High-voltage box copper bar fault diagnosis method, device, equipment, storage medium and product

Abstract

This application provides a method, apparatus, device, storage medium, and product for diagnosing faults in high-voltage box copper busbars. The method includes: acquiring real-time operating data of the high-voltage box in an energy storage system; extracting features from the real-time operating data to generate real-time feature data for the current time window; and inputting the real-time feature data into a pre-trained fault diagnosis model to generate a fault diagnosis result for the high-voltage box copper busbars in the energy storage system. This method enables online diagnosis of faults in the high-voltage box copper busbars of the energy storage system, improving the real-time performance and accuracy of fault diagnosis.

Description

Technical Field

[0001] This application relates to the field of energy storage technology, and in particular to a method, apparatus, equipment, storage medium and product for diagnosing faults in high-voltage box copper busbars. Background Technology

[0002] The high-voltage box is one of the core components of an energy storage system, used to achieve high-voltage power distribution and transmission between battery clusters. The copper busbars inside the high-voltage box are the main conductive components, and their connection status directly affects the conductivity and safety stability of the energy storage system. In actual operation, the high-voltage box copper busbars may experience various malfunctions due to long-term vibration, thermal stress, oxidation corrosion, or installation defects, such as voltage imbalance between battery clusters and abnormal relay temperature rise, directly impacting the reliability and economy of the energy storage system.

[0003] In existing technologies, fault diagnosis methods based on fixed threshold rules rely on manual experience to set fixed thresholds for parameters such as the voltage difference between battery clusters and the temperature rise rate of relays. When real-time monitoring data exceeds the threshold, a fault is determined. This approach is difficult to adapt to complex operating conditions and suffers from diagnostic lag.

[0004] Therefore, there is an urgent need for a high-voltage box copper busbar fault diagnosis method that integrates multi-source data and balances diagnostic efficiency and model interpretability, so as to achieve accurate identification and rapid response to copper busbar faults. Summary of the Invention

[0005] This application provides a method, apparatus, equipment, storage medium, and product for diagnosing faults in high-voltage box copper busbars, enabling online diagnosis of faults in high-voltage box copper busbars of energy storage systems and improving the real-time performance and accuracy of fault diagnosis.

[0006] In a first aspect, embodiments of this application provide a method for diagnosing faults in high-voltage box copper busbars, including:

[0007] Acquire real-time operating data of the high-voltage box in the energy storage system;

[0008] Feature extraction is performed on the real-time running data to generate real-time feature data for the current time window;

[0009] The real-time feature data is input into a pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar of the energy storage system.

[0010] In one possible implementation, the real-time operating data includes: sampled voltage data of each battery cluster connection point within the high-voltage box and temperature data of the relays within the high-voltage box; the step of extracting features from the real-time operating data to generate real-time feature data for the current time window includes:

[0011] Preliminary calculations are performed based on the sampled voltage data and the relay temperature data to obtain derived data, which includes: the voltage difference between battery clusters and the temperature rise rate of the relay;

[0012] Feature extraction is performed on the pressure difference between the battery clusters and the temperature rise rate to obtain real-time feature data for the current time window.

[0013] In one possible implementation, the real-time feature data includes: time-domain statistical features, operating condition correlation features, and temperature rise trend features. The step of extracting features from the pressure difference and the temperature rise rate to obtain the real-time feature data for the current time window includes:

[0014] Based on the inter-cell pressure difference and the temperature rise rate, the statistical values ​​of the current time window are calculated to obtain the time-domain statistical characteristics of the inter-cell pressure difference and the temperature rise rate.

[0015] Calculate the correlation coefficient between the inter-cell pressure difference and the temperature rise rate to obtain the operating condition correlation characteristics;

[0016] Calculate the moving average of the temperature rise rate to obtain the temperature rise trend characteristics.

[0017] In one possible implementation, the fault diagnosis model includes: a fault detection model and a fault classification model; the step of inputting the real-time feature data into the pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar of the energy storage system includes:

[0018] The real-time feature data is input into the fault detection model to generate the fault detection results of the high-voltage box copper busbar of the energy storage system;

[0019] When the fault detection result indicates that there is no high-voltage box copper busbar fault in the energy storage system, the fault diagnosis result of the high-voltage box copper busbar of the energy storage system is determined to be normal.

[0020] When the fault detection result indicates that there is a high-voltage box copper busbar fault in the energy storage system, the real-time feature data is input into the fault classification model to generate the fault classification result of the high-voltage box copper busbar of the energy storage system, and the fault classification result is used as the fault diagnosis result of the high-voltage box copper busbar of the energy storage system.

[0021] In one possible implementation, the real-time feature data is input into the fault detection model to generate fault detection results for the high-voltage box copper busbar of the energy storage system, including:

[0022] The real-time feature data is input into the fault detection model, and the real-time feature data is mapped to anomaly values.

[0023] Determine whether the anomaly value is less than the decision threshold;

[0024] If so, then the fault detection result is determined to be that there is no high-voltage box copper busbar fault in the energy storage system;

[0025] If not, then the fault detection result indicates that there is a high-voltage box copper busbar fault in the energy storage system.

[0026] In one possible implementation, the fault classification model includes multiple predefined fault categories; the step of inputting the real-time feature data into the fault classification model to generate the fault classification result of the high-voltage box copper busbar of the energy storage system includes:

[0027] The real-time feature data is input into the fault classification model to obtain the probability distribution of the current fault belonging to each fault category;

[0028] The fault classification result of the current fault is determined based on the highest probability value in the probability distribution.

