Cross-manufacturer battery module multi-fault positioning detection method and system

Abstract

The invention provides a cross-manufacturer battery module multi-fault positioning detection method and system, and the method comprises the steps: analyzing the manufacturer identification information and communication interaction parameters of a battery module, and determining the structural characteristic data of the battery module based on the manufacturer identification information; matching a corresponding parameter acquisition strategy based on the manufacturer identification information and the communication interaction parameters, and acquiring multi-dimensional operation parameters of the battery module based on the parameter acquisition strategy; and through a multi-parameter fusion analysis model and in combination with the structural characteristic data, carrying out correlation calculation on the multi-dimensional operation parameters to obtain a fault positioning result of the battery module, and outputting a fault positioning result and a fault cause analysis report through the upper computer. According to the invention, the defects of poor universality and insufficient accuracy during fault detection of the current battery module are overcome.

Description

Technical Field

[0001] This invention relates to the technical field of battery module data processing, and in particular to a detection method and system for locating multiple faults in battery modules from different manufacturers. Background Technology

[0002] With the rapid development of new energy vehicles, energy storage equipment, and other fields, battery modules, as the core energy supply unit, directly determine the operational quality of end products through their reliability and safety. Currently, there are numerous battery module manufacturers on the market. Based on their own technological approaches, design standards, and patent portfolios, these manufacturers exhibit significant differences in the structural design (such as cell arrangement, sampling point distribution, interface definition, and heat dissipation structure) and communication protocols (such as data transmission format, baud rate, verification methods, and manufacturer-specific data frame definitions).

[0003] Existing battery module testing technologies suffer from the following main drawbacks: First, the testing equipment lacks versatility. Most testing solutions are designed for specific manufacturers or models of battery modules, relying on dedicated communication protocol parsing modules and parameter acquisition strategies. This makes them unsuitable for the structural and communication differences of modules from different manufacturers, requiring companies to equip themselves with multiple sets of testing equipment, increasing testing costs and operational complexity. Second, the accuracy of fault location is insufficient. Existing testing methods often employ single-parameter threshold judgment or fixed-weight multi-parameter analysis modes, failing to consider the structural characteristics of the battery module (such as the thermal sensitivity of densely packed cell areas and the contact reliability of interface areas) for targeted analysis. This makes it difficult to distinguish whether "parameter anomalies are caused by structural design" or "actual faults," leading to misjudgments and missed judgments, and making it impossible to accurately locate the physical area corresponding to the fault. Summary of the Invention

[0004] The main objective of this invention is to provide a detection method and system for locating multiple faults in battery modules from different manufacturers, aiming to overcome the shortcomings of poor universality and insufficient accuracy in current battery module fault detection.

[0005] To achieve the above objectives, this invention provides a detection method for locating multiple faults in battery modules from different manufacturers, comprising the following steps: The manufacturer identification information and communication interaction parameters of the battery module are analyzed, and the structural characteristic data of the battery module are determined based on the manufacturer identification information. Based on the manufacturer identification information and communication interaction parameters, a corresponding parameter acquisition strategy is matched, and multi-dimensional operating parameters of the battery module are acquired based on the parameter acquisition strategy. By using a multi-parameter fusion analysis model and combining the structural characteristic data, the multi-dimensional operating parameters are correlated and calculated to obtain the fault location results of the battery module. The host computer outputs the fault location results and fault cause analysis report.

[0006] Furthermore, the multi-dimensional operating parameters include voltage, current, and temperature; the fault location results include the fault area and fault type.

[0007] Furthermore, before parsing the manufacturer identification information and communication interaction parameters of the battery module, the following steps are included: A unified detection architecture is constructed, which has a built-in multi-protocol parsing module to adapt to the communication protocols of battery modules from different manufacturers, so as to perform protocol parsing for battery modules from different manufacturers.

[0008] Furthermore, the manufacturer identification information and communication interaction parameters of the battery module are analyzed, and the structural characteristic data of the battery module are determined based on the manufacturer identification information, including: The unified detection architecture uses a built-in multi-protocol parsing module to traverse and adapt to different manufacturers' preset communication protocol formats. After establishing a communication connection with the battery module, it extracts the identification field representing the manufacturer's identity and communication interaction parameters from the communication data frame. Based on the parsed identifier field, a pre-defined manufacturer-structural characteristic mapping database is queried. The manufacturer-structural characteristic mapping database stores structural characteristic data corresponding to different models of battery modules from various manufacturers, including cell arrangement, module interface definition, sampling point distribution, heat dissipation structure parameters, and protection threshold range.

[0009] Furthermore, by using a multi-parameter fusion analysis model and combining the structural characteristic data, the multi-dimensional operating parameters are correlated and calculated to obtain the fault location results of the battery module, including: The multi-dimensional operating parameters are associated and matched with the structural characteristic data. Based on the cell arrangement and sampling point distribution in the structural characteristic data, the cell unit and module area corresponding to each set of operating parameters are calibrated. Using a multi-parameter fusion analysis model, the voltage, current, and temperature parameters of the same cell unit and module area are coupled and analyzed to calculate the parameter deviation, parameter change trend similarity, and multi-parameter collaborative anomaly threshold. By comparing the parameter deviation and parameter change trend similarity with the preset manufacturer-specific anomaly judgment threshold, and combining the heat dissipation structure parameters and protection threshold range in the structural characteristic data, the fault area of ​​the battery module is determined. Based on a preset fault type feature library, the multi-parameter collaborative anomaly pattern corresponding to the multi-parameter collaborative anomaly threshold is matched to determine the corresponding fault type. The fault area and fault type are integrated to form the fault location result.

[0010] Furthermore, by using a multi-parameter fusion analysis model and combining the structural characteristic data, the multi-dimensional operating parameters are correlated and calculated to obtain the fault location results of the battery module, including: Based on the cell arrangement, sampling point distribution, and protection threshold range in the structural characteristic data, differentiated weights are assigned to voltage, current, and temperature in the multi-dimensional operating parameters. Combining the heat dissipation structure parameters, module interface definitions, and pre-stored manufacturer design redundancy parameters in the structural characteristic data, personalized anomaly judgment thresholds adapted to the battery module structure are generated for different cell units and module areas. The time-series correlation calculation is performed on the multi-dimensional operating parameters with assigned weights to extract the spatiotemporal collaborative anomaly features of voltage fluctuation frequency, current mutation amplitude, temperature gradient change rate, and the three parameters. The spatiotemporal collaborative anomaly features are compared with the personalized anomaly judgment threshold. Combined with the cell connection method and module topology in the structural characteristic data, the physical location corresponding to the anomaly parameters is traced back to determine the fault area. Based on a preset multi-parameter anomaly pattern-fault type mapping library, the mapping library pre-stores the correspondence between parameter anomaly combination patterns and fault types corresponding to different structural characteristic modules, and matches the fault type.

