its SOC variation is obtained, and by comparing the SOC variation of the faulty battery with that of a normal battery, the relative capacity value of the faulty battery is calculated.

[0027] Preferably, the process of comparing the relative capacity and the fault diagnosis and classification threshold comprises:

[0028] The charge capacity deviation is q1, the discharge capacity deviation is q2, and the charge and discharge capacity deviation is |q1-q2|=q3;

[0029] Taking q1 as the main criterion for judging low capacity, q3 as the main criterion for judging MSC, and q2 as the auxiliary criterion for judging MSC, the battery with only low capacity fault and the battery with both MSC fault and low capacity fault can be distinguished. The expression is:

[0030]

[0031] Among them, P1 and P2 are the thresholds for fault diagnosis and classification.

[0032] Preferably, the process of comparing the relative capacity with the fault diagnosis and classification threshold further includes: defining the charge and discharge capacity deviation caused by low capacity as q4=(q1+q2) / 2, the capacity deviation caused by MSC as q5=(q2-q1) / 2, and judging the degree of low capacity fault in the battery having both faults, the expression is:

[0033]

[0034] Compared with the prior art, the present invention has the following advantages and technical effects:

[0035] 1. Differentiate between micro short circuit faults and low capacity faults: Micro short circuit faults that are difficult to diagnose and may cause serious accidents in energy storage systems are diagnosed, and low capacity faults that are easily confused with them are effectively distinguished. This innovation not only improves the accuracy of fault diagnosis, but also significantly enhances the safety of energy storage equipment and avoids potential risks caused by misjudgment of fault types.

[0036] 2. Simplified and efficient unsupervised learning framework: A more simplified and efficient unsupervised learning framework is adopted. This framework only requires unified cleaning and processing of data, without complex labeling and feature extraction operations, which greatly reduces the technical complexity and cost of implementation.

[0037] 3. Model and data fusion method for effectively processing sparse and low-frequency data: The proposed model and data fusion-driven method can efficiently process sparse and low-frequency sampled data, and is suitable for common data types and characteristics in practical application scenarios. This method takes into account the battery mechanism and effectively completes the classification of battery faults, which not only improves the efficiency of data processing, but also ensures the accuracy of fault identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0039] Figure 1 A schematic diagram of a method flow of an embodiment of the present invention;

[0040] Figure 2 A schematic diagram of a first-order integer equivalent circuit model of a battery according to an embodiment of the present invention;

[0041] Figure 3 Schematic diagram of a battery equivalent internal resistance model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0042] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0043] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0044] Embodiment 1

[0045] like Figure 1As shown, this embodiment provides a method for diagnosing and classifying micro-short circuit and low capacity faults of energy storage lithium batteries driven by a sparse data model and data fusion, including:

[0046] Step 1: Obtain relevant data of the energy storage battery system based on the cloud platform or other hardware and software platforms, including sampling information and sampling time of each level of architecture;

[0047] Step 2: Perform unified data processing on the information obtained and filter out unusable data information;

[0048] Step 3: Merge the battery cell voltage data collected by all BMUs, extract relevant features, and perform feature analysis;

[0049] Step 4: Use the DBSCAN algorithm to perform data clustering analysis based on the selected features to identify batteries with low capacity faults and MSC faults;

[0050] Step 5: For the BMU with the faulty battery, data interpolation and data alignment are performed based on the battery cluster current and cluster SOC data collected by the BCU to which it belongs, to ensure that the voltage collected by the BMU has corresponding current information.

[0051] Step 6: Based on the first-order integer battery equivalent circuit model, the average ohmic internal resistance of the batteries in each BMU is estimated according to the average voltage and current of the normal batteries in the BMU;

[0052] Step 7: interpolate the single cell voltage information collected in the BMU according to the interpolated cluster current and cluster SOC information of the BCU, and fit the normal battery average OCV-SOC curve according to the Rint model;

[0053] Step 8: Fit a polynomial according to the normal battery OCV-SOC curve to establish a normal battery OCV-SOC data table;

[0054] Step 9: Calculate the relative SOC change value of the faulty battery based on the Rint model, the faulty battery voltage, the cluster current of the BCU where the faulty battery is located, and the normal battery OCV-SOC data table;

[0055] Step 10: Calculate the relative capacity and design the fault diagnosis and classification threshold according to the relative SOC change value;

[0056] Step 11: Output the fault diagnosis results.