[0029] In one possible implementation, each fault category includes at least one of a fault cause and a fault level; determining the fault classification result of the current fault based on the highest probability value in the probability distribution includes:

[0030] Determine whether the highest probability value in the probability distribution is greater than the confidence threshold;

[0031] If so, then at least one of the fault types and fault levels included in the fault category corresponding to the highest probability value shall be used as the fault classification result of the current fault.

[0032] If not, then the fault classification result of the current fault is determined to be an unknown anomaly.

[0033] In one possible implementation, before inputting the real-time feature data into a pre-trained fault diagnosis model to generate fault diagnosis results for the energy storage system, the method further includes:

[0034] Obtain historical operating data and maintenance records of the high-voltage box in the energy storage system for each historical period;

[0035] Based on the operation and maintenance records, each of the historical operation data is labeled to obtain sample labels for each of the historical operation data.

[0036] Feature extraction is performed on each of the historical operational data to obtain historical feature data;

[0037] The fault diagnosis model is obtained by training the model based on the sample labels and historical feature data of each historical operation data.

[0038] Secondly, embodiments of this application provide a high-voltage box copper busbar fault diagnosis device, comprising:

[0039] The acquisition module is used to acquire real-time operating data of the high-voltage box in the energy storage system;

[0040] The extraction module is used to extract features from the real-time running data and generate real-time feature data for the current time window;

[0041] The diagnostic module is used to input the real-time feature data into a pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar of the energy storage system.

[0042] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0043] The memory stores computer-executed instructions;

[0044] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0045] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0046] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0047] The high-voltage box copper busbar fault diagnosis method, apparatus, equipment, storage medium, and product provided in this application acquire real-time operating data of the high-voltage box in the energy storage system, extract features from the real-time operating data to generate real-time feature data for the current time window, and input the real-time feature data into a pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar in the energy storage system. This achieves online fault diagnosis of the high-voltage box copper busbar in the energy storage system, solving the problems of poor adaptability and lag in existing fault diagnosis methods, and improving the real-time performance and accuracy of fault diagnosis. Attached Figure Description

[0048] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0049] Figure 1A schematic diagram illustrating a scenario for fault diagnosis of the copper busbar in a high-voltage box of an energy storage system, provided as an embodiment of this application;

[0050] Figure 2 A flowchart illustrating a high-voltage box copper busbar fault diagnosis method provided in this application embodiment. Figure 1 ;

[0051] Figure 3 A flowchart illustrating a high-voltage box copper busbar fault diagnosis method provided in this application embodiment. Figure 2 ;

[0052] Figure 4 A flowchart illustrating a high-voltage box copper busbar fault diagnosis method provided in this application embodiment. Figure 3 ;

[0053] Figure 5 A flowchart illustrating a high-voltage box copper busbar fault diagnosis method provided in this application embodiment. Figure 4 ;

[0054] Figure 6 This is a schematic diagram of a high-voltage box copper busbar fault diagnosis device provided in an embodiment of this application;

[0055] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0056] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0057] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0058] "Multiple" refers to two or more, and other quantifiers are similar. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following objects have an "or" relationship.

[0059] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, products, or apparatus.

[0060] It should be noted that, in the embodiments of this application, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0061] The high-voltage box is one of the core components of an energy storage system, used to achieve high-voltage power distribution and transmission between battery clusters. The copper busbars inside the high-voltage box are the main conductive components, and their connection status directly affects the conductivity and safety stability of the energy storage system.

[0062] In actual operation, the high-voltage box copper busbar may cause various faults due to long-term vibration, temperature stress, oxidation corrosion or installation defects, such as voltage imbalance between battery clusters and abnormal relay temperature rise, which directly affect the reliability and economy of the energy storage system.

[0063] Figure 1 This application provides a schematic diagram of a scenario for diagnosing a fault in the copper busbar of a high-voltage box in an energy storage system, as illustrated in the embodiments of this application. Figure 1 As shown, the high-voltage box in the energy storage system 10 is connected to each battery cluster 12 through copper busbar 11. During the operation of the energy storage system 10, the energy storage system 10 monitors the operating data of the high-voltage box in real time and sends the operating data to the electronic device 13 for copper busbar fault diagnosis. The electronic device 13 performs copper busbar fault diagnosis by setting fixed thresholds for parameters such as the voltage difference between battery clusters and the relay temperature rise rate based on manual experience. That is, when the real-time monitoring data exceeds the threshold, a fault is determined.

[0064] Based on the above scenarios, it can be seen that existing fault diagnosis methods based on fixed threshold rules are difficult to adapt to complex working conditions and suffer from diagnostic lag.

[0065] The high-voltage box copper busbar fault diagnosis method provided in this application realizes online diagnosis of high-voltage box copper busbar faults in energy storage systems by collecting real-time operating data of the high-voltage box in the energy storage system and combining it with a fault diagnosis model deployed in edge computing electronic devices. This solves the technical problems of poor adaptability and lag in existing fault diagnosis methods.

[0066] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0067] Figure 2 A flowchart illustrating a high-voltage box copper busbar fault diagnosis method provided in this application embodiment. Figure 1 The execution entity in this embodiment can be as follows: Figure 1 The electronic device 13 shown is, for example Figure 2 As shown, the method includes:

[0068] S201. Obtain real-time operating data of the high-voltage box in the energy storage system.

[0069] Among them, high-voltage box copper busbar fault diagnosis refers to fault diagnosis of copper busbar connection faults in the high-voltage box of the energy storage system. Fault diagnosis includes, but is not limited to, fault detection, determining the cause of the fault, and determining the fault level.

