An abnormality diagnosis and repair method and system of an energy storage device and a storage medium

Abstract

The application relates to the technical field of anomaly detection, and discloses an abnormality diagnosis and repair method and system of a storage energy device and a storage medium, the method comprising the following steps: performing abnormal mode identification on real-time operation data to obtain initial abnormal event information; reconstructing the working conditions of a local area to obtain a local state parameter set; performing strategy optimization on a preset repair strategy mapping library to obtain a targeted repair scheme; capturing feedback state data of the targeted repair scheme in real time, evaluating repair expectation of the feedback state data, and obtaining a repair action execution record; synchronously rechecking a full state parameter of a target storage energy device to obtain a full state verification comparison result; archiving and sealing the repair action execution record and the full state verification comparison result to obtain a repair success record; and the application can improve the efficiency of abnormality diagnosis and repair of the storage energy device.

Description

Technical Field

[0001] This invention relates to the field of anomaly detection technology, and in particular to a method, system and storage medium for anomaly diagnosis and repair of energy storage devices. Background Technology

[0002] The existing anomaly diagnosis process for energy storage equipment only performs basic screening of the overall real-time operating data of the equipment. It does not conduct fine decomposition of the data feature parameters, nor does it perform accurate feature space mapping between the operating data and the anomaly features. At the same time, it lacks the reverse tracing and spatial coordinate registration process for the anomaly initiation time. This results in a discrepancy between the anomaly type identifier and the location coordinates of the anomaly occurrence in the initial anomaly event information, which cannot provide accurate and effective data support for subsequent repair work, directly leading to insufficient accuracy in anomaly diagnosis.

[0003] The repair work of existing energy storage equipment lacks a systematic process design. The formulation of repair strategies does not take into account the coupling relationship between anomaly type and local state parameters. Most of them adopt general repair solutions, which have poor targeted repair effects on equipment anomalies. Moreover, after repair, the equipment is not simultaneously re-inspected for all state parameters. The repair results are judged only by simply looking at the state data, which cannot verify whether the equipment has truly returned to normal at the feature space level. At the same time, there is a lack of standardized archiving and preservation of the repair process and results, which is prone to incomplete repair and is not conducive to solving similar anomalies in the future. The overall efficiency of anomaly diagnosis and repair is low. Summary of the Invention

[0004] This invention discloses a method, system, and storage medium for diagnosing and repairing anomalies in energy storage devices.

[0005] In a first aspect, the present invention discloses a method for anomaly diagnosis and repair of energy storage devices, comprising: Pt.1, performing anomaly pattern recognition on real-time operating data of a target energy storage device to obtain initial anomaly event information of the target energy storage device; Pt.2, based on the anomaly occurrence location coordinates in the initial anomaly event information, reconstructing the operating conditions of a local area of ​​the target energy storage device to obtain a local state parameter set of the target energy storage device; Pt.3, based on the coupling relationship between the anomaly type identifier in the initial anomaly event information and the local state parameter set, performing strategy optimization on a preset repair strategy mapping library to obtain a targeted repair scheme for the target energy storage device; Pt. 4. Real-time capture of the feedback status data of the targeted repair scheme, and assessment of the repair expectation of the feedback status data to obtain the repair action execution record of the target energy storage device; Pt.5. Based on the completion status identifier in the repair action execution record, perform a full-state parameter synchronous re-examination of the target energy storage device to obtain the full-state verification comparison result of the target energy storage device; Pt.6. When the full-state verification comparison result is that the full amount of operating data does not match any abnormal feature in the preset abnormal feature library in the feature space, archive and seal the repair action execution record and the full-state verification comparison result to obtain the repair success record of the target energy storage device.

[0006] In a preferred embodiment, the step of identifying abnormal patterns in the real-time operating data of the target energy storage device to obtain the initial abnormal event information of the target energy storage device includes: deconstructing the real-time operating data of the target energy storage device into feature parameters to obtain the real-time operating feature vector of the target energy storage device; mapping the real-time operating feature vector to an abnormal feature template in a preset abnormal feature library to obtain the abnormal type identification result of the target energy storage device; and based on the abnormal type identification result, tracing back the real-time operating data and performing spatial coordinate registration based on the obtained abnormal start time and the location label in the real-time operating data to obtain the initial abnormal event information of the target energy storage device.

[0007] In a preferred embodiment, the step of reconstructing the operating conditions of a local area of ​​the target energy storage device based on the anomaly location coordinates in the initial anomaly event information to obtain a local state parameter set of the target energy storage device includes: delineating the spatial topology of the target energy storage device based on the anomaly location coordinates in the initial anomaly event information to obtain a local detection area range of the target energy storage device; collecting operating condition parameters from the sensor data set of the target energy storage device based on the local detection area range to obtain a local state parameter data package of the target energy storage device; performing inter-parameter correlation verification on the state parameters in the local state parameter data package and performing constant area boundary verification on the local state parameter data package; when the state parameters pass the consistency check in the inter-parameter correlation verification and match the spatial positional relationship with the anomaly location coordinates in the anomaly area boundary verification, the local state parameter data package is marked as valid, and a local state parameter record of the target energy storage device is obtained.

[0008] In a preferred embodiment, the step of optimizing a preset repair strategy mapping library based on the coupling relationship between the anomaly type identifier in the initial anomaly event information and the local state parameter set to obtain a targeted repair scheme for the target energy storage device includes: assigning contribution weights to the state parameters based on the response sensitivity coefficient between the state parameters and the anomaly type identifier in the local state parameter record to obtain a parameter coupling feature vector of the target energy storage device; measuring the similarity between the parameter coupling feature vector and the trigger feature vector of the repair strategy entry in the preset repair strategy mapping library to obtain a candidate repair strategy list for the target energy storage device; projecting the candidate repair strategy list into the feature space based on the contribution weighting result in the parameter coupling feature vector to obtain the optimal repair strategy identifier for the target energy storage device; deconstructing the optimal repair strategy identifier to obtain a repair action sequence for the target energy storage device; and performing action sequence inversion on the repair actions in the repair action sequence to obtain a targeted repair scheme for the target energy storage device.

[0009] In a preferred embodiment, the real-time capture of feedback status data of the targeted repair scheme and the assessment of repair expectations based on the feedback status data to obtain the repair action execution record of the target energy storage device include: synchronously capturing the feedback status data of the targeted repair scheme to obtain a feedback status data stream of the target energy storage device; extracting waveform features from the feedback status data packets in the feedback status data stream to obtain a repair action feedback feature vector of the target energy storage device; quantifying the deviation between the repair action feedback feature vector and the preset expected status feature template in the feature space based on the spatial distance between the repair action feedback feature vector and the expected status feature template to obtain a comprehensive deviation quantization value and an execution status identifier of the target energy storage device; and archiving the comprehensive deviation quantization value and the execution status identifier in a log to obtain the repair action execution record of the target energy storage device.

[0010] In a preferred embodiment, the formula for calculating the comprehensive deviation quantification value is as follows: ;

[0011] In the formula, This is the quantified value of the overall deviation. The voltage recovery feature value is located in the feedback feature vector of the repair action. The voltage expected feature value in the expected state feature template, The current stability feature value in the feedback feature vector of the repair action is... The desired current characteristic value in the desired state characteristic template. The temperature convergence feature value in the feedback feature vector of the repair action is... The desired temperature feature value in the desired state feature template. The preset voltage deviation weighting coefficient, The preset current deviation weighting coefficient, The preset temperature deviation weighting coefficient, This is a preset stability constant.