[0057] Specifically:

[0058] 1. Data cleaning and extraction

[0059] Since the data protocols uploaded by different devices may be different, the valid data (such as battery voltage, current, etc.) in the transmission content usually accounts for only a small proportion. In addition, various sensors may be interfered by noise, resulting in abnormal values ​​or abnormal intervals in the data. Therefore, before officially running the algorithm framework, the uploaded battery data needs to be cleaned and extracted. This step aims to eliminate invalid or abnormal data, extract key parameters, ensure the reliability of the algorithm, and optimize the efficiency of computing and storage resources.

[0060] 2. Data alignment and merging

[0061] The battery management system for industrial and commercial energy storage equipment usually adopts a three-layer architecture to achieve hierarchical management and control of battery packs, clusters, and stacks. The lowest layer is the battery management unit (BMU), whose main function is to collect data such as the voltage and temperature of the single battery and execute the control strategy of the single battery. The middle layer is the battery cluster management unit (BCU), which is mainly responsible for collecting data such as the voltage, current, SOC, etc. of the battery cluster, while summarizing the information transmitted by the BMU, communicating with the top layer and executing cluster-level control instructions. The top layer is the battery array management unit (BAU), which is responsible for storing and displaying data uploaded by the BMU, and communicating and transmitting information with external devices.

[0062] Due to the limitation of information transmission and storage cost, the sampling density of BMU for single cell information is usually low, and BMU usually adopts polling mechanism when reporting information to BCU, that is, each BCU queries its subordinate BMU in turn at fixed time intervals, which may cause the uploaded battery data to have timestamp misalignment problem. In addition, in order to meet the higher voltage requirements of large-scale industrial and commercial storage equipment, the battery cells in the BMU and the BMU under the BCU are usually connected in series. The voltage and temperature data of the single cell are reported by the BMU, while the battery current and cluster-level average SOC data are stored in the BCU, which further aggravates the problem of data timestamp misalignment. Therefore, in order to meet the needs of fault diagnosis, these data need to be reasonably aligned and integrated. The specific alignment and integration principles will be explained in detail in the subsequent steps.

[0063] 3. Feature analysis and selection

[0064] During the discharge process, micro-short circuit (MSC) batteries and low-capacity batteries show similar characteristics, that is, the deviation of their voltage compared to normal batteries will gradually increase. During the charging process, due to the existence of the equipment balancing system, regardless of the type of faulty battery, the voltage difference between it and the normal battery will gradually decrease. In addition, MSC faulty batteries are usually accompanied by low-capacity fault characteristics. Based on the above characteristics, and taking into account the limitations of data transmission strategies and data quality, the present invention uses the original voltage data of the single cell uploaded by the BMU for fault identification, and selects the average value of the battery voltage and the difference between the single cell voltage and the average value as the main features.

[0065] Since lithium iron phosphate batteries inevitably have certain inconsistencies during use, if feature analysis is performed based only on the data of a single BMU, misdiagnosis of faults may occur due to minor inconsistencies. To this end, this solution integrates battery data from all dates for comprehensive analysis. By studying the battery voltage data curve, the voltage average value and the residual of the single cell voltage and the average value are determined as key features. Considering that the distinction between faulty batteries and normal batteries mainly depends on the residual features, in order to enhance the significance of the features, the residual features are multiplied by a factor of 2, thereby strengthening the clustering effect and reducing the sensitivity and complexity of the algorithm parameter adjustment.

[0066] 4. Fault identification based on clustering algorithm

[0067] Common supervised learning methods (such as neural networks or support vector machines) rely on accurately labeled training data, which often requires a lot of manpower and material resources in the early stages of data processing. The data labeling process not only includes data collection and cleaning, but also needs to ensure the accuracy and consistency of the labels, otherwise it may have a negative impact on model performance. In addition, supervised learning methods rely heavily on the quality and diversity of data labels. If there are errors in the labels or insufficient data diversity, the model may not perform well when processing unlabeled or unknown data. In contrast, unsupervised learning methods do not require prior labeling of data, but learn by automatically discovering patterns and structures in the data, making them more flexible and adaptable in processing large-scale and diverse data. Although the accuracy of unsupervised learning may be inferior to that of supervised learning in certain specific tasks, its advantages in exploratory data analysis, data clustering, and dimensionality reduction make it more suitable for the battery fault identification work in this article.