[0070] Real-time operating data refers to data used to reflect the operating status of the high-voltage box. Real-time operating data may include, for example, the output voltage of each battery cluster, the current relay temperature of each battery cluster, and the status of the relay contacts.

[0071] In this step, the acquisition module in the energy storage system acquires real-time operating data of the high-voltage box during the operation of the energy storage system at fixed time intervals.

[0072] The acquisition module can be a sensor system deployed within the energy storage system, including voltage sensors, current sensors, temperature sensors, etc. Alternatively, it can be a receiving device within an electronic device that receives real-time operating data from the high-voltage box, collected and transmitted in real-time by the energy storage system's operation monitoring system.

[0073] A fixed time interval can be, for example, 1 second, 5 seconds, etc.

[0074] For example, by using voltage sensors, current sensors, temperature sensors and other acquisition units deployed within the energy storage system, real-time operating data such as the output voltage of each battery cluster in the high-voltage box, the temperature of the current relay of each battery cluster, and the status of the relay contacts can be acquired at fixed time intervals during the operation of the energy storage system.

[0075] S202. Extract features from real-time running data to generate real-time feature data for the current time window.

[0076] Feature extraction refers to extracting characteristic data from real-time operational data for fault analysis. Feature extraction includes, but is not limited to, statistical calculations, time-domain analysis, and frequency-domain analysis.

[0077] The current time window refers to a time range determined according to a preset duration. For example, the current time window could be the past 5 minutes, the past 10 minutes, etc.

[0078] Real-time feature data refers to feature values ​​obtained by extracting features from real-time running data based on the current time window.

[0079] Real-time feature data can be, for example, statistical features, operating condition features, or trend features.

[0080] Statistical characteristics can include, for example, the mean, variance, and skewness. Operating condition characteristics can include, for example, relay action events and operating condition correlation characteristics, such as the correlation coefficient between the voltage difference between battery clusters and the rate of temperature rise. Trend characteristics can include, for example, the temperature rise trend.

[0081] In this step, real-time feature data for the current time window is generated by extracting features from the real-time running data of the current time window.

[0082] Specifically, by performing time-domain analysis, frequency-domain analysis, statistical calculations, and other feature extraction operations on the real-time operating data of the current time window, real-time feature data such as statistical features, operating condition features, and trend features of the current time window are obtained.

[0083] In existing technologies, fault diagnosis methods based on fixed thresholds cannot fully capture the multi-dimensional characteristics of complex faults such as loose copper busbar connections and oxidation corrosion. Therefore, copper busbar fault diagnosis has a one-sided nature. In one possible implementation, real-time operating data includes: sampled voltage data of each battery cluster connection point in the high-voltage box and temperature data of the relays in the high-voltage box. The process of extracting features from the above-mentioned real-time operating data to generate real-time feature data for the current time window is described in detail, including:

[0084] Preliminary calculations are performed based on the sampled voltage data and the relay temperature data to obtain derived data, including the inter-cell voltage difference and the relay temperature rise rate. Feature extraction is performed on the inter-cell voltage difference and temperature rise rate to obtain real-time feature data for the current time window.

[0085] For each battery cluster, the cluster voltage is determined based on the difference in sampled voltage data at the connection points at both ends of each cluster. The voltage difference between each cluster and its adjacent clusters is calculated to obtain the inter-cluster voltage difference. The temperature difference is determined by comparing the temperature value at each time point with the temperature values ​​at adjacent time points in the relay's temperature data. The ratio of the temperature difference to a fixed time interval is calculated to obtain the relay's temperature rise rate.

[0086] Feature extraction operations such as time domain analysis, frequency domain analysis, and statistical calculation are performed on the pressure difference and temperature rise rate between battery clusters to obtain real-time feature data of the current time window, including time domain statistical features, operating condition correlation features, and temperature rise trend features.

[0087] Time-domain statistical characteristics refer to the statistical characteristics of the pressure difference and temperature rise rate between battery clusters over time, such as the mean and variance.

[0088] Operating condition correlation features are used to reflect the correlation between inter-cell pressure difference and temperature rise rate, such as the correlation coefficient between inter-cell pressure difference and temperature rise rate.

[0089] Temperature rise trend characteristics refer to features used to describe the upward trend of temperature, such as moving averages of the rate of temperature rise.

[0090] In this embodiment, voltage data at each battery cluster connection point can be obtained using voltage sensors installed within the high-voltage box. Temperature data on the relay surface can be obtained using temperature sensors deployed on the relay surface.

[0091] By extracting features from the derived data, multi-dimensional real-time feature data is obtained, which enhances the ability to identify complex faults and improves the comprehensiveness of fault diagnosis.

[0092] Optionally, before extracting features from the pressure difference and temperature rise rate to obtain real-time feature data for the current time window, the pressure difference and temperature rise rate between battery clusters can be standardized first.

[0093] Standardization processing can include noise reduction and normalization.

[0094] Denoising is used to remove outliers from a dataset. For example, it can be median filtering, weighted average filtering, or a combination of median filtering and weighted average filtering.

[0095] Normalization is used to eliminate the dimensional differences between data of different dimensions. For example, it can be min-max normalization, which linearly maps the data to the [0, 1] interval, or Z-Score normalization, etc.

[0096] Specifically, denoising processing is performed on the inter-cluster pressure difference and temperature rise rate to remove outlier data. The denoised inter-cluster pressure difference and temperature rise rate are then normalized to eliminate dimensional differences between different dimensions, ensuring the data are within a uniform scale, which facilitates subsequent training and analysis of machine learning models. This denoising and normalization of the inter-cluster pressure difference and temperature rise rate provides high-quality data for subsequent feature extraction.