[0012] In a preferred embodiment, the step of performing a full-state parameter synchronous re-examination of the target energy storage device based on the completion status identifier in the repair action execution record to obtain the full-state verification comparison result of the target energy storage device includes: when the execution status identifier of the repair action execution record indicates that the repair action has been completed, triggering the full-state detection channel of the target energy storage device to obtain a full-state detection trigger confirmation signal of the target energy storage device; based on the full-state detection trigger confirmation signal, synchronously acquiring the operating parameters of the target energy storage device to obtain a full-scale operating data package of the target energy storage device; performing time-frequency domain feature deconstruction on each operating parameter in the full-scale operating data package to obtain a full-state feature vector of the target energy storage device; measuring the similarity between the full-state feature vector and the abnormal feature template in the preset abnormal feature library to obtain a matching relationship list of the target energy storage device; and encapsulating the verification conclusion of the matching relationship list to obtain the full-state verification comparison result of the target energy storage device.

[0013] In a preferred embodiment, when the full-state verification comparison result indicates that the full operational data does not match any abnormal feature in the preset abnormal feature library in the feature space, the repair action execution record and the full-state verification comparison result are archived and sealed to obtain a successful repair record of the target energy storage device. This includes: analyzing the conclusion of the full-state verification comparison result; when the abnormal feature matching judgment conclusion in the full-state verification comparison result indicates that the full operational data does not match any abnormal feature in the abnormal feature library, a successful repair archiving trigger signal for the target energy storage device is obtained; based on the successful repair archiving trigger signal, the repair action execution record and the full-state verification comparison result are uniquely bound to obtain a repair result associated data packet for the target energy storage device; the storage location of the repair result associated data packet is mapped to obtain a repair archive storage confirmation record for the target energy storage device; and key summary extraction is performed on the archive storage path index and metadata identifier in the repair archive storage confirmation record to obtain a successful repair record for the target energy storage device.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention performs feature parameter deconstruction and feature space mapping analysis on the real-time operation data of energy storage equipment, and obtains accurate initial abnormal event information by combining reverse tracing and spatial coordinate registration. At the same time, it completes the working condition reconstruction of the local area based on the coordinates of the abnormality occurrence location, and obtains an effective local state parameter set through parameter correlation verification and abnormal area boundary verification, which greatly improves the accuracy of abnormal diagnosis of energy storage equipment, provides comprehensive and reliable state data support for subsequent repair work, and enables the results of abnormal diagnosis to accurately point to the core problem and specific location of the equipment abnormality.

[0015] 2. This invention achieves precise optimization of repair strategies based on the coupled correlation between anomaly types and local state parameters, formulating targeted repair schemes adapted to the actual anomaly conditions of the equipment. Simultaneously, it captures feedback status data during the repair process in real time and assesses repair expectations, accurately controlling the execution status of repair actions. Furthermore, through synchronous re-examination of all state parameters, it comprehensively verifies the equipment repair effect at the feature space level, ensuring the effectiveness of the repair work. Moreover, it standardizes and archives repair records and verification results, forming traceable repair success records, significantly improving the overall efficiency of energy storage equipment anomaly repair and accumulating standardized and effective data for subsequent equipment anomaly management. Attached Figure Description

[0016] Figure 1 The flowchart of the anomaly diagnosis and repair method for energy storage devices according to Embodiment 1 of the present invention is shown; Figure 2 The diagram shows the functional modules of the energy storage device anomaly diagnosis and repair system according to Embodiment 2 of the present invention; Figure 3 This diagram shows the changes in the comprehensive deviation quantification value before and after repair for different anomaly types in the anomaly diagnosis and repair method of the energy storage device according to Embodiment 1 of the present invention; Figure 4 This diagram shows the change in the comprehensive deviation quantification value during the repair process of the energy storage device according to the anomaly diagnosis and repair method of the energy storage device in Embodiment 1 of the present invention. Figure 5 This diagram shows the variation of the feature space matching ratio in the process stage of the anomaly diagnosis and repair method for energy storage devices according to Embodiment 1 of the present invention; Figure 6 This demonstrates the influence of the stability constant of the anomaly diagnosis and repair method for energy storage devices according to Embodiment 1 of the present invention on the comprehensive deviation quantification value. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions of this disclosure, and to fully understand and implement the process of how this disclosure applies technical means to solve technical problems and achieve corresponding technical effects, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, not all embodiments. The embodiments of this disclosure and the various features within them can be combined with each other without conflict, and the resulting technical solutions are all within the protection scope of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without creative effort should fall within the protection scope of this disclosure.

[0018] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0019] Example 1

[0020] Figure 1 This is a flowchart illustrating a method for diagnosing and repairing anomalies in an energy storage device, as provided in an embodiment of this disclosure. Figure 1 As shown, a method for anomaly diagnosis and repair of an energy storage device includes: Pt.1, performing anomaly pattern recognition on the real-time operating data of the target energy storage device to obtain initial anomaly event information of the target energy storage device; In this embodiment of the invention, the step of performing anomaly pattern recognition on the real-time operating data of the target energy storage device to obtain initial anomaly event information of the target energy storage device includes: performing feature parameter deconstruction on the real-time operating data of the target energy storage device to obtain the real-time operating feature vector of the target energy storage device; performing feature space mapping on the real-time operating feature vector and the anomaly feature template in the preset anomaly feature library to obtain the anomaly type identification result of the target energy storage device; based on the anomaly type identification result, performing reverse tracing on the real-time operating data, and performing spatial coordinate registration based on the obtained anomaly start time and the location label in the real-time operating data to obtain the initial anomaly event information of the target energy storage device.

[0021] Real-time operational data continuously collected during the operation of the target energy storage device is extracted. The data includes core parameters such as voltage, current, temperature, charging and discharging power, and operating frequency at various monitoring points of the device. The specific values ​​of each parameter are extracted at a fixed time interval of 0.2 seconds. For parameter values ​​with different dimensions, dimension normalization is performed according to the preset normal operating extreme value range of each parameter. All parameter values ​​are uniformly converted to a fixed value range of 0-1. Then, according to the preset parameter arrangement order of voltage, current, temperature, charging and discharging power, and operating frequency, all normalized parameter values ​​are sequentially and orderly combined to form an ordered vector structure with the same number dimension as the core parameters. This ordered vector structure is the real-time operational feature vector of the target energy storage device.

[0022] A pre-set anomaly feature library is retrieved, which contains anomaly feature templates corresponding to various abnormal states of pre-stored energy storage devices. Each anomaly feature template is an ordered vector structure with dimensions completely consistent with the real-time operating feature vector, and each anomaly feature template corresponds to a unique anomaly type identifier for the energy storage device. The real-time operating feature vector of the target energy storage device is compared with the feature values ​​of each anomaly feature template in the anomaly feature library dimension by dimension. If the difference between the feature values ​​of the two in the same dimension is within the preset feature value matching threshold of 0.03, the feature value of that dimension is determined to be matched. The matching degree of all anomaly feature templates is statistically analyzed one by one, and the number of matching dimensions between each anomaly feature template and the real-time operating feature vector is counted. The proportion of the number of matching dimensions to the total number of dimensions is calculated. The anomaly type identifier corresponding to the anomaly feature template whose proportion reaches the preset matching proportion threshold of 92% is determined as the matching result. This matching result is the anomaly type identification result of the target energy storage device.