[0068] Data clustering is one of the core methods of unsupervised learning, which achieves classification by grouping disordered data into meaningful subclasses. Among many clustering algorithms, the DBSCAN algorithm has greater flexibility in processing unknown data sets because it does not need to pre-specify the number of clusters. In addition, the DBSCAN algorithm can effectively identify and process noise and outliers in the data, marking these points as outliers, so that it can still perform well in the case of irregular data distribution. However, the parameter selection of this algorithm is relatively complex and needs to be accurately set to ensure the best clustering effect.

[0069] The main input parameters of the DBSCAN algorithm include the radius of the cluster (Eps) and the minimum number of data points (MinPts). Based on these two parameters, the data points in the data set are divided into three categories: core points, boundary points, and noise points. A core point refers to a data point that contains at least MinPts points in its Eps neighborhood; a boundary point refers to a data point that has less than MinPts points in its Eps neighborhood but is located in the core point neighborhood; a noise point is neither a core point nor a boundary point. The main process of the algorithm is as follows:

[0070] (1) Counting the number of points in the neighborhood: Calculate the number of points in the neighborhood of Eps for each data point, and mark the points with a number greater than MinPts as core points. The neighborhood calculation is based on a specified distance metric. The present invention uses the Euclidean distance, and its formula is:

[0071]

[0072] Where x, y represent the feature vectors of the data points; n represents the dimension of the feature.

[0073] (2) Create clusters: randomly select a core point and add it to the first cluster, and then continue to add all the core points in the neighborhood of the core point to the cluster, and repeat this process until all the core points are included.

[0074] (3) Add boundary points: All non-core points (i.e., boundary points) within the neighborhood of the core point are added to the cluster.

[0075] (4) Processing the remaining core points: Select the core points that have not been added to the cluster, add them to the next cluster, and repeat the above steps until all core points are included.

[0076] (5) Identify noise points: mark non-core points that are not added to any cluster as noise points.

[0077] In the present invention, the Euclidean distance used by the DBSCAN algorithm can effectively measure the similarity between battery features. By reasonably setting the Eps and MinPts parameters, normal batteries will be classified into the same cluster, while faulty batteries will be identified as noise points due to the significant deviation of their features, thereby achieving accurate fault identification.

[0078] 5. Identification of battery ohmic internal resistance

[0079] Since the BMU usually adopts a polling mechanism when reporting information to the BCU, the sampling density of the single battery voltage reported by the BMU is low, while the sampling density of the cluster current and cluster SOC data reported by the BCU is high. In order to match the data with different sampling frequencies, the BCU data is first interpolated. Then, the corresponding cluster current data is matched based on the time point of the BMU data, and based on the first-order integer model (such as Figure 2 As shown) completes the estimation of the battery ohmic internal resistance R0.

[0080] Based on Kirchhoff's laws, the mathematical description of the first-order model is:

[0081]

[0082] Among them U k is the battery terminal voltage U t The discrete expression of I k is the battery terminal current I t Discrete expression of battery, OCV is the open circuit voltage of the battery, R0 is the DC internal resistance of the battery, R1, C1 are the polarization resistance and capacitance, and U1 is the polarization voltage on the RC pair.

[0083] Polarization coefficient Then there is U 1,k =α1U 1,k-1 +R1(1-α1)I k-1 , substitute U 1,k =OCV k -U k -R0I k have:

[0084] α1U 1,k-1 =OCV k -U k -R0I k -R1(1-α1)I k-1 ;

[0085] Type U 1,k =α1U 1,k-1 +R1(1-α1)I k-1 U 1,k Change to U 1,k-1 , then:

[0086] U 1,k-1 =α1U 1,k-2 +R1(1-α1)I k-2 ;

[0087] Further:

[0088]

[0089] α1U 1,k-1 =OCV k -U k -R0I k -R1(1-α1)I k-1 U 1,k-1 Change to U 1,k-2 Then we have:

[0090] α1U 1,k-2 =OCV k-1 -U k-1 -R0I k-1 -R1(1-α1)I k-2 ;

[0091] Further:

[0092] OCV k -U k -R0I k -R1(1-α1)I k-1 =α1[OCV k-1 -U k-1 -R0I k-1 -R1(1-α1)I k-2 ]+α1R1(1-α1)I k-2 ;

[0093] After simplification, we have:

[0094] U k =α1U k-1 +OCV k -α1OCV k-1 -R0I k +α1R0I k-1 -R1(1-α1)I k-1

[0095] In the above formula, U k Change to U k-1 , using the slow time-varying characteristics of OCV, that is, OCV k -OCV k-1 ≈0, after subtraction we have:

[0096] U k -U k-1 =α1(U k-1 -U k-2 )-R0(I k -I k-1 )-[R1(1-α1)-α1R0](I k-1 -I k-2 );

[0097] Remember k =U k -U k-1 , x k =I k -I k-1 , θ1=α1, θ2=-R0, θ3=-[R1(1-α1)-α1R0], then there is:

[0098]

[0099] Where the parameter vector θ = [θ1θ2θ3] T , k is the bounded noise and model error.