[0011] Furthermore, by using a multi-parameter fusion analysis model and combining the structural characteristic data, the multi-dimensional operating parameters are correlated and calculated to obtain the fault location results of the battery module, including: Based on the cell arrangement, sampling point distribution, and protection threshold range in the structural characteristic data, a dynamic parameter weight allocation rule adapted to the battery module is constructed to dynamically allocate weights to voltage, current, and temperature among the multi-dimensional operating parameters. The dynamic change curves of voltage, current and temperature after weight allocation are extracted by the time-series correlation algorithm. Combined with the heat dissipation structure parameters in the structural characteristic data, the spatiotemporal coupling degree of parameters in different module regions is calculated. By mapping parameter combinations whose spatiotemporal coupling exceeds a preset threshold with module interface definitions and cell connection methods in structural characteristic data, the physical location corresponding to abnormal parameters is located, and the fault area is obtained. Based on a pre-defined fault-feature mapping library, combined with the structural characteristics corresponding to the physical location, the cause of the fault is deduced in reverse and the fault type is determined.

[0012] This invention also provides a detection system for locating multiple faults in battery modules from different manufacturers, comprising: The parsing module is used to parse the manufacturer identification information and communication interaction parameters of the battery module, and determine the structural characteristic data of the battery module based on the manufacturer identification information; The acquisition module is used to match the corresponding parameter acquisition strategy based on the manufacturer identification information and communication interaction parameters, and to acquire multi-dimensional operating parameters of the battery module based on the parameter acquisition strategy. The calculation module is used to perform correlation calculations on the multi-dimensional operating parameters by combining the structural characteristic data with a multi-parameter fusion analysis model to obtain the fault location results of the battery module. The output module is used to output fault location results and fault cause analysis reports to the host computer.

[0013] The present invention provides a detection method and system for multi-fault location of battery modules across different manufacturers, comprising: parsing the manufacturer identification information and communication interaction parameters of the battery module; determining the structural characteristic data of the battery module based on the manufacturer identification information; matching a corresponding parameter acquisition strategy based on the manufacturer identification information and communication interaction parameters; acquiring multi-dimensional operating parameters of the battery module based on the parameter acquisition strategy; performing correlation calculations on the multi-dimensional operating parameters through a multi-parameter fusion analysis model and the structural characteristic data to obtain the fault location result of the battery module; and outputting the fault location result and fault cause analysis report through a host computer. In this invention, parsing the manufacturer identification information and communication interaction parameters of the battery module and then matching a corresponding parameter acquisition strategy improves versatility; combining the structural characteristic data and performing correlation calculations on the multi-dimensional operating parameters to obtain the fault location result of the battery module improves the accuracy of fault detection; and overcomes the shortcomings of poor versatility and insufficient accuracy in current battery module fault detection methods. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the detection steps for multi-fault location of battery modules across different manufacturers in one embodiment of the present invention; Figure 2 This is a block diagram of a detection system for locating multiple faults in battery modules across different manufacturers, according to one embodiment of the present invention. Figure 3 This is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0015] The implementation, functional features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0017] It is particularly important to note that all technical steps, algorithm applications, and parameter settings in the technical solution of this application have clear technical objectives and application value. They do not utilize complex steps and algorithmic formulas to achieve simple functions. To provide detailed explanations of each step and avoid ambiguity, some conventional algorithms are used for illustration. However, this does not mean that the algorithms and technical features listed herein are the only way to implement the technical solution of this application, nor is it intended to limit the scope of protection of this application. This application is not a combination or stacking of the listed algorithms and technical features; its essence is to exemplify the implementation methods of this application to fully explain it. It does not pursue formal complexity by adding meaningless technical steps, nor does it involve the accumulation of technologies divorced from practical needs; it conforms to the conventional logic of technical improvement and design.

[0018] Reference Figure 1 One embodiment of the present invention provides a detection method for locating multiple faults in battery modules from different manufacturers, comprising the following steps: Step S1: Parse the manufacturer identification information and communication interaction parameters of the battery module, and determine the structural characteristic data of the battery module based on the manufacturer identification information; Step S2: Based on the manufacturer identification information and communication interaction parameters, match the corresponding parameter acquisition strategy, and acquire multi-dimensional operating parameters of the battery module based on the parameter acquisition strategy; Step S3: Using a multi-parameter fusion analysis model and combining the structural characteristic data, perform correlation calculations on the multi-dimensional operating parameters to obtain the fault location results of the battery module. Step S4: Output the fault location results and fault cause analysis report through the host computer.

[0019] In this embodiment, as described in step S1 above, firstly, relying on the multi-protocol parsing module built into the preset unified detection architecture, this module pre-integrates the communication protocol specifications of mainstream battery module manufacturers in the market, and can automatically adapt to the protocol formats of different manufacturers (including data transmission baud rate, verification method, data frame structure, manufacturer-specific field definitions, etc.). By establishing a communication connection with the battery module to be tested, it traverses and parses the key information in the communication data frame, extracts the identification information representing the manufacturer's identity (such as manufacturer code, brand feature code, product model code, etc.), and simultaneously obtains the interaction parameters that ensure the stability of subsequent communication (such as data frame length, key parameter field offset, communication handshake protocol rules, etc.). Subsequently, based on the parsed manufacturer identification information, it determines the core structural characteristic data corresponding to the battery module, including the cell arrangement (such as the number of series and parallel connections, cell spacing), the sampling point distribution location (such as the cell unit corresponding to the voltage / temperature sampling channel), the module interface definition (such as signal pin function, power supply interface specification), heat dissipation structure parameters (such as heat dissipation channel distribution, thermal conductive material type), and protection threshold range (such as overvoltage / undervoltage threshold, overcurrent protection threshold, temperature protection range, etc.).

[0020] As described in step S2 above, based on the parsed manufacturer identification information and communication interaction parameters, the corresponding parameter acquisition strategy is matched from the preset acquisition strategy library. The acquisition strategy library pre-sets differentiated acquisition parameters (such as acquisition frequency and sampling accuracy), acquisition channel configuration (such as activating the corresponding voltage / current / temperature acquisition channel according to the distribution location of sampling points), data transmission rules (such as the data frame sending / receiving timing set based on the communication interaction parameters), and acquisition triggering conditions (such as the threshold triggering logic for starting acquisition) for the module characteristics of different manufacturers, thus avoiding the data acquisition omission or redundancy problems caused by traditional general acquisition strategies. Subsequently, following the matched parameter acquisition strategy, the acquisition module of the detection system is activated to synchronously acquire multi-dimensional operating parameters of the battery module. These multi-dimensional operating parameters cover the core state indicators of the battery module during operation, including voltage parameters (such as individual cell voltage, total module voltage, and cell voltage difference), current parameters (such as charging and discharging current and loop current), and temperature parameters (such as cell temperature at each sampling point, module surface temperature, and temperature of key areas of the heat dissipation structure). During the acquisition process, the matched communication interaction parameters and acquisition timing requirements are strictly followed to ensure the real-time performance and reliability of the acquired data.