[0023] Starting from the current time point when the anomaly type identification result is obtained, the real-time operation data of the target energy storage device is extracted backwards on the time axis in 0.2-second increments. The values ​​of each core parameter in the extracted operation data are checked at each moment to ensure they meet the normal operating parameter threshold range of the target energy storage device. The first time point where a core parameter value exceeds the normal operating parameter threshold range is determined as the anomaly start time of the target energy storage device. Pre-bound location tags are retrieved from the real-time operation data. These location tags represent the three-dimensional spatial coordinates of each monitoring point of the energy storage device. Each time point of the target energy storage device's real-time operation data is uniquely associated with the corresponding three-dimensional spatial coordinates of the monitoring point. The determined anomaly start time is bound one-to-one with the three-dimensional spatial coordinates of the monitoring point associated with the real-time operation data at that time. Simultaneously, the anomaly type identification result is standardized and integrated with the bound anomaly start time and the three-dimensional spatial coordinates of the monitoring point to form a complete information set containing anomaly type identifier, anomaly start time, and anomaly occurrence location coordinates. This complete information set is the initial anomaly event information of the target energy storage device.

[0024] Pt.2. Based on the anomaly location coordinates in the initial anomaly event information, the operating conditions of a local area of ​​the target energy storage device are reconstructed to obtain a local state parameter set of the target energy storage device. In this embodiment of the invention, the step of reconstructing the operating conditions of a local area of ​​the target energy storage device based on the anomaly location coordinates in the initial anomaly event information to obtain a local state parameter set of the target energy storage device includes: delineating the spatial topology of the target energy storage device based on the anomaly location coordinates in the initial anomaly event information to obtain a local detection area range of the target energy storage device; collecting operating condition parameters from the sensor data set of the target energy storage device based on the local detection area range to obtain a local state parameter data package of the target energy storage device; performing inter-parameter correlation verification on the state parameters in the local state parameter data package and performing constant area boundary verification on the local state parameter data package; when the state parameters pass the consistency check in the inter-parameter correlation verification and match the spatial positional relationship with the anomaly location coordinates in the anomaly area boundary verification, the local state parameter data package is marked as valid, and a local state parameter record of the target energy storage device is obtained.

[0025] Retrieve the complete spatial topology file of the target energy storage device. This file contains the precise three-dimensional spatial coordinate information of each functional component and each sensing and monitoring node of the device. Using the coordinates of the location of the anomaly in the initial abnormal event information as the core point of the three-dimensional space, extend outward by a preset fixed distance of 6cm along the X, Y, and Z axes to form a closed three-dimensional cubic space region. Based on the boundary coordinates of this three-dimensional cubic space region, accurately delineate the corresponding region in the spatial topology file of the target energy storage device. The delineated three-dimensional cubic space region is the local detection area range of the target energy storage device.

[0026] All sensor monitoring nodes deployed within the local detection area are identified. These nodes are all components of the target energy storage device's sensor data set, specifically including voltage, current, temperature, and internal resistance sensors. Real-time operating condition data of all sensor monitoring nodes within this area are continuously collected at a preset fixed data acquisition interval of 0.3 seconds for a preset duration of 20 seconds. All collected operating condition data are categorized and organized according to sensor node type, data acquisition timestamp, and sensor node three-dimensional spatial coordinates to form a structured data set containing specific acquired values ​​of various operating condition parameters, acquisition time information, and node location information. This structured data set is the local state parameter data package of the target energy storage device.

[0027] Based on the inherent linkage relationship of various state parameters during normal operation of the target energy storage device, a unified standard for parameter correlation verification is formulated. This standard requires that the combination relationship of various state parameter values ​​corresponding to the same acquisition timestamp and the same sensor monitoring node completely match the preset normal parameter combination relationship. According to this standard, all state parameters in the local state parameter data package are cross-verified group by group to complete the parameter correlation verification of the local state parameter data package. At the same time, the three-dimensional spatial coordinates of all sensor monitoring nodes in the local state parameter data package are extracted to check whether the coordinates of each node are within the three-dimensional spatial boundary of the local detection area. In addition, it is checked whether the spatial straight-line distance between the sensor node with numerical deviation and the location of the anomaly is within the preset 8cm boundary threshold. Based on this verification requirement, the boundary verification of the abnormal area of ​​the local state parameter data package is completed.

[0028] The results of the parameter correlation verification are comprehensively evaluated item by item to confirm that all state parameters in the local state parameter data package meet the verification standards and pass the consistency test. At the same time, the results of the abnormal area boundary verification are evaluated as a whole to confirm that the verification results of the data package match the spatial position relationship of the abnormal location coordinates. A preset valid state character encoding identifier is added to the metadata information area of ​​the local state parameter data package. The local state parameter data package with the added identifier is then standardized in data format and converted into a preset structured data format. The local state parameter data package with the completed format processing and the valid state identifier is the local state parameter record of the target energy storage device.

[0029] Pt.3. Based on the coupling relationship between the anomaly type identifier in the initial anomaly event information and the local state parameter set, strategy optimization is performed on the preset repair strategy mapping library to obtain a targeted repair scheme for the target energy storage device; In this embodiment of the invention, the step of optimizing the preset repair strategy mapping library based on the coupling relationship between the anomaly type identifier in the initial anomaly event information and the local state parameter set to obtain a targeted repair scheme for the target energy storage device includes: assigning contribution weights to the state parameters based on the response sensitivity coefficient between the state parameters in the local state parameter record and the anomaly type identifier to obtain... The parameter coupling feature vector of the target energy storage device is used; the similarity between the parameter coupling feature vector and the trigger feature vector of the repair strategy entry in the preset repair strategy mapping library is measured to obtain a candidate repair strategy list for the target energy storage device; based on the contribution weighting result in the parameter coupling feature vector, the candidate repair strategy list is projected into the feature space to obtain the optimal repair strategy identifier of the target energy storage device; the optimal repair strategy identifier is deconstructed to obtain the repair action sequence of the target energy storage device; the repair actions in the repair action sequence are inverted to obtain a targeted repair scheme for the target energy storage device.

[0030] A pre-defined table of anomaly types and state parameter response sensitivity coefficients for energy storage devices is retrieved. Each anomaly type identifier in this table has a unique correspondence with all state parameters in the local state parameter record, and each correspondence is configured with a fixed response sensitivity coefficient in the range of 0-1. The coefficient value directly represents the degree of correlation influence of the corresponding state parameter on the anomaly type. The anomaly type identifier of the target energy storage device and all state parameters in the local state parameter record are extracted. The response sensitivity coefficient corresponding to each state parameter is accurately matched from the table, and this coefficient is used as the contribution weight value of each state parameter and uniquely bound to the corresponding state parameter. According to the original arrangement order of the state parameters in the local state parameter record, all state parameters after binding the contribution weight value are sequentially and orderly combined to form an ordered vector structure containing the specific value of the state parameter and the corresponding contribution weight. This ordered vector structure is the parameter coupling feature vector of the target energy storage device.

