A method and system for processing vehicle leakage current data

Abstract

This invention relates to the field of vehicle electrical system testing technology, specifically to a method and system for detecting leakage current in vehicles without power interruption. The method includes: acquiring vehicle status signals and determining the vehicle's current operating condition; if the vehicle's current operating condition is a sleep preparation state, outputting a sleep guidance command to establish an uninterrupted power interruption detection path; connecting the detection device in parallel with the vehicle's battery negative terminal circuit to form a bypass path, switching to a series state while maintaining the power supply to the vehicle's electronic control unit, thus obtaining a series detection circuit; acquiring current signals and performing conversion and amplification processing to obtain an amplified voltage signal; performing multi-dimensional error compensation processing on the amplified voltage signal to generate compensated current data; if it is determined that the vehicle has entered a stable sleep state, comparing the compensated current data with a preset threshold to generate a leakage current determination result; and outputting a detection report. This invention achieves uninterrupted power interruption detection in all scenarios, avoids data loss, and improves detection accuracy and efficiency.

Description

Technical Field

[0001] This invention relates to the field of vehicle data analysis and fault diagnosis technology, specifically to a method and system for processing vehicle leakage data. Background Technology

[0002] The higher the degree of electrification and intelligence of a vehicle, the more important the stability of its low-voltage electrical system becomes. However, intermittent and hidden leakage faults caused by retrofitting, aging wiring harnesses, and module failures are recognized as difficult problems in the repair industry.

[0003] Traditional leakage current detection methods have fundamental shortcomings in dealing with such problems:

[0004] First, there are data silos and reliance on experience. Traditional multimeters or testing instruments only provide instantaneous, isolated readings. Fault diagnosis relies entirely on the technician's personal experience. For intermittent faults that sometimes leak and sometimes don't, technicians often need to spend several days repeatedly testing, or eventually give up, resulting in extremely low customer satisfaction.

[0005] Secondly, there is a lack of historical data tracing and trend analysis capabilities. Repair shops cannot obtain historical data on vehicle electrical leakage. When a vehicle returns to the shop due to a dead battery, technicians cannot access the detailed waveforms and data from the previous inspection, making it impossible to determine whether the fault is a recurrence or a new one, often requiring repairs to start from scratch.

[0006] Third, fault knowledge cannot be accumulated and reused. The valuable detection waveform data and repair steps of every successfully resolved complex leakage current case are lost as the repairman forgets them. Repair companies lack a tool to transform individual experience into an organizational knowledge base, hindering technological progress across the industry.

[0007] One solution in the related technology is a current logger with Bluetooth / WiFi data transmission capabilities. Specifically, the device can record current data for extended periods and allow users to view simple data curves via a mobile application.

[0008] Compared to pure instruments, this technology enables data logging and wireless viewing, making it easier to detect intermittent faults. However, it has the following technical defects and limitations:

[0009] The data processing capabilities are weak, and analysis still relies on manual labor: the equipment is only responsible for recording and replaying raw data, without providing any intelligent analysis. Users still need to stare at the graphs and visually search for abnormal peaks, making the analysis difficult and inefficient.

[0010] Zero value mining of data: The data is limited to single-use of a single device, without cloud storage, cross-vehicle or cross-time period data comparison and analysis, and no fault model or knowledge base that can be used for industry reference.

[0011] Limited functionality and lack of a closed-loop system: It is merely a data logger and cannot be integrated with the repair shop's management system, the insurance company's damage assessment system, or the car owner's mobile application, thus failing to form a service loop from detection, diagnosis, repair to reporting. Summary of the Invention

[0012] To address the aforementioned technical problems, this invention provides a vehicle leakage current data processing method and system, aiming to transform discrete detection behaviors into continuous data assets, achieve a leap from manual experience-based diagnosis to data-driven intelligent diagnosis, significantly improve the efficiency of resolving difficult and intermittent leakage current faults, and provide a data foundation for vehicle lifecycle health management.

[0013] To achieve the above objectives, the present invention provides the following technical solution:

[0014] On one hand, embodiments of the present invention provide a method for processing vehicle leakage current data, the method comprising the following steps:

[0015] S100: Obtain the raw data of vehicle leakage detection and the trained intelligent diagnostic model for leakage fault. The intelligent diagnostic model for leakage fault includes a leakage feature map library, a time-series diagnostic network based on a long short-term memory network, and a fault prediction model.