[0097] For example, the inter-cell pressure difference and temperature rise rate are sequentially processed using median filtering and weighted average filtering through sliding window technology to obtain the filtered inter-cell pressure difference and temperature rise rate. The filtered inter-cell pressure difference and temperature rise rate are then normalized to linearly map the data to the [0, 1] interval.

[0098] When a copper busbar fails, it may exhibit characteristics such as increased contact resistance leading to localized overheating and abnormal voltage drop. In one possible implementation, real-time characteristic data includes: time-domain statistical characteristics, operating condition correlation characteristics, and temperature rise trend characteristics. A detailed explanation is provided regarding the extraction of features from the pressure difference and temperature rise rate to obtain the real-time characteristic data for the current time window, including:

[0099] Based on the inter-cell pressure difference and temperature rise rate, the statistical values ​​of the current time window are calculated to obtain the time-domain statistical characteristics of the inter-cell pressure difference and temperature rise rate; the correlation coefficient between the inter-cell pressure difference and temperature rise rate is calculated to obtain the operating condition correlation characteristics; and the moving average of the temperature rise rate is calculated to obtain the temperature rise trend characteristics.

[0100] Among them, statistical values ​​can be, for example, the mean, variance, and skewness of the pressure difference between battery clusters within a fixed time window (such as 5 minutes); and the mean and maximum values ​​of the relay's temperature rise rate.

[0101] Under normal operating conditions, the correlation between the voltage difference between battery clusters and the rate of temperature rise is relatively weak; when the copper busbar connection is loose, the contact resistance increases, and the heat generated at this point will be more significant when current passes through, leading to a stronger positive correlation between the voltage difference and the rate of temperature rise.

[0102] Based on the inter-cell pressure difference and temperature rise rate, the mean, variance, and skewness of the inter-cell pressure difference and the mean and maximum values ​​of the relay temperature rise rate are calculated for the current time window, respectively, so as to obtain the time-domain statistical characteristics of the inter-cell pressure difference and temperature rise rate.

[0103] The correlation coefficient between the pressure difference between battery clusters and the temperature rise rate is calculated, and this correlation coefficient is used as the operating condition correlation characteristic. By calculating the average temperature rise rate of each fixed time window in the current time window, a moving average line of the temperature rise rate is determined, and the temperature rise trend characteristic is obtained.

[0104] For example, in the early stages of copper busbar oxidation, the pressure difference between battery clusters may not change significantly, but the moving average of the temperature rise rate has shown a slow upward trend. The fault diagnosis model uses this feature to provide early warning.

[0105] By fusing time-domain statistical features, operating condition correlation features, and temperature rise trend features, the ability to identify complex faults is enhanced, and the comprehensiveness of fault diagnosis is improved.

[0106] S203. Input the real-time feature data into the pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar of the energy storage system.

[0107] The fault diagnosis results are used to reflect the fault status of the high-voltage box copper busbar in the energy storage system. The fault diagnosis results may include fault type, fault cause, fault level, etc.

[0108] In existing technologies, fault diagnosis methods based on fixed thresholds cannot adapt to the complex operating conditions of energy storage systems and individual differences in equipment, which can easily lead to misjudgment or missed judgment.

[0109] By inputting real-time feature data into a pre-trained fault diagnosis model, the fault status of the high-voltage box copper busbar in the energy storage system can be determined, including whether there is a fault, the type of fault, the cause of the fault, and the fault level.

[0110] For example, real-time feature data is input into a pre-trained fault diagnosis model. The fault diagnosis model outputs the probability distribution of the high-voltage box copper busbar of the energy storage system belonging to multiple predefined fault categories. Based on the highest probability value in the probability distribution, the fault diagnosis result of the high-voltage box copper busbar of the energy storage system is determined, thereby determining whether there is a fault in the high-voltage box copper busbar of the energy storage system and including the fault type, fault cause, fault level, etc.

[0111] This application provides a high-voltage box copper busbar fault diagnosis method. By acquiring real-time operating data of the high-voltage box in an energy storage system, extracting features from the real-time operating data, generating real-time feature data for the current time window, and inputting the real-time feature data into a pre-trained fault diagnosis model, a fault diagnosis result for the high-voltage box copper busbar of the energy storage system is generated. This achieves online fault diagnosis of the high-voltage box copper busbar in the energy storage system, solving the problems of poor adaptability and lag in existing fault diagnosis methods, and improving the real-time performance and accuracy of fault diagnosis.

[0112] In actual high-voltage box copper busbar fault diagnosis, the fault diagnosis result may indicate the presence of a fault or the absence of a fault. Each fault diagnosis outputs a probability distribution of the high-voltage box copper busbar belonging to multiple predefined fault categories. This is especially problematic when the high-voltage box copper busbar does not exhibit a fault, as outputting the probability distribution wastes computational resources and reduces the real-time performance of the fault diagnosis.

[0113] Figure 3 A flowchart illustrating a high-voltage box copper busbar fault diagnosis method provided in this application embodiment. Figure 2 In this embodiment, the fault diagnosis model includes: a fault detection model and a fault classification model. Figure 2 Based on the examples, this paper provides a detailed explanation of how real-time feature data is input into a pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar of the energy storage system. Figure 3 As shown, the method includes:

[0114] S301. Input real-time feature data into the fault detection model to generate fault detection results for the high-voltage box copper busbar of the energy storage system.