[0021] As described in step S3 above, the collected multi-dimensional operating parameters are preprocessed (including noise reduction, outlier removal, and data standardization) to eliminate interference factors during the acquisition process and ensure data quality. Simultaneously, the obtained structural characteristic data is used as input constraints for the multi-parameter fusion analysis model. Subsequently, the preprocessed multi-dimensional operating parameters are input into the multi-parameter fusion analysis model, which is trained using machine learning algorithms (such as gradient boosting trees and neural networks). This model combines key information from the structural characteristic data (such as cell arrangement, sampling point distribution, and heat dissipation structure parameters) to perform multi-dimensional correlation calculations on voltage, current, and temperature parameters. In one embodiment, based on the cell arrangement and sampling point distribution, the specific cell unit and module region corresponding to each set of operating parameters are identified, establishing a mapping relationship between parameter data and physical location. The dynamic change curves of each parameter are extracted using a time-series correlation algorithm, and the spatiotemporal coupling degree of parameters in different module regions is calculated using heat dissipation structure parameters (including parameter differences between different regions at the same time, parameter change rates at different times in the same region, and time-series deviations in the occurrence of multi-parameter peaks), capturing the collaborative anomaly characteristics between parameters. Finally, by comparing the spatiotemporal coupling degree of the parameters with the manufacturer-specific anomaly thresholds calibrated based on structural characteristic data, the physical location (i.e., the fault area) corresponding to the abnormal parameters is located. Combined with the preset fault-feature mapping library, the fault type corresponding to the abnormal parameter combination (such as cell aging, poor interface contact, heat dissipation failure, overcurrent fault, etc.) is matched, and finally a complete fault location result containing the fault area and fault type is formed.

[0022] As described in step S4 above, the host computer, acting as the core of the human-machine interaction of the detection system, first receives the fault location result obtained in step S3. Then, it calls the built-in cause analysis algorithm, combining the obtained structural characteristic data with the multi-parameter correlation calculation process of step S3 to deduce the fault cause in reverse: for example, if the fault area is a densely packed area of ​​battery cells and the temperature parameters are abnormal, it analyzes whether heat accumulation is caused by blocked heat dissipation channels or by excessively dense battery cell arrangement preventing heat dissipation, based on the heat dissipation structure parameters; if the fault type is abnormal voltage, it analyzes whether it is caused by aging of individual battery cells or loose battery cell connections, based on the battery cell arrangement and sampling point distribution. Finally, the host computer outputs the fault location result (including the specific physical location of the fault area and the fault type name) in a visual format (such as text description, graphic annotation, data tables, etc.) and generates a detailed fault cause analysis report. The report covers the core cause of the fault, the specific values ​​and trends of abnormal parameters, the possible impact of the fault, and targeted maintenance suggestions.

[0023] In one embodiment, the multi-dimensional operating parameters include voltage, current, and temperature; the fault location results include the fault area and the fault type.

[0024] In one embodiment, before parsing the manufacturer identification information and communication interaction parameters of the battery module, the process includes: A unified detection architecture is constructed, which has a built-in multi-protocol parsing module to adapt to the communication protocols of battery modules from different manufacturers, so as to perform protocol parsing for battery modules from different manufacturers.

[0025] In this embodiment, the above steps form the core hardware and software framework supporting cross-manufacturer battery module protocol parsing and adaptation. This aims to break down the detection barriers caused by differences in communication protocols and structures among different manufacturers' modules, providing a unified technical platform for subsequent accurate parsing of manufacturer identifiers, collection of operating parameters, and fault analysis. First, based on the fragmented communication protocols and differentiated structural designs of battery modules from multiple manufacturers in the market, a unified detection architecture is specifically constructed. This architecture is not a simple integration of single functional modules, but rather an integrated system encompassing a communication adaptation layer, a data processing layer, and a strategy storage layer. Its core component is a built-in multi-protocol parsing module. This multi-protocol parsing module pre-integrates the communication protocol specifications of mainstream battery module manufacturers (including but not limited to CAN, CAN FD, Modbus, and custom serial port protocols) to build an extensible protocol adaptation library. This library not only stores the basic formats of each protocol (such as data transmission baud rate, verification method, frame header and trailer identifiers), but also includes field definitions, data encoding rules, and communication handshake processes for manufacturer-specific protocols. It can quickly adapt to the communication protocol type of the battery module under test through an automatic identification and traversal matching mechanism. Meanwhile, this multi-protocol parsing module possesses protocol compatibility and scalability, allowing for the inclusion of communication protocol specifications from new manufacturers through firmware upgrades or parameter configurations. This prevents the testing architecture from failing due to the introduction of new module models in the market. By establishing this unified testing architecture, standardized access to battery modules from different manufacturers and with different protocol types can be achieved, ensuring the accuracy and efficiency of subsequent manufacturer identification information and communication interaction parameter parsing. This lays a solid foundation for the smooth implementation of the entire cross-manufacturer testing process.

[0026] In one embodiment, the manufacturer identification information and communication interaction parameters of the battery module are parsed, and the structural characteristic data of the battery module are determined based on the manufacturer identification information, including: The unified detection architecture uses a built-in multi-protocol parsing module to traverse and adapt to different manufacturers' preset communication protocol formats. After establishing a communication connection with the battery module, it extracts the identification field representing the manufacturer's identity and communication interaction parameters from the communication data frame. Based on the parsed identifier field, a pre-defined manufacturer-structural characteristic mapping database is queried. The manufacturer-structural characteristic mapping database stores structural characteristic data corresponding to different models of battery modules from various manufacturers, including cell arrangement, module interface definition, sampling point distribution, heat dissipation structure parameters, and protection threshold range.

[0027] In this embodiment, firstly, the multi-protocol parsing module pre-integrated in the unified detection architecture is invoked. This module, through prior technical research and protocol analysis, integrates various communication protocol formats preset by mainstream battery module manufacturers in the market, covering general communication protocols (such as CAN, CAN FD, Modbus) and manufacturer-defined serial port protocols, and has the ability to automatically identify and adapt protocols. During the detection process, the multi-protocol parsing module first sends a protocol adaptation probe signal to the battery module under test. By traversing and matching the preset protocol formats of various manufacturers, it gradually adjusts the communication parameters (such as baud rate, parity check method, frame header and frame tail identifiers) until a stable bidirectional communication connection is established with the battery module, ensuring that data transmission is free of packet loss and errors. After the communication connection is established, the module receives communication data frames sent by the battery module in real time. Through the built-in field parsing algorithm, it accurately extracts two types of key information from the data frames: First, an identification field that represents the manufacturer's identity. This field is a pre-set exclusive identification information for each manufacturer, including but not limited to the manufacturer's unique code, brand feature code, product model code, protocol version number, etc., which can uniquely identify the manufacturer and specific model of the battery module. Second, communication interaction parameters that ensure the smooth operation of subsequent testing processes. These parameters include data transmission baud rate, data frame length, offset of key parameter fields, communication handshake protocol rules, data verification methods (such as parity check, CRC check), etc., which provide a standardized basis for data transmission in the subsequent parameter acquisition process and avoid data acquisition failure or data distortion due to mismatched communication parameters.