[0100] The parameter identification of the battery ohmic internal resistance is completed using the recursive least squares algorithm with forgetting factor (FFRLS):

[0101]

[0102] in

[0103] 6. Construction of battery OCV-SOC relationship table

[0104] After the ohmic internal resistance is estimated, in order to obtain a smoother and more accurate OCV-SOC curve, the interpolated BCU data is used to further perform secondary interpolation and alignment on the single battery voltage data of the BMU to ensure that the voltage data is synchronized with the cluster SOC and cluster current after the primary interpolation.

[0105] Then, the equivalent internal resistance (Rint) model is used to fit the OCV-SOC function of the normal battery based on the battery voltage data after secondary interpolation and the BCU cluster SOC and cluster current data after primary interpolation. Figure 3 As shown, the OCV of the Rint model is calculated by the following formula:

[0106] OCV=U t +I t R0;

[0107] According to the above formula, the OCV of the longest charging process in each time period is estimated and aligned with the SOC. In order to obtain a smooth and accurate OCV-SOC curve, a multi-order polynomial function is used to fit the calculation results, and the convergence of the fitting is judged by the root mean square error (RMSE), and the battery data that can be used is further screened.

[0108] Since the fitting function uses a multi-order polynomial, the computational complexity of directly reversing the SOC through the OCV of the faulty battery is high. In order to simplify the SOC estimation process of the faulty battery, this study established an OCV-SOC data table based on the fitting function, and used the OCV of the faulty battery to reversely query the SOC through a table lookup method to improve the efficiency of SOC estimation.

[0109] 7. Calculation of relative SOC and relative capacity of faulty battery

[0110] Fault identification has been completed based on the DBSCAN algorithm, but since the characteristics of MSC and low-capacity batteries are similar during discharge, the key to classification lies in analyzing the SOC changes and capacity calculation during charging. To this end, it is necessary to further pay attention to the SOC changes during charging and combine them with capacity calculation to distinguish the different characteristics of the two.

[0111] At the same time, considering that MSC batteries in real equipment are usually accompanied by low-capacity faults and are affected by short-circuit internal resistance, the discharge process of the faulty battery will be greatly affected. Therefore, in the process of fault identification, it is necessary to combine the SOC changes during the discharge process and the capacity calculation results for auxiliary judgment.

[0112] Therefore, according to OCV = U t +I t R0 calculates the OCV of the faulty battery, and obtains the SOC change trend of the faulty battery in the normal battery SOC section by querying the OCV-SOC data table of the normal battery. Further, by calculating the change interval of the faulty battery SOC, the SOC change is obtained, and by comparing the SOC change of the faulty battery with that of the normal battery, the relative capacity value of the faulty battery is calculated, thereby achieving more accurate fault diagnosis.

[0113] 8. Design and classification of fault thresholds

[0114] By calculating the average charge and discharge capacity and capacity deviation of each faulty battery, the fault classification threshold is designed. Let the charge capacity deviation be q1, the discharge capacity deviation be q2, and the charge and discharge capacity deviation be |q1-q2|=q3. Combined with the influence of the whole process interpolation processing and data accuracy, q1 is the main criterion for judging low capacity, q3 is the main criterion for judging MSC, and q2 is the auxiliary criterion for judging MSC. According to the following formula, batteries with only low capacity faults and batteries with both MSC faults and low capacity faults can be distinguished.

[0115]

[0116] Among them, P1 and P2 are thresholds for fault diagnosis and classification, which are set according to the specific situation of the battery. Usually, P1 can be set to 3% and P2 to 5%.

[0117] Furthermore, for a battery with two faults at the same time, the charge and discharge capacity deviation caused by low capacity is defined as q4 = (q1 + q2) / 2, and the capacity deviation caused by MSC is defined as q5 = (q2 - q1) / 2. According to the following formula, the degree of low capacity fault in a battery with two faults at the same time can be clearly determined.