[0031] A pre-defined repair strategy mapping library is retrieved. This library contains repair strategy entries corresponding to various anomalies of pre-stored energy devices. Each repair strategy entry has a unique strategy identifier, and each entry is configured with a trigger feature vector that is completely consistent with the dimension of the parameter coupling feature vector. The parameter coupling feature vector of the target energy storage device is compared with the trigger feature vectors of all repair strategy entries in the repair strategy mapping library dimension by dimension. The matching degree of the feature values ​​of the two in the same dimension is judged by the numerical deviation being within the preset range of 0.04. The matching degree of the trigger feature vectors of all repair strategy entries is statistically analyzed, and the proportion of the number of matching dimensions to the total number of dimensions is calculated. All repair strategy entries with a matching ratio of 85% or higher are sorted and arranged in descending order of matching ratio to form an ordered set of strategy entries. This set of strategy entries is the candidate repair strategy list for the target energy storage device.

[0032] The contribution weights of each state parameter in the parameter coupling feature vector are used as feature space projection weights. Feature space projection processing is performed on the trigger feature vectors corresponding to each repair strategy entry in the candidate repair strategy list. The projection processing involves associating and matching the dimensional values ​​of each trigger feature vector with the corresponding projection weights to complete the dimensional value fusion processing, resulting in the projected feature vectors of each repair strategy entry. Then, the projected feature vectors and parameter coupling feature vectors are judged for full-dimensional fit. The fit judgment criterion is the proportion of the number of matches between the fused value and the original value in the same dimension to the total number of dimensions. The repair strategy entry corresponding to the projected feature vector with a proportion reaching a preset 95% fit threshold is determined as the optimal repair strategy. The unique strategy identifier corresponding to the optimal repair strategy is the optimal repair strategy identifier of the target energy storage device.

[0033] The system retrieves complete repair strategy information uniquely bound to the optimal repair strategy identifier from the preset repair strategy mapping library. This information includes all repair actions required to complete the repair strategy, the basic execution requirements of each repair action, and the associated execution conditions between each action. In accordance with the general process specifications for hardware maintenance and operation parameter debugging of energy storage equipment, all repair actions in the complete repair strategy information are decomposed into independently executable repair action units. During the decomposition process, it is ensured that the execution target of each action unit is singular and the operation process is complete. Then, according to the process execution sequence requirements of energy storage equipment repair, all the decomposed independent repair action units are arranged in an orderly manner according to the preset operation sequence logic to form an ordered set composed of independent repair action units. This ordered set is the repair action sequence of the target energy storage equipment.

[0034] All independent repair action units in the repair action sequence are extracted. A pre-set energy storage device repair action standard library is retrieved, and the complete execution specification information corresponding to each repair action unit is obtained from the library. This information includes the executing entity of the repair action, the dedicated execution tool, the specific operation steps, the fixed execution duration, and the immediate status judgment standard after execution. The repair action sequence is then subjected to action sequence inversion processing. The inversion processing involves associating and integrating each repair action unit with its corresponding complete execution specification information one by one according to the predetermined execution order of the repair action sequence. At the same time, based on the overall repair process requirements of the energy storage device, the execution connection conditions, fixed connection duration, and status verification requirements after connection are supplemented between adjacent repair action units. The repair action units, after integrating all execution specification information and supplementing all connection and verification requirements, are then structurally integrated according to the execution order to form a standardized repair document containing a complete execution process, detailed operation specifications, clear connection requirements, and unified status verification standards. This standardized repair document is the targeted repair plan for the target energy storage device.

[0035] Pt.4. Real-time capture of the feedback status data of the targeted repair scheme and assessment of the repair expectation of the feedback status data to obtain the repair action execution record of the target energy storage device; In this embodiment of the invention, the real-time capture of the feedback status data of the targeted repair scheme and assessment of the repair expectation of the feedback status data to obtain the repair action execution record of the target energy storage device includes: synchronously capturing the feedback status data of the targeted repair scheme to obtain the feedback status data stream of the target energy storage device; extracting waveform features from the feedback status data packets in the feedback status data stream to obtain the repair action feedback feature vector of the target energy storage device; quantifying the deviation between the repair action feedback feature vector and the preset expected status feature template in the feature space based on the spatial distance between the repair action feedback feature vector and the expected status feature template to obtain the comprehensive deviation quantization value and execution status identifier of the target energy storage device; and archiving the comprehensive deviation quantization value and the execution status identifier in the log to obtain the repair action execution record of the target energy storage device.

[0036] The formula for calculating the comprehensive deviation quantification value is as follows: In the formula, This is the quantified value of the overall deviation. The voltage recovery feature value is located in the feedback feature vector of the repair action. The voltage expected feature value in the expected state feature template, The current stability feature value in the feedback feature vector of the repair action is... The desired current characteristic value in the desired state characteristic template. The temperature convergence feature value in the feedback feature vector of the repair action is... The desired temperature feature value in the desired state feature template. The preset voltage deviation weighting coefficient, The preset current deviation weighting coefficient, The preset temperature deviation weighting coefficient, This is a preset stability constant.

[0037] The execution trigger signals of each repair action in the targeted repair plan for the target energy storage device are retrieved. At the same time as the repair action is started, the status data synchronous capture process of the corresponding monitoring point of the device is initiated. The captured status data includes the core parameters of voltage, current, temperature and internal resistance of the equipment repair area. Each core parameter is continuously collected at a fixed acquisition interval of 0.1 seconds. The acquisition process is completely synchronized with the execution process of the repair action until the repair action is completed. All the acquired status data are continuously arranged in the order of acquisition timestamps to form a continuous data sequence containing parameter acquisition values, acquisition time and monitoring point information. This continuous data sequence is the feedback status data stream of the target energy storage device.

[0038] The feedback status data stream is segmented according to the execution time nodes of each independent repair action unit in the targeted repair scheme. The segmented data corresponding to each repair action unit forms an independent feedback status data packet. For each core parameter data in the feedback status data packet, the corresponding time-value change waveform is plotted. From the waveform, three types of core waveform feature values ​​of each parameter are extracted: steady-state holding value, fluctuation amplitude, and convergence trend value. After extraction, according to the preset parameter arrangement order of voltage, current, temperature, and operating internal resistance, the three types of waveform feature values ​​corresponding to each parameter are sequentially and orderly combined to form an ordered vector structure with the same dimension as the total number of feature values. This ordered vector structure is the repair action feedback feature vector of the target energy storage device.

[0039] A preset desired state feature template is retrieved. This template is an ordered vector structure with dimensions completely consistent with the feedback feature vector of the repair action. The values ​​of each dimension are the standard waveform feature values ​​that each parameter should reach after the corresponding repair action of the energy storage device is completed. The feedback feature vector of the repair action and the desired state feature template are placed in the same feature space for a dimension-by-dimensional numerical comparison. The numerical difference between the feedback feature value and the desired feature value in each dimension is calculated. The numerical differences of all dimensions are integrated according to a preset equal proportion to obtain a comprehensive deviation quantification value that represents the overall degree of difference. This quantification value is in a fixed numerical range of 0-1. At the same time, a quantification value judgment threshold is set. When the comprehensive deviation quantification value is ≤0.06, an execution status identifier representing the normal completion of the repair action is generated. When the comprehensive deviation quantification value is >0.06, an execution status identifier representing the incomplete repair action is generated. In this way, the comprehensive deviation quantification value and execution status identifier of the target energy storage device are obtained.

[0040] The basic information of the repair actions of the target energy storage device is retrieved. This information includes the unique identifier of each repair action unit, the execution start time stamp, the execution end time stamp, and the executing entity. This basic information is associated with the corresponding comprehensive deviation quantification value and execution status identifier. According to the preset log archiving format, all associated information is standardized and organized. The organized content includes information classification tags, original data values, and judgment result descriptions. At the same time, a unique archiving number is added to each organized log information. All standardized and organized log information is integrated in an orderly manner according to the execution order of the repair actions to form a structured log set containing complete repair action execution information, deviation assessment results, and status judgment identifiers. This structured log set is the repair action execution record of the target energy storage device.