[0115] Among them, the fault detection model is a binary classification model used to determine whether there is a fault in the copper busbar of the high-voltage box of the energy storage system. For example, it can be a support vector machine, a deep neural network, a random forest model, a gradient boosting tree, etc.

[0116] In this step, real-time feature data is used as input features and fed into the fault detection model. The fault detection model maps the input features to the corresponding categories, i.e., whether a fault exists or not, to obtain the fault detection results of the high-voltage box copper busbar of the energy storage system.

[0117] In one possible implementation, the specific process for generating fault detection results is described in detail, including:

[0118] Real-time feature data is input into the fault detection model, and the real-time feature data is mapped to an anomaly value. It is then determined whether the anomaly value is less than the decision threshold. If it is, the fault detection result is determined to be that there is no high-voltage box copper busbar fault in the energy storage system. If not, the fault detection result is determined to be that there is a high-voltage box copper busbar fault in the energy storage system.

[0119] The anomaly value is used to quantify the probability of a current fault state in the high-voltage copper busbar of the energy storage system. For example, the anomaly value ranges from 0 to 1, and the decision threshold can be, for example, 0.5, 0.6, etc.

[0120] Real-time features are input into the fault detection model, which maps the real-time feature data into anomaly values ​​and compares the anomaly values ​​with the decision threshold. If the anomaly value is less than the decision threshold, it is determined that there is no high-voltage box copper busbar fault in the energy storage system; if the anomaly value is greater than or equal to the decision threshold, it is determined that there is a high-voltage box copper busbar fault in the energy storage system.

[0121] By mapping real-time feature data to anomaly values ​​through a fault detection model, rapid fault diagnosis of the high-voltage box copper busbar in the energy storage system is achieved, reducing computational overhead.

[0122] S302. When the fault detection result indicates that there is no fault in the high-voltage box copper busbar in the energy storage system, the fault diagnosis result of the high-voltage box copper busbar in the energy storage system is determined to be normal.

[0123] When the fault detection result indicates that there is no fault in the high-voltage box copper busbar in the energy storage system, the fault diagnosis result of the high-voltage box copper busbar in the energy storage system is determined to be normal. At this time, the fault diagnosis for this real-time characteristic data ends.

[0124] By using a fault detection model with relatively low computational cost, rapid screening and real-time response to fault diagnosis of copper busbars in high-voltage boxes of energy storage systems were achieved.

[0125] S303. When the fault detection result indicates that there is a fault in the high-voltage box copper busbar in the energy storage system, the real-time feature data is input into the fault classification model to generate the fault classification result of the high-voltage box copper busbar in the energy storage system, and the fault classification result is used as the fault diagnosis result of the high-voltage box copper busbar in the energy storage system.

[0126] The fault classification model is a multi-classification model used to determine the fault category of the current fault in the high-voltage box copper busbar of the energy storage system. Examples include support vector machines, deep neural networks, random forest models, and gradient boosting trees.

[0127] When the fault detection result indicates that there is a fault in the high-voltage box copper busbar in the energy storage system, the real-time feature data is input into the fault classification model to start the fault classification model. The fault classification model outputs the fault classification result of the high-voltage box copper busbar in the energy storage system and uses the fault classification result as the fault diagnosis result of the high-voltage box copper busbar in the energy storage system.

[0128] By activating the fault classification model when a fault exists, a fine classification of the current fault is achieved, avoiding waste of resources.

[0129] For example, in the daily operation of an energy storage system, normal operating conditions account for up to 90%. At this time, the fault diagnosis model only needs to run the fault detection model to perform binary classification judgment, thereby improving the model response time. The fault classification model is only activated in the 10% of abnormal operating conditions to perform multi-classification analysis on the current fault, thus avoiding the waste of computing resources.

[0130] In one possible implementation, the fault classification model includes multiple predefined fault categories, and provides a detailed description of the generated fault classification results, including:

[0131] Real-time feature data is input into the fault classification model to obtain the probability distribution of the current fault belonging to each fault category; based on the highest probability value in the probability distribution, the fault classification result of the current fault is determined.

[0132] Real-time feature data is input into the fault classification model, which calculates the probability distribution of the real-time feature data belonging to each predefined fault category. The fault category corresponding to the maximum probability value in the probability distribution can be directly used as the fault diagnosis result of the high-voltage box copper busbar of the energy storage system; or the reliability of the fault classification result can be determined based on the maximum probability value first, thereby determining the fault classification result of the current fault.

[0133] When a fault occurs in the copper busbar of the high-voltage box in the energy storage system, a fault classification model is activated to determine the fault category of the current fault, thus achieving rapid identification of the fault category.

[0134] Furthermore, each fault category includes at least one of fault cause and fault level; the fault classification result, determined based on the highest probability value in the probability distribution, is described in detail, including:

[0135] Determine whether the highest probability value in the probability distribution is greater than the confidence threshold; if so, take at least one of the fault type and fault level included in the fault category corresponding to the highest probability value as the fault classification result of the current fault; if not, determine the fault classification result of the current fault as an unknown anomaly.

[0136] The confidence threshold is used to indicate the reliability of the current fault category. For example, the confidence threshold can be 0.7, 0.8, etc.

[0137] Multiple predefined fault categories may not exhaustively cover all fault categories. When the highest probability value is less than or equal to the confidence threshold, it indicates a low match between the real-time feature data and each predefined fault category. Therefore, the current fault classification result is determined to be an unknown anomaly. When the highest probability value is greater than the confidence threshold, it indicates a high match between the real-time feature data and each predefined fault category. Therefore, the current fault classification result is determined to be the corresponding fault category, such as loose connection - minor, loose connection - severe, oxidation corrosion, etc.