[0118]

[0119] At this point, the diagnosis and classification of battery MSC and low capacity faults in the energy storage battery system have been completed.

[0120] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for diagnosing and classifying micro-short circuit and low-capacity faults of energy storage lithium batteries driven by a sparse data model and data fusion, characterized in that: include: Obtaining the operating data uploaded by the battery management system of the industrial and commercial energy storage equipment, performing data cleaning on the operating data, and obtaining cleaned data; Performing feature analysis and selection on the cleaned data to obtain a feature data set; Calculating the characteristic data set based on the DBSCAN algorithm to obtain the faulty battery, performing data interpolation and data alignment processing on the battery cluster current and cluster SOC data of the faulty battery to obtain the cluster current and cluster SOC information of the interpolated BCU; Calculate the average ohmic internal resistance of the faulty battery based on a first-order integer-order battery equivalent circuit model; Interpolating the single cell voltage information collected in the BMU based on the interpolated cluster current and cluster SOC information of the BCU, and fitting the normal battery average OCV-SOC curve according to the Rint model and the average ohmic internal resistance; Calculating a relative SOC change value based on the average OCV-SOC curve, calculating a relative capacity based on the relative SOC change value, and designing a fault diagnosis and classification threshold; The relative capacity is compared with the fault diagnosis and classification threshold to obtain the diagnosis and classification results of the battery MSC and the low capacity fault.

2. The method according to claim 1, characterized in that The process of calculating the characteristic data set based on the DBSCAN algorithm to obtain the faulty battery includes: Count the number of points in the neighborhood of each data point; Randomly select a core point and add it to the first cluster, and continue to add all the core points in the neighborhood of the core point to the cluster, repeating this process until all core points are included; Select the core points that have not been added to the cluster, add them to the next cluster, and repeat the above steps until all core points are included; The non-core points that are not added to any cluster are marked as noise points, and the faulty battery data is obtained.

3. The method according to claim 1, characterized in that The expression of the first-order integer battery equivalent circuit model is: Among them, U k is the battery terminal voltage U t The discrete expression of I k is the battery terminal current I t Discrete expression of battery, OCV is the open circuit voltage of the battery, R0 is the DC internal resistance of the battery, R1, C1 are the polarization resistance and capacitance, and U1 is the polarization voltage on the RC pair.

4. The method according to claim 1, characterized in that: The expression for fitting the normal battery average OCV-SOC curve according to the Rint model and the average ohmic internal resistance is: OCV=U t +I t R0; Among them, U t is the open circuit voltage of the battery, I t is the battery terminal current.

5. The method according to claim 1, characterized in that The process of fitting the normal battery average OCV-SOC curve according to the Rint model and the average ohmic internal resistance also includes: fitting the calculation results using a multi-order polynomial function, judging the convergence of the fitting by the root mean square error, and screening the usable battery data.

6. The method according to claim 1, characterized in that The relative SOC change value is calculated based on the average OCV-SOC curve, and the process of calculating the relative capacity based on the relative SOC change value includes: Calculate the OCV of the faulty battery, and obtain the SOC change trend of the faulty battery in the normal battery SOC section by querying the OCV-SOC data table of the normal battery; By calculating the variation range of the SOC of the faulty battery, its SOC variation is obtained, and by comparing the SOC variation of the faulty battery with that of a normal battery, the relative capacity value of the faulty battery is calculated.

7. The method according to claim 1, characterized in that The process of comparing the relative capacity and the fault diagnosis and classification threshold comprises: The charge capacity deviation is q1, the discharge capacity deviation is q2, and the charge and discharge capacity deviation is |q1-q2|=q3; Taking q1 as the main criterion for judging low capacity, q3 as the main criterion for judging MSC, and q2 as the auxiliary criterion for judging MSC, the battery with only low capacity fault and the battery with both MSC fault and low capacity fault can be distinguished. The expression is: Among them, P1 and P2 are the thresholds for fault diagnosis and classification.

8. The method according to claim 1, characterized in that The process of comparing the relative capacity with the fault diagnosis and classification threshold also includes: defining the charge and discharge capacity deviation caused by low capacity as q4=(q1+q2) / 2, the capacity deviation caused by MSC as q5=(q2-q1) / 2, and judging the degree of low capacity fault in the battery having two faults at the same time, the expression is:

Images