[0041] The voltage recovery feature value in the repair action feedback feature vector is a core feature value related to the voltage of the energy storage device repair area, which is extracted from the repair action feedback feature vector obtained after extracting the waveform features of the feedback status data packet of the feedback status data stream. This feature value directly reflects the actual recovery state of the voltage parameter during the repair process.

[0042] The current stability feature value in the repair action feedback feature vector is the core feature value related to the current in the repair area of ​​the energy storage device, which is extracted from the repair action feedback feature vector obtained after extracting the waveform features of the feedback state data packet of the feedback state data stream. This feature value directly reflects the actual stable state of the current parameter during the repair process.

[0043] The temperature convergence feature value in the repair action feedback feature vector is a core feature value related to the temperature of the energy storage device repair area, which is extracted from the repair action feedback feature vector obtained after extracting the waveform features of the feedback state data packet of the feedback state data stream. This feature value directly reflects the actual convergence state of the temperature parameter during the repair process.

[0044] The voltage expectation feature value in the expected state feature template is a standard feature value related to the voltage of the energy storage device repair area extracted from the preset expected state feature template. This feature value is the standard recovery state value that the voltage parameter should reach after the repair is completed.

[0045] The expected current characteristic value in the expected state characteristic template is a standard characteristic value related to the current in the repair area of ​​the energy storage device, which is extracted from the preset expected state characteristic template. This characteristic value is the standard stable state value that the current parameter should reach after the repair is completed.

[0046] The expected temperature feature value in the expected state feature template is a standard feature value related to the temperature of the energy storage device repair area extracted from the preset expected state feature template. This feature value is the standard convergence state value that the temperature parameter should reach after the repair is completed.

[0047] The voltage deviation weighting coefficient is a fixed coefficient that is set in advance for the evaluation of the repair effect of energy storage equipment. This coefficient is determined based on the core influence of voltage parameters in the normal operation of energy storage equipment and is specifically used to characterize the proportion of voltage parameter deviation in the overall deviation evaluation.

[0048] The current deviation weighting coefficient is a fixed coefficient that is set in advance for the evaluation of the repair effect of energy storage equipment. This coefficient is determined based on the core influence of the current parameter in the normal operation of the energy storage equipment and is specifically used to characterize the proportion of the current parameter deviation in the overall deviation evaluation.

[0049] The temperature deviation weighting coefficient is a fixed coefficient that is set in advance for the evaluation of the repair effect of energy storage equipment. This coefficient is determined based on the core influence of temperature parameters in the normal operation of energy storage equipment and is specifically used to characterize the proportion of temperature parameter deviation in the overall deviation evaluation.

[0050] The stability constant is a fixed minimum value set in advance to avoid the abnormal situation of the denominator being zero during the calculation of parameter deviations of energy storage equipment. This value is adapted to the parameter calculation range of voltage, current and temperature of energy storage equipment and will not affect the actual calculation results of parameter deviations.

[0051] First, calculate the actual deviation values ​​of the three types of parameters: voltage, current, and temperature. The calculation method is to subtract the expected characteristic value from the corresponding parameter's repair action feedback characteristic value, and use the difference as the numerator. Use the sum of the expected characteristic value of the corresponding parameter and the stability constant as the denominator. Divide the numerator by the denominator to obtain the relative deviation value of each type of parameter.

[0052] The relative deviation values ​​of the three parameters, voltage, current and temperature, are squared to obtain the squared deviation value of each parameter. Then, the squared deviation value of each parameter is multiplied by the corresponding voltage, current and temperature deviation weighting coefficients to obtain the weighted squared deviation value of each parameter.

[0053] The weighted squared deviations of voltage, current, and temperature are summed to obtain the weighted sum of squared deviations of the three parameters. The square root of this weighted sum of squared deviations is then taken to obtain the comprehensive deviation quantification value that can characterize the overall degree of deviation.

[0054] The obtained comprehensive deviation quantization value is compared with a preset fixed judgment threshold. Based on the comparison result, a corresponding execution status identifier is generated. When the quantization value does not exceed the threshold, an execution status identifier indicating that the repair action has been completed normally is generated. When the quantization value exceeds the threshold, an execution status identifier indicating that the repair action has not been completed is generated.

[0055] The core significance of this calculation is to accurately quantify the actual deviations of three core parameters—voltage, current, and temperature—between the feedback feature vector of energy storage device repair actions and the preset expected state feature template. By configuring corresponding deviation weight coefficients for each of the three parameters, the different influences of different parameters in the evaluation of energy storage device repair effects are accurately reflected. At the same time, a stability constant is introduced to avoid abnormal situations in the calculation. After comprehensively integrating the deviation quantification results of the three parameters, a comprehensive quantitative index is obtained that can comprehensively and objectively characterize the overall difference between the actual feedback state of the repair action and the expected state.

[0056] The practical significance of this calculation is that the obtained comprehensive deviation quantification value provides an accurate and unified numerical basis for determining the execution status of energy storage equipment repair actions. By comparing the comprehensive deviation quantification value with the preset judgment threshold, the completion status of the repair action can be clearly distinguished, ensuring the objectivity and consistency of the execution status judgment.

[0057] The subsequent application of this calculation is to provide accurate and quantitative deviation assessment data to support the generation of subsequent repair action execution records, ensuring that the deviation assessment results contained in the repair action execution records are true and effective, and providing referenceable quantitative data for the subsequent verification of the repair effect of energy storage equipment and the optimization of repair strategies for similar anomalies.

[0058] Pt.5. Based on the completion status identifier in the repair action execution record, the target energy storage device is subjected to a full-state parameter synchronous re-examination to obtain the full-state verification comparison result of the target energy storage device. In this embodiment of the invention, the step of performing a full-state parameter synchronous re-examination on the target energy storage device based on the completion status identifier in the repair action execution record to obtain the full-state verification comparison result of the target energy storage device includes: when the execution status identifier of the repair action execution record indicates that the repair action has been completed, triggering the full-state detection channel of the target energy storage device to obtain a full-state detection trigger confirmation signal of the target energy storage device; based on the full-state detection trigger confirmation signal, synchronously acquiring the operating parameters of the target energy storage device to obtain a full-scale operating data package of the target energy storage device; performing time-frequency domain feature deconstruction on each operating parameter in the full-scale operating data package to obtain a full-state feature vector of the target energy storage device; measuring the similarity between the full-state feature vector and the abnormal feature template in the preset abnormal feature library to obtain a matching relationship list of the target energy storage device; encapsulating the verification conclusion of the matching relationship list to obtain the full-state verification comparison result of the target energy storage device.

[0059] The execution status identifier in the repair action execution record is checked. This identifier is a preset character encoding format. When the identifier is the preset "01" code, it indicates that the repair action has been completed. At this time, a detection start command is sent to the full-state detection and control system of the target energy storage device. This command contains the unique device identifier of the target energy storage device, the detection start timestamp, and the full parameter detection command code. After receiving the command, the full-state detection and control system completes a self-test. After the self-test is passed, a standardized signal containing command reception confirmation, detection channel ready status, and detection start time is generated. This standardized signal is the full-state detection trigger confirmation signal of the target energy storage device.