[0028] Next, the parsed manufacturer identification field (such as manufacturer code + product model code) is used as the search keyword, and the preset manufacturer-structural characteristic mapping database is called. This database is a structured data storage system built in the early stage through multi-channel technology accumulation. Its data sources include official technical documents of manufacturers, actual test verification data, industry standards and specifications, etc., to ensure the accuracy and completeness of the data. The database employs a hierarchical storage structure of "manufacturer-model-structural characteristics," pre-storing core structural characteristic data for different battery module models from various manufacturers. This data covers five key categories: 1) Cell arrangement, including the number of cells connected in series and parallel, cell spacing, and the physical layout of cells within the module; 2) Module interface definition, including the functional allocation of signal pins, power supply interface specifications, and data transmission interface type and pin definitions; 3) Sampling point distribution, clearly defining the specific cell unit and module area corresponding to the sampling channels for parameters such as voltage and temperature, ensuring accurate correspondence between sampling data and physical location; 4) Heat dissipation structural parameters, including the distribution of heat dissipation channels, the type and thickness of thermally conductive materials, and the installation location of cooling fans / heat sinks; and 5) Protection threshold ranges, including manufacturer-preset safety operating parameters such as overvoltage protection thresholds, undervoltage protection thresholds, overcurrent protection thresholds, and temperature protection ranges for the battery module. By accurately matching the identification field with the database, the complete structural characteristic data uniquely corresponding to the battery module to be tested can be quickly retrieved. This provides customized support for the matching of parameter acquisition strategies and the constraint calculation of multi-parameter fusion analysis models in subsequent steps, ensuring that the testing process is highly compatible with the structural characteristics of the battery module and avoiding the problem of insufficient accuracy caused by generalized testing.

[0029] In one embodiment, a multi-parameter fusion analysis model is used to perform correlation calculations on the multi-dimensional operating parameters in conjunction with the structural characteristic data to obtain the fault location results of the battery module, including: The multi-dimensional operating parameters are associated and matched with the structural characteristic data. Based on the cell arrangement and sampling point distribution in the structural characteristic data, the cell unit and module area corresponding to each set of operating parameters are calibrated. Using a multi-parameter fusion analysis model, the voltage, current, and temperature parameters of the same cell unit and module area are coupled and analyzed to calculate the parameter deviation, parameter change trend similarity, and multi-parameter collaborative anomaly threshold. By comparing the parameter deviation and parameter change trend similarity with the preset manufacturer-specific anomaly judgment threshold, and combining the heat dissipation structure parameters and protection threshold range in the structural characteristic data, the fault area of ​​the battery module is determined. Based on a preset fault type feature library, the multi-parameter collaborative anomaly pattern corresponding to the multi-parameter collaborative anomaly threshold is matched to determine the corresponding fault type. The fault area and fault type are integrated to form the fault location result.

[0030] In this embodiment, firstly, the collected multi-dimensional operating parameters (voltage, current, temperature) are structured and organized according to the acquisition channel and timestamp to form a standardized data matrix; at the same time, the obtained structural characteristic data is called to extract the core information directly related to spatial positioning: the arrangement of the cells (such as the number of series and parallel groups, the physical coordinate distribution of the cells in the module, and the grouping rules of the cell units) and the distribution of sampling points (such as the cell number corresponding to each voltage sampling channel, the installation coordinates of each temperature sensor, and the module area division corresponding to the current sampling circuit). Subsequently, through an association matching algorithm, each set of operating parameters in the standardized data matrix (including the voltage value of a single sampling point, the current value of a specific circuit, and the temperature value at a specified location) is matched one by one with the spatial information in the structural characteristic data. This clarifies the specific cell unit and module area corresponding to each parameter data (such as the heat dissipation channel coverage area at the front of the module, the dense cell area in the middle of the module, and the interface connection area at the rear of the module). This completes the triple calibration of "parameter data - cell unit - module area", ensuring that all subsequent analyses can accurately anchor the physical spatial location and avoid the defects of traditional testing that "only the parameter is abnormal, but the abnormal location is unknown".

[0031] For the same calibrated cell unit and module area, a multi-parameter fusion analysis model (which is trained and built based on machine learning algorithms and has the ability to perform parameter collaborative analysis) is invoked. First, the voltage, current and temperature parameters of this area are coupled in terms of time and numerical dimensions: in the time dimension, the dynamic change sequence of each parameter is extracted synchronously to eliminate the analysis bias caused by the difference in the acquisition time of different parameters; in the numerical dimension, each parameter is standardized to the same order of magnitude to avoid the influence of parameter unit differences on the correlation calculation. Subsequently, the model completes feature calculations through three core algorithms: First, parameter deviation calculation, using the corresponding protection threshold range in the manufacturer's module structural characteristic data as a benchmark, combined with the statistical mean of parameters from normal modules of the same type, to quantify the degree of deviation between the current parameter and the benchmark value (e.g., voltage deviation = |actual voltage - benchmark voltage| / benchmark voltage); Second, parameter change trend similarity calculation, using algorithms such as cosine similarity and Pearson correlation coefficient to analyze the similarity of the slope of change, peak occurrence sequence, and fluctuation frequency of voltage, current, and temperature parameters within the same time interval, capturing coordinated change features such as "sudden current increase accompanied by a sudden temperature rise"; Third, multi-parameter coordinated anomaly threshold calculation, based on the deviation and change trend similarity of each parameter, using a weighted fusion algorithm (weights are dynamically allocated according to the structural importance of the region in the structural characteristic data) to generate a quantitative threshold that comprehensively reflects the degree of multi-parameter coordinated anomaly, providing a unified standard for subsequent anomaly judgment.