[0138] By comparing the highest probability value with the confidence threshold, the credibility of the fault category is determined, which improves the robustness and generalization ability of the fault classification model under complex working conditions.

[0139] This application provides a fault diagnosis method for high-voltage box copper busbars. By inputting real-time feature data into a fault detection model, it generates fault detection results for the high-voltage box copper busbars in an energy storage system. When the fault detection results indicate that there is no fault in the high-voltage box copper busbars in the energy storage system, the fault diagnosis result is determined to be normal. When the fault detection results indicate that there is a fault in the high-voltage box copper busbars in the energy storage system, the real-time feature data is input into a fault classification model to generate a fault classification result for the high-voltage box copper busbars in the energy storage system. This fault classification result is then used as the fault diagnosis result for the high-voltage box copper busbars in the energy storage system. By mapping real-time feature data to anomaly values ​​for fault detection through the fault detection model, rapid fault diagnosis of high-voltage box copper busbars in energy storage systems is achieved, reducing computational overhead. The fault classification model is only activated when a fault exists, achieving fine classification of the current fault, avoiding resource waste, and improving the real-time performance and accuracy of fault diagnosis.

[0140] For example, Figure 4 A flowchart illustrating a high-voltage box copper busbar fault diagnosis method provided in this application embodiment. Figure 3 ,like Figure 4 As shown, the method includes:

[0141] Data acquisition phase: Acquire real-time operating data of the high-voltage box in the energy storage system.

[0142] Edge processing stage: Features are extracted from real-time operating data to obtain real-time feature data, namely: time-domain statistical features, operating condition correlation features, and temperature rise trend features.

[0143] Real-time feature data is input into the fault detection model to calculate the anomaly value. The anomaly value is compared with the judgment anomaly value. If the anomaly value is less than the judgment threshold, it means that there is no fault in the high-voltage box copper busbar of the energy storage system, that is, the high-voltage box copper busbar is normal. If the anomaly value is greater than or equal to the judgment threshold, it means that there is a fault in the high-voltage box copper busbar of the energy storage system, and the fault classification model is triggered.

[0144] Real-time feature data is input into the fault classification model to calculate the probability distribution of the current fault belonging to each fault category. It is then determined whether the maximum probability value in the probability distribution is greater than the confidence threshold. If it is greater, the fault category corresponding to the maximum probability value is determined as the fault classification result of the current fault; if it is less, it indicates that it is an unknown anomaly.

[0145] Optionally, a cloud-based analytics phase can also be included. This phase stores each fault diagnosis result and fault maintenance record to update the model parameters of the fault detection and fault classification models.

[0146] Figure 5 A flowchart illustrating a high-voltage box copper busbar fault diagnosis method provided in this application embodiment. Figure 4 In this embodiment Figure 2 or Figure 3 Based on the examples, the training process of the fault diagnosis model is described in detail, such as... Figure 5 As shown, the method includes:

[0147] S501. Obtain historical operating data and maintenance records of the high-voltage box in the energy storage system for each historical period.

[0148] In this step, historical operating data of the energy storage system under known conditions is acquired for each historical period.

[0149] Known states include: normal state, different fault types, and different severity levels.

[0150] S502. Based on the operation and maintenance records, each historical operation data is labeled to obtain sample labels for each historical operation data.

[0151] Based on the operation and maintenance records, each segment of historical operation data is labeled to obtain sample labels for each historical operation data.

[0152] Sample labels can be, for example, 0 (normal) or 1 (abnormal); they can also be specific fault categories, such as 10 (loose connection - minor), 11 (loose connection - severe), 20 (oxidation and corrosion), etc.

[0153] S503. Extract features from each historical operation data to obtain historical feature data.

[0154] Step S503 is similar to step S102, and will not be described again here.

[0155] S504. Based on the sample labels and historical feature data of each historical operation data, the model is trained to obtain the fault diagnosis model.

[0156] Based on the sample labels and historical feature data of each historical running data, the sample labels and historical feature data corresponding to each historical running data, i.e., the sample dataset, can be divided into training set and test set for model training and validation.

[0157] For example, 70% of the sample dataset can be divided into a training set for model training, and 30% of the sample dataset can be divided into a test set for model validation.

[0158] In one possible implementation, the fault diagnosis model includes a fault detection model and a fault classification model. Correspondingly, based on operation and maintenance records, each segment of historical operational data is labeled, including:

[0159] Each historical operational data point is labeled based on the first label of the fault detection model and the second label of the fault classification model to obtain sample labels.

[0160] The fault diagnosis model is trained using sample labels and historical feature data from various historical operational data. A detailed explanation follows, including:

[0161] The fault detection model is trained based on the training set in the sample dataset, with normal / abnormal as the target variable.

[0162] The fault classification model is trained based on the training set in the sample dataset, using each predefined fault category as the target variable.

[0163] For example: Based on the label 0 (normal) or 1 (abnormal) of the fault detection model, the historical operating data is labeled, and based on the label of the fault classification model, that is, the specific fault category, such as 10 (loose connection - minor), 11 (loose connection - severe), 20 (oxidation corrosion), etc., the historical operating data is labeled to obtain the sample label of each historical operating data.

[0164] 1) Training the fault detection model:

[0165] Using the training set, historical feature data are used as input feature vectors, and 0 / 1 labels are used as target variables. The random forest algorithm is used for training.