[0060] Upon receiving the full-state detection trigger confirmation signal, the synchronous acquisition process of operating parameters of all monitoring points of the target energy storage device is immediately initiated. The acquired parameters include the core parameters of voltage, current, temperature, charging and discharging power, operating internal resistance, and operating frequency of all points. All core parameters are continuously and synchronously acquired at a fixed acquisition interval of 0.2 seconds for a preset acquisition duration of 30 seconds. All acquired parameter data are structured and classified according to the spatial coordinates of the monitoring points, data acquisition timestamps, and parameter types to form a complete data set containing the acquired parameter values ​​of all points, acquisition time information, and point identification information. This complete data set is the full operation data package of the target energy storage device.

[0061] The time-frequency domain feature deconstruction process is carried out on each operating parameter in the full-volume operation data package. The time-domain feature deconstruction is to extract three types of feature values ​​from the time series data of each parameter: steady-state value, maximum fluctuation value, and average rate of change. The frequency-domain feature deconstruction is to extract three types of feature values: main frequency value, harmonic ratio value, and frequency stability value after frequency domain transformation of the time series data of each parameter. After the time-frequency domain feature values ​​of all parameters are extracted, the six types of time-frequency domain feature values ​​corresponding to each parameter are sequentially and orderly combined according to the preset parameter arrangement order of voltage, current, temperature, charging and discharging power, operating internal resistance, and operating frequency to form an ordered vector structure with the same dimension as the total number of feature values. This ordered vector structure is the full-state feature vector of the target energy storage device.

[0062] A pre-set anomaly feature library is retrieved, which contains anomaly feature templates corresponding to various anomaly states of pre-stored energy devices. Each anomaly feature template is an ordered vector structure with dimensions completely consistent with the full-state feature vector, and each anomaly feature template corresponds to a unique anomaly type identifier. The full-state feature vector of the target energy storage device is compared with the feature values ​​of each anomaly feature template in the anomaly feature library dimension by dimension. If the difference between the feature values ​​of the two in the same dimension is within the preset feature value matching threshold of 0.04, the feature value of that dimension is determined to be matched. The matching degree of all anomaly feature templates is statistically analyzed one by one. The number of matching dimensions between each anomaly feature template and the full-state feature vector is counted and the proportion of the number of matching dimensions to the total number of dimensions is calculated. The identifiers, corresponding anomaly types, and matching dimension proportions of all anomaly feature templates are sorted in an orderly manner to form a list containing complete matching information. This list is the matching relationship list of the target energy storage device.

[0063] A unified analysis is performed on the proportion data of all matching dimensions in the matching relationship list. A preset 90% matching ratio threshold is set. When the matching dimension ratio of a certain abnormal feature template reaches the threshold, it is determined that the template is successfully matched with the full-state feature vector. If it does not reach the threshold, it is determined that the matching is unsuccessful. The number of all successfully matched and unsuccessfully matched abnormal feature templates in the matching relationship list and their corresponding information are counted. According to the preset verification conclusion encapsulation format, the matching result judgment, the abnormal type identifier of the successfully matched, the specific matching dimension ratio of each abnormal feature template, the verification execution time, and the detection point information are standardized, integrated, and encapsulated to form a standardized verification document containing complete verification judgment results and detailed matching data. This standardized verification document is the full-state verification comparison result of the target energy storage device.

[0064] Pt.6 When the full-state verification comparison result shows that the full amount of operating data does not match any abnormal feature in the preset abnormal feature library in the feature space, the repair action execution record and the full-state verification comparison result are archived and sealed to obtain the repair success record of the target energy storage device.

[0065] In this embodiment of the invention, when the full-state verification comparison result indicates that the full operational data does not match any abnormal feature in the preset abnormal feature library in the feature space, the repair action execution record and the full-state verification comparison result are archived and sealed to obtain a repair success record of the target energy storage device. This includes: performing conclusion analysis on the full-state verification comparison result; when the abnormal feature matching judgment conclusion in the full-state verification comparison result indicates that the full operational data does not match any abnormal feature in the abnormal feature library, a repair success archiving trigger signal for the target energy storage device is obtained; based on the repair success archiving trigger signal, the repair action execution record and the full-state verification comparison result are uniquely bound to obtain a repair result associated data packet for the target energy storage device; the repair result associated data packet is mapped to a storage location to obtain a repair archive storage confirmation record for the target energy storage device; and key summary extraction is performed on the archive storage path index and metadata identifier in the repair archive storage confirmation record to obtain a repair success record for the target energy storage device.

[0066] Extract the pre-defined anomaly feature matching judgment conclusion field from the full-state verification comparison results. This field uses a fixed character code to represent the judgment result. The pre-defined "00" code specifically indicates that the full amount of running data does not match any anomaly feature in the pre-defined anomaly feature library. Read the code through field parsing and complete the judgment result confirmation. After confirming that the judgment result is that there is no anomaly feature matching, generate a standardized electrical signal containing the unique device identifier of the target energy storage device, the archive trigger timestamp, and the anomaly feature matching judgment code. This standardized electrical signal is the archive trigger signal for the successful repair of the target energy storage device.

[0067] Upon receiving the successful repair archiving trigger signal, a unique archiving identifier is generated based on the unique device identifier of the target energy storage device and the timestamp of the repair completion. This identifier is a fixed-length code consisting of numbers and letters. This unique archiving identifier is then added to the metadata fields of the repair action execution record and the full-state verification comparison result to complete global binding. Finally, all data content of the bound repair action execution record and the full-state verification comparison result is structured and integrated to form a structured data set containing the unique archiving identifier, the complete record of repair action execution, the complete result of the full-state verification comparison, and the data integration time. This structured data set is the repair result associated data package of the target energy storage device.

[0068] The system retrieves the hierarchical directory rules of the energy storage equipment repair file storage system. These rules establish a four-level storage directory based on the equipment's maintenance region, equipment model, and the year and month of repair completion. The repair result-related data packets are matched to the corresponding four-level storage directory according to the actual information of the target energy storage equipment. A unique storage address is assigned to each data packet, and the data packet is written and stored in the storage system. After confirming that the data packet has been written and that the data verification is correct, the storage system generates a standardized record containing the file storage path index, data packet storage byte size, data writing completion timestamp, and storage data verification code. This standardized record is the repair file storage confirmation record for the target energy storage equipment.

[0069] Extract the complete file storage path index from the repair archive storage confirmation record, and simultaneously extract the core information from the metadata identifier field of this record. This core information includes a unique archive identifier, a unique device identifier for the target energy storage device, an anomaly type identifier for this repair, a repair completion timestamp, and a full-state verification conclusion code. The extracted core information is then structured and organized, and the organized metadata core information is then systematically integrated with the complete archive storage path index to form a concise structured data record containing key retrieval information, archive storage location information, and core verification conclusions. This concise structured data record is the successful repair record for the target energy storage device.

[0070] Example 2

[0071] like Figure 2 As shown in the figure, this embodiment also provides a functional module diagram of an energy storage device anomaly diagnosis and repair system.