[0032] Then, the calculated parameter deviation and parameter change trend similarity are compared one by one with the preset manufacturer-specific anomaly judgment threshold. The manufacturer-specific anomaly judgment threshold is not a general threshold, but is based on the structural characteristic data of the manufacturer's module (such as protection threshold range, cell material characteristics), combined with a large amount of measured fault data to ensure that the threshold is highly adapted to the characteristics of the module itself (such as the temperature parameter anomaly judgment threshold in the weak heat dissipation area is lower than that in the normal area). If the parameter deviation of a certain area exceeds the corresponding threshold, or the parameter change trend similarity is lower than / higher than the anomaly threshold, it is initially judged that there is an anomaly risk in the area. Subsequently, the heat dissipation structure parameters in the structural characteristic data (such as the heat dissipation channel distribution, thermal conductive material type, heat dissipation efficiency parameters in the area) and the protection threshold range are combined to perform secondary verification on the initial anomaly area: for example, if the temperature parameter deviation of a certain area exceeds the standard, the heat dissipation structure parameters are combined to analyze whether the area is a heat dissipation blind zone (such as the heat dissipation channel blockage caused by dense cell density), and to determine whether the anomaly is a normal deviation caused by the heat dissipation structure design or an actual fault; at the same time, referring to the protection threshold range, the minor anomaly of "parameters close to the threshold but not triggered protection" and the serious fault of "parameters exceeding the threshold" are distinguished. Through the above dual verification, non-fault-related abnormal interference is eliminated, and the specific cell unit and module area corresponding to the fault is accurately located, and the physical coordinates and range of the fault area are clarified.

[0033] Finally, a pre-defined fault type feature library is invoked. This database is a structured knowledge base built through the accumulation of numerous fault cases and the review of manufacturer technical documents. It stores the correspondence between common battery module fault types (such as cell aging, poor interface contact, heat dissipation failure, overcurrent fault, cell short circuit, loose connection, etc.) and multi-parameter collaborative anomaly modes. The multi-parameter collaborative anomaly mode is a quantitative description of the collaborative changes in voltage, current, and temperature parameters when various faults occur (e.g., "cell aging" corresponds to the mode of "slowly increasing voltage deviation, stable current change trend, and no obvious temperature abnormality," and "poor interface contact" corresponds to the mode of "frequent voltage fluctuations, intermittent current interruption, and local slight temperature rise"). Subsequently, the anomaly features corresponding to the calculated multi-parameter collaborative anomaly threshold (including the degree of deviation of each parameter, correlation of change trends, and temporal characteristics of collaborative anomalies) are matched with various fault modes in the fault type feature library for similarity. The most suitable fault type is then found through a pattern recognition algorithm. During the matching process, the structural characteristics of the area are referenced simultaneously (e.g., anomalies in the interface area are prioritized for matching "poor contact" faults, and anomalies in densely packed cell areas are prioritized for matching "precursor to thermal runaway" or "cell short circuit" faults), further improving the accuracy of fault type determination and avoiding misjudgments caused by the crossover of abnormal modes of different fault types.

[0034] The identified fault area information (including specific cell unit number, module area coordinates, and fault area range description) is structured and integrated with the determined fault type, and key parameter data at the time of the fault occurrence (such as the specific values ​​of abnormal parameters, parameter deviation, and collaborative anomaly threshold) is supplemented to form a complete fault location result that includes the physical location of the fault area, the fault type name, and the core abnormal parameters.

[0035] In one embodiment, a multi-parameter fusion analysis model is used to perform correlation calculations on the multi-dimensional operating parameters in conjunction with the structural characteristic data to obtain the fault location results of the battery module, including: Based on the cell arrangement, sampling point distribution, and protection threshold range in the structural characteristic data, differentiated weights are assigned to voltage, current, and temperature in the multi-dimensional operating parameters. Combining the heat dissipation structure parameters, module interface definitions, and pre-stored manufacturer design redundancy parameters in the structural characteristic data, personalized anomaly judgment thresholds adapted to the battery module structure are generated for different cell units and module areas. The time-series correlation calculation is performed on the multi-dimensional operating parameters with assigned weights to extract the spatiotemporal collaborative anomaly features of voltage fluctuation frequency, current mutation amplitude, temperature gradient change rate, and the three parameters. The spatiotemporal collaborative anomaly features are compared with the personalized anomaly judgment threshold. Combined with the cell connection method and module topology in the structural characteristic data, the physical location corresponding to the anomaly parameters is traced back to determine the fault area. Based on a preset multi-parameter anomaly pattern-fault type mapping library, the mapping library pre-stores the correspondence between parameter anomaly combination patterns and fault types corresponding to different structural characteristic modules, and matches the fault type.

[0036] In this embodiment, firstly, three core types of information directly related to parameter importance from the structural characteristic data are retrieved: cell arrangement (e.g., cells in high series-parallel density areas have stronger mutual influence and a higher risk of parameter anomalies), sampling point distribution location (e.g., sampling points near module interfaces or areas with weak heat dissipation have parameters that better reflect the state of critical areas), and protection threshold range (e.g., parameters close to the protection threshold have higher reference value for fault determination). Then, based on a preset weight allocation algorithm, differentiated weights are assigned according to the above structural characteristics: for voltage parameters corresponding to sampling points in densely populated cell areas, higher weights are assigned because the voltage balance of cells in this area directly affects the overall performance of the module; for temperature parameters in areas with weak heat dissipation structures, their weight is increased because heat accumulation in these areas easily leads to faults; for current parameters close to the overcurrent protection threshold, priority weights are given because sudden current changes may cause serious safety issues; and for parameters corresponding to areas with stable structures and low fault risk, the weights are appropriately reduced. This differentiated weight allocation allows subsequent analysis to focus more on key parameters and high-risk areas at the module structure level, avoiding interference from irrelevant parameters and improving the targeting and efficiency of fault identification.

[0037] Next, heat dissipation structural parameters (such as heat dissipation channel distribution, thermal conductivity of thermally conductive materials, and heat dissipation power) and module interface definitions (such as interface contact resistance design values ​​and signal transmission stability parameters) are extracted from the structural characteristic data. Pre-stored manufacturer design redundancy parameters (including the manufacturer's preset allowable range for parameter fluctuations, parameter tolerance thresholds corresponding to structural redundancy, and factory calibration deviation ranges for the same model of module) are also retrieved. Subsequently, for each calibrated cell unit and module area, a dynamic threshold generation algorithm is used to construct personalized anomaly judgment thresholds: for densely populated cell areas with poor heat dissipation structural parameters, the reasonable temperature rise rate and peak upper limit of the area are calculated based on heat dissipation efficiency, generating a temperature anomaly judgment threshold lower than that of conventional areas; for module interface areas, the anomaly judgment thresholds for voltage fluctuations and current transmission interruptions are set by combining the interface contact resistance design value and manufacturer redundancy parameters to adapt to the contact characteristics of the interface area; for areas where the cell connection method is flexible, the judgment threshold for small current fluctuations is relaxed by referring to the vibration tolerance parameters in the manufacturer's design redundancy to avoid misjudgment. The personalized anomaly detection threshold is not a fixed value, but a dynamic threshold range that is highly bound to the structural characteristics of each area and the manufacturer's design standards, ensuring that the anomaly detection not only conforms to the module's own design logic, but also accurately captures real fault signals.