[0166] Key hyperparameters of the model (such as the number of decision trees n_estimators, the maximum depth of the trees max_depth, the minimum number of samples required for internal node re-splitting min_samples_split, etc.) can be optimized through grid search or other methods to achieve the expected accuracy and recall on the test set.

[0167] After training, a binary random forest model that can output the probability of "normal" or "abnormal" is obtained, namely the fault detection model M1, and saved.

[0168] 2) Training the fault classification model:

[0169] The training set for the second-level model is obtained by filtering out all historical running data with anomaly labels from the sample training set.

[0170] The historical feature data in the training set of the second-level model is used as input, and the specific fault category label is used as the target variable.

[0171] The random forest algorithm was also used, and hyperparameters were optimized to train a multi-classification model, namely the fault classification model M2, and then saved.

[0172] Preferably, when using historical feature data from the training set of the second-level model as input, the weights of the operating condition correlation features and the temperature rise trend features can be increased, or the weights of the time-domain statistical features can be reduced, so that the operating condition correlation features and the temperature rise trend features have a larger numerical range when input into the model, thereby enhancing the importance of the operating condition correlation features and the temperature rise trend features.

[0173] This application provides a high-voltage box copper busbar fault diagnosis method. It acquires historical operating data and maintenance records of the high-voltage box in the energy storage system for various historical periods. Based on the maintenance records, it labels each historical operating data point to obtain sample labels. Features are extracted from each historical operating data point to obtain historical feature data. A fault diagnosis model is then trained based on the sample labels and historical feature data. This method enables online fault diagnosis of the high-voltage box copper busbar in the energy storage system, solving the problems of poor adaptability and lag in existing fault diagnosis methods, and improving the real-time performance and accuracy of fault diagnosis.

[0174] Figure 6 This is a schematic diagram of a high-voltage box copper busbar fault diagnosis device provided in an embodiment of this application, as shown below. Figure 6 As shown, the high-voltage box copper busbar fault diagnosis device 60 provided in this embodiment includes: an acquisition module 601, an extraction module 602, and a diagnosis module 603.

[0175] The acquisition module 601 is used to acquire real-time operating data of the high-voltage box in the energy storage system;

[0176] The extraction module 602 is used to extract features from real-time running data and generate real-time feature data for the current time window;

[0177] The diagnostic module 603 is used to input real-time feature data into a pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar of the energy storage system.

[0178] In one possible implementation, the real-time operating data includes: sampled voltage data of each battery cluster connection point in the high-voltage box and temperature data of the relays in the high-voltage box. The extraction module 602 is also used to perform preliminary calculations based on the sampled voltage data and the relay temperature data to obtain derived data, which includes: the voltage difference between battery clusters and the temperature rise rate of the relays. The voltage difference between battery clusters and the temperature rise rate are used to extract features to obtain real-time feature data for the current time window.

[0179] In one possible implementation, the real-time feature data includes: time-domain statistical features, operating condition correlation features, and temperature rise trend features. The extraction module 602 is also used to calculate the statistical values ​​of the current time window based on the inter-cell pressure difference and temperature rise rate, respectively, to obtain the time-domain statistical features of the inter-cell pressure difference and temperature rise rate; calculate the correlation coefficient between the inter-cell pressure difference and temperature rise rate to obtain the operating condition correlation features; and calculate the moving average of the temperature rise rate to obtain the temperature rise trend features.

[0180] In one possible implementation, the fault diagnosis model includes a fault detection model and a fault classification model. The diagnosis module 603 is further used to input real-time feature data into the fault detection model to generate a fault detection result for the high-voltage box copper busbar of the energy storage system; when the fault detection result indicates that there is no fault in the high-voltage box copper busbar of the energy storage system, the fault diagnosis result of the high-voltage box copper busbar of the energy storage system is determined to be normal; when the fault detection result indicates that there is a fault in the high-voltage box copper busbar of the energy storage system, the real-time feature data is input into the fault classification model to generate a fault classification result for the high-voltage box copper busbar of the energy storage system, and the fault classification result is used as the fault diagnosis result of the high-voltage box copper busbar of the energy storage system.

[0181] In one possible implementation, the diagnostic module 603 is further configured to input real-time feature data into the fault detection model, map the real-time feature data to anomaly values, determine whether the anomaly value is less than a decision threshold, and if so, determine that the fault detection result is that there is no high-voltage box copper busbar fault in the energy storage system; if not, determine that there is a high-voltage box copper busbar fault in the energy storage system.

[0182] In one possible implementation, the fault classification model includes multiple predefined fault categories; the diagnostic module 603 is also used to input real-time feature data into the fault classification model to obtain the probability distribution of the current fault belonging to each fault category; and to determine the fault classification result of the current fault based on the highest probability value in the probability distribution.

[0183] In one possible implementation, each fault category includes at least one of fault cause and fault level. The diagnostic module 603 is further used to determine whether the highest probability value in the probability distribution is greater than the confidence threshold. If so, at least one of the fault type and fault level included in the fault category corresponding to the highest probability value is used as the fault classification result of the current fault. If not, the fault classification result of the current fault is determined to be an unknown anomaly.

[0184] In one possible implementation, the high-voltage box copper busbar fault diagnosis device 60 further includes a training module 604;

[0185] The acquisition module 601 is also used to acquire historical operating data and operation and maintenance records of the high-voltage box in the energy storage system during various historical periods; based on the operation and maintenance records, each historical operating data is labeled to obtain sample labels for each historical operating data.