[0016] S200, the original data is cleaned, feature extracted and compressed by the edge computing front end to obtain a processed high-efficiency data packet;

[0017] S300, the high-efficiency data packet is uploaded to the cloud data platform, and the cloud data platform associates the high-efficiency data packet with the vehicle identification code to form a vehicle leakage current health record;

[0018] S400, by comparing the data in the vehicle's leakage health record with the leakage feature map in the leakage feature map library through the leakage fault intelligent diagnosis model, the fault type is identified, the fault point is located and the severity of leakage is quantified, and a diagnosis result is obtained.

[0019] S500, the diagnostic results are sent to the multi-application layer, which generates a diagnostic report and outputs warning information based on the diagnostic results.

[0020] Optionally, in S200, the step of performing data cleaning, feature extraction, and data compression on the original data through an edge computing front-end to obtain a processed high-efficiency data packet includes:

[0021] S210, the edge computing front end performs data cleaning on the original data, filtering out environmental noise and occasional spike interference to obtain cleaned data;

[0022] S220, the edge computing front end performs feature extraction on the cleaned data, identifies and marks key events, including current spikes at the moment of locking the car, the stepped platforms of each electronic control unit going offline, the time point of entering stable sleep mode, and the statistical characteristics of sleep current, to obtain data after feature extraction;

[0023] S230, the edge computing front end packages and compresses the data after feature extraction, and encapsulates the original key waveforms and event labels of the key events to obtain the high-efficiency data packet.

[0024] Optionally, in S220, the edge computing front end performs feature extraction on the cleaned data, identifies and labels key events, including:

[0025] The edge computing front end performs time-series analysis on the cleaned data and detects abrupt changes in the current waveform as the current spike at the moment of locking the car.

[0026] The edge computing front end tracks the stepped platforms during the current decrease process and marks each stepped platform as the offline event of the corresponding electronic control unit.

[0027] The edge computing front end monitors the current fluctuation amplitude. When the current fluctuation amplitude is continuously lower than a preset threshold, the corresponding time point is marked as the time point for entering stable sleep mode.

[0028] The edge computing front end calculates the mean current and variance of the current during the stable sleep phase, which are used as statistical characteristics of the sleep current.

[0029] Optionally, in S300, the cloud data platform associates the high-efficiency data packets with the vehicle identification code to form a vehicle leakage current health record, including:

[0030] The cloud-based data platform receives high-efficiency data packets uploaded from different edge computing front-ends and parses the vehicle identification codes in the high-efficiency data packets;

[0031] The cloud-based data platform associates historical high-efficiency data packets of the same vehicle with the vehicle identification code in a time sequence to construct leakage current detection time-series data for the vehicle.

[0032] The cloud-based data platform binds the leakage current detection time-series data with the vehicle's basic information to form the vehicle's leakage current health record.

[0033] Optionally, in S400, the step of comparing the data in the vehicle's leakage current health record with the leakage current feature map in the leakage current feature map library through the intelligent leakage current fault diagnosis model to identify the fault type, locate the fault point, and quantify the severity of the leakage current, and obtain a diagnostic result, includes:

[0034] S410, input the leakage detection time series data in the vehicle leakage health record into the time series diagnostic network based on the long short-term memory network. The time series diagnostic network models the long-term dependency relationship in the leakage detection time series data through a gating mechanism and outputs the fault type identification result and the corresponding confidence level.

[0035] S420, perform pattern matching between the fault type identification result and the typical current-time waveform patterns and characteristic parameters corresponding to each vehicle model and each fault cause stored in the leakage current feature map library to locate the fault point and quantify the severity of leakage current.

[0036] S430, the fault type identification result, the fault location result, and the leakage severity quantification result are integrated to obtain the diagnostic result.

[0037] Optionally, in S410, the time-series diagnostic network models the long-term dependencies in the leakage current detection time-series data through a gating mechanism, and outputs fault type identification results and corresponding confidence levels, including:

[0038] The forget gate in the time-series diagnostic network performs a linear transformation on the hidden state of the previous time step and the input data of the current time step, and then applies a Sigmoid activation function to generate a forget gate vector. The forget gate vector determines the proportion of information retained in the hidden state of the previous time step.

[0039] The input gate in the time-series diagnostic network performs a linear transformation on the hidden state of the previous time step and the input data of the current time step, and then applies the Sigmoid activation function and the hyperbolic tangent activation function respectively to generate the input gating vector and the candidate memory vector.