[0038] Furthermore, the weighted voltage, current, and temperature parameters are subjected to time-series correlation processing to eliminate analytical biases caused by differences in the timing of parameter acquisition, and the dynamic change data of each parameter is synchronized based on timestamps. Subsequently, key dynamic features are mined through feature extraction algorithms: voltage fluctuation frequency (the number of times the voltage exceeds the stable range per unit time), current surge amplitude (the maximum change in current within a short period of time), and temperature gradient change rate (the rate of temperature change and spatial distribution gradient per unit time). At the same time, the spatiotemporal coordinated anomaly features of the three parameters are extracted in detail (such as "current surge amplitude exceeding the standard accompanied by a sudden increase in local temperature gradient" and "abnormal voltage fluctuation frequency with time-series correlation with the current changes of adjacent cell units," etc., which are cross-parameter and cross-regional coordinated anomaly patterns). Then, the extracted spatiotemporal coordinated anomaly features are compared one by one with the generated personalized anomaly judgment thresholds, and anomaly feature combinations that exceed the threshold range are screened out. Finally, by combining the cell connection methods (such as series / parallel / hybrid, rigid / flexible connections) in the structural characteristic data with the module topology (such as cell grouping methods, signal transmission paths, and power distribution circuits), the abnormal feature combinations are traced back in reverse. For example, if a region exhibits "abnormal voltage fluctuation frequency + coordinated current change in adjacent regions," combining the cell series connection method with the grouping rules in the module topology, it can be determined that the abnormality originates from a loose cell connection within that series group. If the temperature gradient change rate exceeds the standard and is concentrated at the end of the heat dissipation channel, combining the heat dissipation path design in the module topology, the fault area can be located as the cell area corresponding to the blockage of the heat dissipation channel. Through the above process, the specific cell unit and module area corresponding to the abnormal parameters are accurately located, and the physical coordinates, range, and structural unit to which the fault area belongs are clarified, achieving precise anchoring of the fault area.

[0039] Finally, a pre-defined multi-parameter anomaly mode-fault type mapping library is invoked. This database is a structured knowledge base built through the accumulation of massive fault cases, the compilation of manufacturer fault diagnosis manuals, and experimental testing verification. Its core feature is that it stores data in categories according to "module structural characteristics - anomaly mode - fault type". It pre-stores exclusive fault modes corresponding to modules with different structural characteristics: for example, for module structures with dense cells and weak heat dissipation, it stores the corresponding relationship of "temperature gradient change rate exceeding the standard + voltage fluctuation frequency abnormal → thermal runaway precursor fault"; for module interfaces with plug-in structures, it stores the corresponding relationship of "current sudden change amplitude exceeding the standard + voltage intermittent interruption → interface poor contact fault"; for cell connections with flexible connections, it stores the corresponding relationship of "small current fluctuation + adjacent cell voltage coordination deviation → connection loose fault". Subsequently, the extracted spatiotemporal coordinated anomaly feature combinations (such as "current surge amplitude exceeding the threshold by 20% + local temperature gradient change rate exceeding the standard + voltage fluctuation frequency reaching 5 times / minute") are used as search criteria for precise matching in the mapping library to filter out the fault type that best matches the anomaly pattern and module structural characteristics. During the matching process, the structural characteristics of the fault area are simultaneously referenced (e.g., if the fault area is a cell connection area, connection-related faults are prioritized; if it is a heat dissipation area, heat dissipation-related faults are prioritized) to further eliminate cross-interference and ensure the accuracy of fault type determination. Finally, a clear fault type corresponding to the fault area is obtained (such as cell aging, interface oxidation, heat dissipation channel blockage, overcurrent burnout, etc.).

[0040] In one embodiment, a multi-parameter fusion analysis model is used to perform correlation calculations on the multi-dimensional operating parameters in conjunction with the structural characteristic data to obtain the fault location results of the battery module, including: Based on the cell arrangement, sampling point distribution, and protection threshold range in the structural characteristic data, a dynamic parameter weight allocation rule adapted to the battery module is constructed to dynamically allocate weights to voltage, current, and temperature among the multi-dimensional operating parameters. The dynamic change curves of voltage, current and temperature after weight allocation are extracted by the time-series correlation algorithm. Combined with the heat dissipation structure parameters in the structural characteristic data, the spatiotemporal coupling degree of parameters in different module regions is calculated. By mapping parameter combinations whose spatiotemporal coupling exceeds a preset threshold with module interface definitions and cell connection methods in structural characteristic data, the physical location corresponding to abnormal parameters is located, and the fault area is obtained. Based on a pre-defined fault-feature mapping library, combined with the structural characteristics corresponding to the physical location, the cause of the fault is deduced in reverse and the fault type is determined.

[0041] In this embodiment, firstly, three types of core information highly relevant to parameter importance from the structural characteristic data are retrieved: cell arrangement (e.g., cells in high series-parallel density areas significantly influence each other, and abnormal parameters can easily trigger cascading failures), sampling point distribution location (e.g., sampling points near the core area of ​​the module or at interface connections, whose parameters better reflect the operating status of critical components), and protection threshold range (e.g., parameters close to overvoltage, overcurrent, and overtemperature protection thresholds have higher reference value for fault warning and judgment). Subsequently, based on a preset weight allocation algorithm (e.g., a hybrid weight algorithm combining the analytic hierarchy process and entropy weighting), a dynamic parameter weight allocation rule specifically adapted to this battery module is constructed: for areas with dense cells and limited heat dissipation space, temperature parameters are more sensitive to faults and are assigned higher weights; for voltage parameters with sampling points located in the module interface area, their weight is increased because poor interface contact is a high-frequency fault in this area; for current parameters close to the manufacturer's preset protection threshold, priority weight is given because they may trigger safety protection mechanisms; and for parameters corresponding to areas with stable structures and low fault rates, the weight is appropriately reduced. This dynamic allocation rule ensures that the weights of voltage, current, and temperature parameters are highly matched with the module's own structural characteristics and fault risk distribution, thus avoiding interference from irrelevant parameters.