[0072] The anomaly diagnosis and repair system for energy storage devices described in this embodiment can be installed in electronic devices. Depending on the functions implemented, the anomaly diagnosis and repair system may include an anomaly identification module, a working condition reconstruction module, a strategy optimization module, a repair evaluation module, a status re-inspection module, and a record archiving module. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0073] In this embodiment, the functions of each module / unit are as follows: The anomaly identification module is used to identify anomaly patterns in the real-time operating data of the target energy storage device to obtain the initial anomaly event information of the target energy storage device; the operating condition reconstruction module is used to reconstruct the operating conditions of a local area of ​​the target energy storage device based on the anomaly occurrence location coordinates in the initial anomaly event information to obtain the local state parameter set of the target energy storage device; the strategy optimization module is used to optimize a preset repair strategy mapping library based on the coupling relationship between the anomaly type identifier in the initial anomaly event information and the local state parameter set to obtain a targeted repair scheme for the target energy storage device; the repair evaluation module... The system is used to capture feedback status data of the targeted repair scheme in real time and evaluate the repair expectation of the feedback status data to obtain the repair action execution record of the target energy storage device; the status re-inspection module is used to perform a full-state parameter synchronous re-inspection of the target energy storage device based on the completion status identifier in the repair action execution record to obtain the full-state verification comparison result of the target energy storage device; the record archiving module is used to archive and preserve the repair action execution record and the full-state verification comparison result when the full-state verification comparison result is that the full amount of operating data does not match any abnormal feature in the preset abnormal feature library in the feature space, to obtain the repair success record of the target energy storage device.

[0074] In detail, each module in the abnormal diagnosis and repair system of the energy storage device described in the embodiments of the present invention adopts the same technical means as the abnormal diagnosis and repair method of the energy storage device described in Embodiment 1 and Embodiment 2, and can produce the same technical effect, which will not be repeated here.

[0075] Figure 3 The graph shows the changes in the comprehensive deviation quantification value before and after repair for different anomaly types. The vertical axis represents the comprehensive deviation quantification value, and the horizontal axis represents the four common anomaly types of energy storage devices: voltage anomaly, current anomaly, temperature exceeding limits, and multi-parameter coupling. The graph indicates that the threshold value for the comprehensive deviation quantification value used to determine the repair effect is 0.06. The graph visually demonstrates that the comprehensive deviation quantification values ​​of the above four anomaly types were at a high level before repair using the method of this invention, and that the comprehensive deviation quantification values ​​were significantly reduced and lower than the preset threshold after repair, reflecting the effectiveness of the method of this invention in repairing different types of anomalies.

[0076] Figure 4This graph shows the change in the comprehensive deviation quantification value during the repair process of energy storage equipment. The vertical axis represents the comprehensive deviation quantification value, and the horizontal axis represents the repair action sequence number. A comprehensive deviation quantification value judgment threshold of 0.06 is set in the graph. At the same time, the graph shows the change curves of the comprehensive deviation quantification value as the repair action progresses under four different repair schemes: targeted, optimized, ordinary, and generalized. The graph clearly shows that the comprehensive deviation quantification value of the targeted repair scheme formulated in this invention decreases rapidly as the repair action is executed and stabilizes below the threshold, which has better repair efficiency and effect compared with other repair schemes.

[0077] Figure 5 The graph shows the change in the feature space matching ratio during the process stages. The vertical axis represents the feature space matching ratio (unit: %), and the horizontal axis represents the key process stages of energy storage equipment anomaly diagnosis and repair, including anomaly identification, full-state re-inspection, and repair. The graph marks the 90% anomaly feature matching ratio threshold. This graph reflects that the feature space matching ratio between the equipment operation data and the anomaly feature library is at a high level during the anomaly identification stage. After repair by the method of this invention, the matching ratio in the full-state re-inspection and repair stages drops significantly and is far below the threshold, indicating that the equipment has effectively escaped the abnormal state.

[0078] Figure 6 This graph shows the influence of the stability constant on the quantified value of the comprehensive deviation. The vertical axis represents the quantified value of the comprehensive deviation, and the horizontal axis represents the stability constant. It uses a logarithmic scale with a value range of 10⁻ 6 Up to 10⁻²; This figure shows that when the stability constant changes within the above range, the comprehensive deviation quantification value has almost no obvious fluctuations and remains stable, verifying that the preset stability constant in this invention is only used to avoid the abnormal situation of the denominator being zero in the parameter deviation calculation, and will not affect the actual calculation result of the comprehensive deviation quantification value.

[0079] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0080] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for diagnosing and repairing anomalies in an energy storage device, characterized in that, The method includes: Pt.

1. Perform anomaly pattern identification on the real-time operating data of the target energy storage device to obtain the initial anomaly event information of the target energy storage device; Pt.

2. Based on the coordinates of the location of the anomaly in the initial anomaly event information, the operating conditions of the local area of ​​the target energy storage device are reconstructed to obtain the local state parameter set of the target energy storage device. Pt.

3. Based on the coupling relationship between the anomaly type identifier in the initial anomaly event information and the local state parameter set, the preset repair strategy mapping library is optimized to obtain a targeted repair scheme for the target energy storage device. Pt.

4. Real-time capture of the feedback status data of the targeted repair scheme, and assessment of the repair expectation of the feedback status data to obtain the repair action execution record of the target energy storage device; Pt.

5. Based on the completion status identifier in the repair action execution record, perform a full-state parameter synchronous re-examination of the target energy storage device to obtain the full-state verification comparison result of the target energy storage device. Pt.

6. When the full-state verification comparison result shows that the full operational data does not match any abnormal feature in the preset abnormal feature library in the feature space, the repair action execution record and the full-state verification comparison result are archived and sealed to obtain the repair success record of the target energy storage device. Based on the coupling relationship between the abnormal type identifier in the initial abnormal event information and the local state parameter set, the preset repair strategy mapping library is optimized to obtain a targeted repair scheme for the target energy storage device, including: Based on the response sensitivity coefficient between the state parameters and the anomaly type identifier in the local state parameter record, the state parameters are weighted according to their contribution to obtain the parameter coupling feature vector of the target energy storage device. The similarity between the parameter coupling feature vector and the trigger feature vector of the repair strategy entry in the preset repair strategy mapping library is measured to obtain a candidate repair strategy list for the target energy storage device. Based on the contribution weighting result in the parameter coupling feature vector, the candidate repair strategy list is projected into the feature space to obtain the optimal repair strategy identifier of the target energy storage device. The optimal repair strategy identifier is deconstructed to obtain the repair action sequence of the target energy storage device; The repair actions in the repair action sequence are inverted to obtain a targeted repair plan for the target energy storage device. The feedback status data of the targeted repair plan is captured in real time, and the repair expectation is evaluated based on the feedback status data to obtain the repair action execution record of the target energy storage device, including: The feedback status data of the targeted repair scheme is captured synchronously to obtain the feedback status data stream of the target energy storage device; Waveform features are extracted from the feedback status data packets in the feedback status data stream to obtain the repair action feedback feature vector of the target energy storage device. Based on the spatial distance between the repair action feedback feature vector and the preset expected state feature template in the feature space, the deviation between the repair action feedback feature vector and the expected state feature template is quantized to obtain the comprehensive deviation quantization value and execution status identifier of the target energy storage device. The comprehensive deviation quantification value and the execution status identifier are archived in a log to obtain the repair action execution record of the target energy storage device. The calculation formula for the comprehensive deviation quantification value is as follows: ； In the formula, This is the quantified value of the overall deviation. The voltage recovery feature value is located in the feedback feature vector of the repair action. The voltage expected feature value in the expected state feature template, The current stability feature value in the feedback feature vector of the repair action is... The desired current characteristic value in the desired state characteristic template. The temperature convergence feature value in the feedback feature vector of the repair action is... The desired temperature feature value in the desired state feature template. The preset voltage deviation weighting coefficient, The preset current deviation weighting coefficient, The preset temperature deviation weighting coefficient, This is a preset stability constant.