[0186] The extraction module 602 is also used to extract features from each historical running data to obtain historical feature data;

[0187] The training module 604 is also used to train the model based on the sample labels and historical feature data of each historical operation data to obtain the fault diagnosis model.

[0188] The high-voltage box copper busbar fault diagnosis device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0189] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 7 As shown, the electronic device 70 provided in this embodiment includes at least one processor 701 and a memory 702. Optionally, the device 70 further includes a communication component 703. The processor 701, memory 702, and communication component 703 are connected via a bus 704.

[0190] In a specific implementation, at least one processor 701 executes computer execution instructions stored in memory 702, causing at least one processor 701 to perform the above-described method.

[0191] The specific implementation process of processor 701 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0192] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0193] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0194] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0195] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0196] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0197] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0198] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0199] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0200] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0201] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0202] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0203] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0204] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for diagnosing faults in high-voltage box copper busbars, characterized in that, include: Acquire real-time operating data of the high-voltage box in the energy storage system; Feature extraction is performed on the real-time running data to generate real-time feature data for the current time window; The real-time feature data is input into a pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar of the energy storage system.

2. The method according to claim 1, characterized in that, The real-time operating data includes: sampled voltage data of each battery cluster connection point within the high-voltage box and temperature data of the relays within the high-voltage box; the step of extracting features from the real-time operating data to generate real-time feature data for the current time window includes: Preliminary calculations are performed based on the sampled voltage data and the relay temperature data to obtain derived data, which includes: the voltage difference between battery clusters and the temperature rise rate of the relay; Feature extraction is performed on the pressure difference between the battery clusters and the temperature rise rate to obtain real-time feature data for the current time window.

3. The method according to claim 2, characterized in that, The real-time feature data includes: time-domain statistical features, operating condition correlation features, and temperature rise trend features. The feature extraction of the pressure difference and the temperature rise rate to obtain the real-time feature data for the current time window includes: Based on the inter-cell pressure difference and the temperature rise rate, the statistical values ​​of the current time window are calculated to obtain the time-domain statistical characteristics of the inter-cell pressure difference and the temperature rise rate. Calculate the correlation coefficient between the inter-cell pressure difference and the temperature rise rate to obtain the operating condition correlation characteristics; Calculate the moving average of the temperature rise rate to obtain the temperature rise trend characteristics.

4. The method according to claim 1, characterized in that, The fault diagnosis model includes a fault detection model and a fault classification model; the step of inputting the real-time feature data into the pre-trained fault diagnosis model to generate the fault diagnosis results of the high-voltage box copper busbar of the energy storage system includes: The real-time feature data is input into the fault detection model to generate the fault detection results of the high-voltage box copper busbar of the energy storage system; When the fault detection result indicates that there is no fault in the high-voltage box copper busbar in the energy storage system, the fault diagnosis result of the high-voltage box copper busbar in the energy storage system is determined to be normal. When the fault detection result indicates that there is a high-voltage box copper busbar fault in the energy storage system, the real-time feature data is input into the fault classification model to generate the fault classification result of the high-voltage box copper busbar of the energy storage system, and the fault classification result is used as the fault diagnosis result of the high-voltage box copper busbar of the energy storage system.

5. The method according to claim 4, characterized in that, The step of inputting the real-time feature data into the fault detection model to generate the fault detection result of the high-voltage box copper busbar of the energy storage system includes: The real-time feature data is input into the fault detection model, and the real-time feature data is mapped to anomaly values. Determine whether the anomaly value is less than the decision threshold; If so, then the fault detection result is determined to be that there is no high-voltage box copper busbar fault in the energy storage system; If not, then the fault detection result indicates that there is a high-voltage box copper busbar fault in the energy storage system.

6. The method according to claim 4, characterized in that, The fault classification model includes multiple predefined fault categories; the step of inputting the real-time feature data into the fault classification model to generate fault classification results for the high-voltage copper busbar of the energy storage system includes: The real-time feature data is input into the fault classification model to obtain the probability distribution of the current fault belonging to each fault category; The fault classification result of the current fault is determined based on the highest probability value in the probability distribution.

7. The method according to claim 6, characterized in that, Each fault category includes at least one of fault cause and fault level; determining the fault classification result of the current fault based on the highest probability value in the probability distribution includes: Determine whether the highest probability value in the probability distribution is greater than the confidence threshold; If so, then at least one of the fault types and fault levels included in the fault category corresponding to the highest probability value shall be used as the fault classification result of the current fault. If not, then the fault classification result of the current fault is determined to be an unknown anomaly.

8. The method according to any one of claims 1-7, characterized in that, Before inputting the real-time feature data into the pre-trained fault diagnosis model to generate the fault diagnosis result of the energy storage system, the method further includes: Obtain historical operating data and maintenance records of the high-voltage box in the energy storage system for each historical period; Based on the operation and maintenance records, each of the historical operation data is labeled to obtain sample labels for each of the historical operation data. Feature extraction is performed on each of the historical operational data to obtain historical feature data; The fault diagnosis model is obtained by training the model based on the sample labels and historical feature data of each historical operation data.

9. A fault diagnosis device for high-voltage box copper busbars, characterized in that, include: The acquisition module is used to acquire real-time operating data of the high-voltage box in the energy storage system; The extraction module is used to extract features from the real-time running data and generate real-time feature data for the current time window; The diagnostic module is used to input the real-time feature data into a pre-trained fault diagnosis model to generate fault diagnosis results for the high-voltage box copper busbar of the energy storage system.

10. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-8.

12. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1-8.

Images