[0042] Next, for the weighted multi-dimensional operating parameters of voltage, current, and temperature, time-series correlation algorithms (such as dynamic time warping and sliding window time alignment algorithms) are used for time-dimensional synchronization processing to eliminate analysis biases caused by differences in the timing of parameter acquisition. The dynamic change curves of each parameter throughout the entire detection cycle are extracted, intuitively presenting the voltage fluctuation pattern, current change trend, and temperature rise and fall characteristics. Subsequently, the heat dissipation structure parameters in the structural characteristic data (such as heat dissipation channel distribution, thermal conductivity of thermally conductive materials, heat dissipation power design value, and heat dissipation blind zone location markers) are called, and combined with the module area division rules (based on the cell arrangement and preset sampling point distribution locations), the dynamic change curves are bound to specific module areas. Based on this, the spatiotemporal coupling degree of parameters in different module regions is calculated. The spatiotemporal coupling degree index is a quantitative value that comprehensively reflects the degree of coordinated change of voltage, current, and temperature in the time dimension and the correlation characteristics in the spatial dimension within the same region. The specific calculation logic includes: in the time dimension, analyzing the temporal synchronicity of voltage mutation and temperature surge, current fluctuation and voltage instability within the same region (such as whether the temperature rises synchronously within a preset time after a current surge); in the spatial dimension, analyzing the correlation between temperature gradient and voltage difference between adjacent regions in combination with heat dissipation structure parameters (such as whether the blockage of heat dissipation channels leads to abnormal temperature coupling between upstream and downstream regions). Finally, the spatiotemporal coupling degree of parameters in each module region is obtained through weighted calculation to quantify the degree of coordinated anomaly of multiple parameters.

[0043] Then, the calculated spatiotemporal coupling degree of each module's regional parameters is compared with a preset threshold. This preset threshold is not a universal value, but rather a specific threshold calibrated based on the structural characteristic data of the manufacturer's modules, statistical data of parameters from normal operation of the same model of modules, and numerous fault cases, ensuring the adaptability of anomaly detection. Abnormal parameter combinations (such as the synergistic abnormal combination of "voltage fluctuation + current mutation + temperature surge in a certain area") that exceed the preset threshold are selected. Subsequently, the module interface definitions (such as interface pin function allocation, contact resistance design values, and signal transmission paths) and cell connection methods (such as series / parallel / hybrid modes, rigid / flexible connection structures, and connection point distribution) in the structural characteristic data are invoked to perform spatial mapping analysis on abnormal parameter combinations. For example, if the sampling point corresponding to the abnormal parameter combination is located in the module interface area, combined with the distribution of signal pins and power supply pins in the module interface definition, the abnormality can be located as originating from poor contact of a certain pin in the interface. If the abnormal parameter combination is concentrated in the cell area of ​​a certain series path, combined with the cell series connection method and connection point distribution, the fault can be identified as a loose cell connection point or cell failure in that path. If the abnormal parameter combination involves multiple adjacent areas, combined with the heat dissipation structure parameters and cell arrangement, the fault can be determined as a regional heat accumulation abnormality caused by blockage of the heat dissipation channel. Through the above mapping analysis, the specific physical location corresponding to the abnormal parameters is accurately located, the cell unit number, module area coordinates, and range of the fault area are clarified, and a clear fault area definition result is formed.

[0044] Then, a pre-defined fault-feature mapping library is invoked. This database is a structured knowledge base built through the accumulation of massive amounts of cross-manufacturer fault cases, the sorting of manufacturer technical documents, and experimental testing verification. It stores the correspondence between common battery module fault types (such as cell aging, poor interface contact, heat dissipation failure, loose connection, overcurrent burnout, insulation damage, etc.) and abnormal feature combinations and structural characteristics. Subsequently, the obtained abnormal parameter combinations and the physical location information of the fault area are used as the retrieval basis to perform preliminary matching in the mapping library and filter out candidate fault types. Based on this, secondary constraints and reverse deductions are performed by combining the structural characteristics of the fault area: If the fault area is the module interface area, its structural characteristic is a plug-in connection. Combined with the abnormal parameter combination "voltage fluctuation + intermittent current interruption", the fault cause can be deduced to be the increased contact resistance caused by long-term plugging and unplugging of the interface, and the fault type is determined to be "poor interface contact". If the fault area is a densely packed cell area, its structural characteristic is a narrow heat dissipation channel. Combined with the abnormal parameter combination "sudden temperature rise + deterioration of voltage balance", the fault cause can be deduced to be the heat accumulation in the cell due to blockage of the heat dissipation channel, and the fault type is determined to be "cell performance degradation caused by heat dissipation failure". If the fault area is a cell connection point, its structural characteristic is a flexible connection. Combined with the abnormal parameter combination "sudden current change + slight local temperature rise", the fault cause can be deduced to be the loosening of the connection point due to vibration, and the fault type is determined to be "loose cell connection". Through this process, both the accurate determination of the fault type and the source analysis of the fault cause are achieved.

[0045] In the above embodiments, this application incorporates some existing algorithms and technical features for explanation and description to make the specification more detailed, clear, and complete, thus complying with the provisions of the Patent Law. However, this is not achieved by using a series of complex steps and algorithmic formulas, nor by complicating the technical solution, nor by combining or stacking conventional or simple features. The existing algorithms and technical features listed are for the purpose of disclosing the specific implementation methods of each step of this application (not to limit this application) and to avoid situations where this application cannot be implemented.

[0046] Reference Figure 2 Another embodiment of the present invention also provides a detection system for locating multiple faults in battery modules from different manufacturers, comprising: The parsing module is used to parse the manufacturer identification information and communication interaction parameters of the battery module, and determine the structural characteristic data of the battery module based on the manufacturer identification information; The acquisition module is used to match the corresponding parameter acquisition strategy based on the manufacturer identification information and communication interaction parameters, and to acquire multi-dimensional operating parameters of the battery module based on the parameter acquisition strategy. The calculation module is used to perform correlation calculations on the multi-dimensional operating parameters by combining the structural characteristic data with a multi-parameter fusion analysis model to obtain the fault location results of the battery module. The output module is used to output fault location results and fault cause analysis reports to the host computer.

[0047] In this embodiment, the specific implementation of each module in the above system embodiment is described in the above method embodiment, and will not be repeated here.

[0048] Reference Figure 3 This invention also provides a computer device, which can be a server, and its internal structure can be as follows: Figure 3 As shown, the computer device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores the data corresponding to this embodiment. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0049] Those skilled in the art will understand that Figure 3 The structures shown are merely block diagrams of some structures related to the present invention and do not constitute a limitation on the computer devices on which the present invention is applied.

[0050] An embodiment of the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method. It is understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0051] In summary, the detection method and system for multi-fault localization of battery modules across different manufacturers provided in this embodiment of the invention includes: parsing the manufacturer identification information and communication interaction parameters of the battery module; determining the structural characteristic data of the battery module based on the manufacturer identification information; matching a corresponding parameter acquisition strategy based on the manufacturer identification information and communication interaction parameters; acquiring multi-dimensional operating parameters of the battery module based on the parameter acquisition strategy; performing correlation calculations on the multi-dimensional operating parameters using a multi-parameter fusion analysis model combined with the structural characteristic data to obtain the fault localization result of the battery module; and outputting the fault localization result and fault cause analysis report through a host computer. In this invention, parsing the manufacturer identification information and communication interaction parameters of the battery module and then matching a corresponding parameter acquisition strategy improves versatility; combining the structural characteristic data and performing correlation calculations on the multi-dimensional operating parameters to obtain the fault localization result of the battery module improves the accuracy of fault detection; and overcomes the shortcomings of poor versatility and insufficient accuracy in current battery module fault detection methods.