2. The method for anomaly diagnosis and repair of an energy storage device as described in claim 1, characterized in that, The step of identifying abnormal patterns in the real-time operating data of the target energy storage device to obtain initial abnormal event information of the target energy storage device includes: The real-time operation data of the target energy storage device is deconstructed using feature parameters to obtain the real-time operation feature vector of the target energy storage device; The real-time running feature vector is mapped to the abnormal feature template in the preset abnormal feature library to obtain the abnormal type identification result of the target energy storage device. Based on the anomaly type identification result, the real-time operation data is traced back in reverse, and spatial coordinates are registered based on the obtained anomaly start time and the location label in the real-time operation data to obtain the initial anomaly event information of the target energy storage device.

3. The method for anomaly diagnosis and repair of an energy storage device as described in claim 1, characterized in that, Based on the coordinates of the location of the anomaly in the initial anomaly event information, the operating conditions of a local area of ​​the target energy storage device are reconstructed to obtain a set of local state parameters of the target energy storage device, including: Based on the coordinates of the location of the anomaly occurrence in the initial anomaly event information, the spatial topology of the target energy storage device is delineated to obtain the local detection area of ​​the target energy storage device. Based on the local detection area, the operating parameters of the sensor data group of the target energy storage device are collected to obtain the local state parameter data package of the target energy storage device. The state parameters in the local state parameter data packet are subjected to parameter correlation verification, and the local state parameter data packet is subjected to constant region boundary verification. When the state parameter passes the consistency check in the parameter correlation verification and matches the spatial position relationship with the coordinates of the location where the anomaly occurred in the anomaly area boundary verification, the local state parameter data packet is marked as a valid state, and the local state parameter record of the target energy storage device is obtained.

4. The method for abnormal diagnosis and repair of an energy storage device as described in claim 1, characterized in that, Based on the completion status identifier in the repair action execution record, the target energy storage device undergoes a full-state parameter synchronous re-check to obtain the full-state verification comparison result of the target energy storage device, including: When the execution status identifier of the repair action execution record indicates that the repair action has been completed, the full state detection channel of the target energy storage device is triggered, and the full state detection trigger confirmation signal of the target energy storage device is obtained; Based on the full-state detection trigger confirmation signal, the operating parameters of the target energy storage device are synchronously acquired to obtain the full operating data package of the target energy storage device. The time-frequency domain features of each operating parameter in the full operating data packet are deconstructed to obtain the full-state feature vector of the target energy storage device. The similarity between the full-state feature vector and the abnormal feature templates in the preset abnormal feature library is measured to obtain a matching relationship list of the target energy storage device. The verification conclusions of the matching relationship list are encapsulated to obtain the full-state verification comparison results of the target energy storage device.

5. The method for anomaly diagnosis and repair of an energy storage device as described in claim 1, characterized in that, When the full-state verification comparison result shows that the full operational data does not match any abnormal feature in the preset abnormal feature library in the feature space, the repair action execution record and the full-state verification comparison result are archived and sealed to obtain the successful repair record of the target energy storage device, including: The conclusion of the full-state verification comparison result is analyzed. When the abnormal feature matching judgment conclusion in the full-state verification comparison result indicates that the full amount of operating data does not match any abnormal feature in the abnormal feature library, the repair success archiving trigger signal of the target energy storage device is obtained. Based on the successful repair archiving trigger signal, the repair action execution record and the full-state verification comparison result are uniquely bound to an identifier to obtain the repair result associated data packet of the target energy storage device; The storage location of the data packets associated with the repair results is mapped to obtain the repair file storage confirmation record of the target energy storage device; By extracting the key summary from the file storage path index and metadata identifier in the repair file storage confirmation record, the successful repair record of the target energy storage device is obtained.

6. An anomaly diagnosis and repair system for an energy storage device, characterized in that, include: An anomaly identification module is used to identify anomaly patterns in the real-time operating data of the target energy storage device to obtain the initial anomaly event information of the target energy storage device. The operating condition reconstruction module is used to reconstruct the operating conditions of a local area of ​​the target energy storage device based on the coordinates of the location of the anomaly in the initial anomaly event information, so as to obtain the local state parameter set of the target energy storage device. The strategy optimization module is used to optimize the preset repair strategy mapping library based on the coupling relationship between the anomaly type identifier in the initial anomaly event information and the local state parameter set, so as to obtain a targeted repair scheme for the target energy storage device. The repair assessment module is used to capture the feedback status data of the targeted repair scheme in real time, and to assess the repair expectation of the feedback status data to obtain the repair action execution record of the target energy storage device. The status re-inspection module is used to perform a full-state parameter synchronous re-inspection of the target energy storage device based on the completion status identifier in the repair action execution record, and obtain the full-state verification comparison result of the target energy storage device. The record archiving module is used to archive and preserve the repair action execution record and the full-state verification comparison result when the full-state verification comparison result shows that the full-scale operating data does not match any abnormal feature in the preset abnormal feature library in the feature space, thereby obtaining a successful repair record of the target energy storage device. The step of optimizing the preset repair strategy mapping library based on the coupling relationship between the abnormality type identifier in the initial abnormal event information and the local state parameter set to obtain a targeted repair scheme for the target energy storage device includes: Based on the response sensitivity coefficient between the state parameters and the anomaly type identifier in the local state parameter record, the state parameters are weighted according to their contribution to obtain the parameter coupling feature vector of the target energy storage device. The similarity between the parameter coupling feature vector and the trigger feature vector of the repair strategy entry in the preset repair strategy mapping library is measured to obtain a candidate repair strategy list for the target energy storage device. Based on the contribution weighting result in the parameter coupling feature vector, the candidate repair strategy list is projected into the feature space to obtain the optimal repair strategy identifier of the target energy storage device. The optimal repair strategy identifier is deconstructed to obtain the repair action sequence of the target energy storage device; The repair actions in the repair action sequence are inverted to obtain a targeted repair plan for the target energy storage device. The feedback status data of the targeted repair plan is captured in real time, and the repair expectation is evaluated based on the feedback status data to obtain the repair action execution record of the target energy storage device, including: The feedback status data of the targeted repair scheme is captured synchronously to obtain the feedback status data stream of the target energy storage device; Waveform features are extracted from the feedback status data packets in the feedback status data stream to obtain the repair action feedback feature vector of the target energy storage device. Based on the spatial distance between the repair action feedback feature vector and the preset expected state feature template in the feature space, the deviation between the repair action feedback feature vector and the expected state feature template is quantized to obtain the comprehensive deviation quantization value and execution status identifier of the target energy storage device. The comprehensive deviation quantification value and the execution status identifier are archived in a log to obtain the repair action execution record of the target energy storage device. The calculation formula for the comprehensive deviation quantification value is as follows: ； In the formula, This is the quantified value of the overall deviation. The voltage recovery feature value is located in the feedback feature vector of the repair action. The voltage expected feature value in the expected state feature template, The current stability feature value in the feedback feature vector of the repair action is... The desired current characteristic value in the desired state characteristic template. The temperature convergence feature value in the feedback feature vector of the repair action is... The desired temperature feature value in the desired state feature template. The preset voltage deviation weighting coefficient, The preset current deviation weighting coefficient, The preset temperature deviation weighting coefficient, This is a preset stability constant.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 5.

Images