[0052] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the present invention and embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0053] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0054] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A detection method for locating multiple faults in battery modules from different manufacturers, characterized in that, Includes the following steps: The manufacturer identification information and communication interaction parameters of the battery module are analyzed, and the structural characteristic data of the battery module are determined based on the manufacturer identification information. Based on the manufacturer identification information and communication interaction parameters, a corresponding parameter acquisition strategy is matched, and multi-dimensional operating parameters of the battery module are acquired based on the parameter acquisition strategy. By using a multi-parameter fusion analysis model and combining the structural characteristic data, the multi-dimensional operating parameters are correlated and calculated to obtain the fault location results of the battery module. The host computer outputs the fault location results and fault cause analysis report.

2. The detection method for multi-fault location of battery modules across manufacturers according to claim 1, characterized in that, The multi-dimensional operating parameters include voltage, current, and temperature; the fault location results include the fault area and fault type.

3. The detection method for multi-fault location of battery modules across manufacturers according to claim 1, characterized in that, Before parsing the manufacturer identification information and communication interaction parameters of the battery module, the following is included: A unified detection architecture is constructed, which has a built-in multi-protocol parsing module to adapt to the communication protocols of battery modules from different manufacturers, so as to perform protocol parsing for battery modules from different manufacturers.

4. The detection method for multi-fault location of battery modules across manufacturers according to claim 3, characterized in that, The manufacturer identification information and communication interaction parameters of the battery module are analyzed, and the structural characteristic data of the battery module are determined based on the manufacturer identification information, including: The unified detection architecture uses a built-in multi-protocol parsing module to traverse and adapt to different manufacturers' preset communication protocol formats. After establishing a communication connection with the battery module, it extracts the identification field representing the manufacturer's identity and communication interaction parameters from the communication data frame. Based on the parsed identifier field, a pre-defined manufacturer-structural characteristic mapping database is queried. The manufacturer-structural characteristic mapping database stores structural characteristic data corresponding to different models of battery modules from various manufacturers, including cell arrangement, module interface definition, sampling point distribution, heat dissipation structure parameters, and protection threshold range.

5. The detection method for multi-fault location of battery modules across manufacturers according to claim 1, characterized in that, By using a multi-parameter fusion analysis model and combining the structural characteristic data, the multi-dimensional operating parameters are correlated and calculated to obtain the fault location results of the battery module, including: The multi-dimensional operating parameters are associated and matched with the structural characteristic data. Based on the cell arrangement and sampling point distribution in the structural characteristic data, the cell unit and module area corresponding to each set of operating parameters are calibrated. Using a multi-parameter fusion analysis model, the voltage, current, and temperature parameters of the same cell unit and module area are coupled and analyzed to calculate the parameter deviation, parameter change trend similarity, and multi-parameter collaborative anomaly threshold. By comparing the parameter deviation and parameter change trend similarity with the preset manufacturer-specific anomaly judgment threshold, and combining the heat dissipation structure parameters and protection threshold range in the structural characteristic data, the fault area of ​​the battery module is determined. Based on a preset fault type feature library, the multi-parameter collaborative anomaly pattern corresponding to the multi-parameter collaborative anomaly threshold is matched to determine the corresponding fault type. The fault area and fault type are integrated to form the fault location result.

6. The detection method for multi-fault location of battery modules across manufacturers according to claim 1, characterized in that, By using a multi-parameter fusion analysis model and combining the structural characteristic data, the multi-dimensional operating parameters are correlated and calculated to obtain the fault location results of the battery module, including: Based on the cell arrangement, sampling point distribution, and protection threshold range in the structural characteristic data, differentiated weights are assigned to voltage, current, and temperature in the multi-dimensional operating parameters. Combining the heat dissipation structure parameters, module interface definitions, and pre-stored manufacturer design redundancy parameters in the structural characteristic data, personalized anomaly judgment thresholds adapted to the battery module structure are generated for different cell units and module areas. The time-series correlation calculation is performed on the multi-dimensional operating parameters with assigned weights to extract the spatiotemporal collaborative anomaly features of voltage fluctuation frequency, current mutation amplitude, temperature gradient change rate, and the three parameters. The spatiotemporal collaborative anomaly features are compared with the personalized anomaly judgment threshold. Combined with the cell connection method and module topology in the structural characteristic data, the physical location corresponding to the anomaly parameters is traced back to determine the fault area. Based on a preset multi-parameter anomaly pattern-fault type mapping library, the mapping library pre-stores the correspondence between parameter anomaly combination patterns and fault types corresponding to different structural characteristic modules, and matches the fault type.

7. The detection method for multi-fault location of battery modules across manufacturers according to claim 1, characterized in that, By using a multi-parameter fusion analysis model and combining the structural characteristic data, the multi-dimensional operating parameters are correlated and calculated to obtain the fault location results of the battery module, including: Based on the cell arrangement, sampling point distribution, and protection threshold range in the structural characteristic data, a dynamic parameter weight allocation rule adapted to the battery module is constructed to dynamically allocate weights to voltage, current, and temperature among the multi-dimensional operating parameters. The dynamic change curves of voltage, current and temperature after weight allocation are extracted by the time-series correlation algorithm. Combined with the heat dissipation structure parameters in the structural characteristic data, the spatiotemporal coupling degree of parameters in different module regions is calculated. By mapping parameter combinations whose spatiotemporal coupling exceeds a preset threshold with module interface definitions and cell connection methods in structural characteristic data, the physical location corresponding to abnormal parameters is located, and the fault area is obtained. Based on a pre-defined fault-feature mapping library, combined with the structural characteristics corresponding to the physical location, the cause of the fault is deduced in reverse and the fault type is determined.

8. A detection system for locating multiple faults in battery modules from different manufacturers, characterized in that, include: The parsing module is used to parse the manufacturer identification information and communication interaction parameters of the battery module, and determine the structural characteristic data of the battery module based on the manufacturer identification information; The acquisition module is used to match the corresponding parameter acquisition strategy based on the manufacturer identification information and communication interaction parameters, and to acquire multi-dimensional operating parameters of the battery module based on the parameter acquisition strategy. The calculation module is used to perform correlation calculations on the multi-dimensional operating parameters by combining the structural characteristic data with a multi-parameter fusion analysis model to obtain the fault location results of the battery module. The output module is used to output fault location results and fault cause analysis reports to the host computer.

Images