[0040] The temporal diagnostic network multiplies the forgetting gate vector element-wise with the cell state of the previous time step, and multiplies the input gate vector element-wise with the candidate memory vector and then adds them together to update the cell state at the current time step.

[0041] The output gate in the time-series diagnostic network applies a hyperbolic tangent activation function to the cell state at the current time, and then multiplies element-wise with the output gating vector obtained by linearly transforming the hidden state at the previous time and the input data at the current time and applying a sigmoid activation function to generate the hidden state at the current time.

[0042] The time-series diagnostic network inputs the hidden state at the last moment into the fully connected layer for Softmax normalization classification, and outputs the fault type identification result and corresponding confidence level.

[0043] Optionally, in S400, the method further includes:

[0044] The fault prediction model acquires historical vehicle leakage data, battery health data, and user vehicle usage habit data.

[0045] The fault prediction model performs multi-source feature fusion on the vehicle's historical leakage data, the battery health status, and the user's vehicle usage habit data to construct a vehicle electrical health status vector.

[0046] The fault prediction model inputs the vehicle electrical health status vector into the prediction network, calculates the probability of the vehicle failing to start due to leakage within a preset time period, and generates proactive warning information.

[0047] On the other hand, embodiments of the present invention provide a vehicle leakage current data processing system, including:

[0048] At least one processor;

[0049] At least one memory for storing at least one program;

[0050] When the at least one program is executed by the at least one processor, the at least one processor performs the method described above.

[0051] On the other hand, embodiments of the present invention provide a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to perform the above-described method.

[0052] On the other hand, embodiments of the present invention provide a computer program product, including a computer program or computer instructions, the computer program or computer instructions being stored in a memory, a processor of a computer device reading the computer program or computer instructions from the memory, and the processor executing the computer program or computer instructions to cause the computer device to perform the above-described method.

[0053] The embodiments of the present invention have the following beneficial effects:

[0054] The edge-cloud collaborative vehicle leakage current data processing architecture proposed in this invention uses an edge computing front-end for data cleaning and feature extraction, and a cloud platform for model training and intelligent diagnosis. This effectively solves the problems of weak data processing capabilities and reliance on manual analysis in related technologies, and reduces the diagnosis time from days to minutes.

[0055] The leakage current feature map library and the time-series diagnostic network based on long short-term memory network constructed by this invention can automatically learn the long-term dependencies in leakage current data, accurately identify intermittent and hidden leakage current fault modes, and significantly reduce the reliance on the personal experience of maintenance workers.

[0056] The vehicle leakage health record established by this invention realizes the aggregation, management, storage and analysis of scattered detection data, forming an industry-level leakage characteristic knowledge base, so that every successful diagnostic data can feed back into the cloud model, and the whole system becomes smarter with use.

[0057] This invention empowers intelligent diagnostic capabilities to diverse users such as repair shops, insurance companies, and car owners through a multi-application layer, providing value-added services such as inspection reports, health warnings, and damage assessment basis, and creating a brand-new business model based on vehicle electrical health data. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is a flowchart illustrating the vehicle leakage data processing method in an embodiment of the present invention;

[0060] Figure 2 This is a schematic diagram of the overall architecture of the vehicle leakage current data processing system in an embodiment of the present invention. Detailed Implementation

[0061] The following will provide a clear and complete description of the concept, specific structure, and technical effects of the present invention in conjunction with embodiments and accompanying drawings, so as to fully understand the purpose, solution, and effects of the present invention. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0062] 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 of the invention and are not intended to limit the invention. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of this invention; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this invention as detailed in the appended claims.

[0063] Unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this invention is for descriptive purposes only and is not intended to limit the invention.

[0064] refer to Figure 1 ,like Figure 1 The figure shown is a vehicle leakage current data processing method provided by an embodiment of the present invention. The method includes the following steps:

[0065] S100: Obtain the raw data of vehicle leakage detection and the trained intelligent diagnostic model for leakage fault. The intelligent diagnostic model for leakage fault includes a leakage feature map library, a time-series diagnostic network based on a long short-term memory network, and a fault prediction model.

[0066] Specifically, the system collects raw data on vehicle leakage current detection through an intelligent detection terminal. This raw data includes current data, voltage data, and temperature data, with a sampling rate of no less than 1 Hz. Simultaneously, a pre-trained intelligent leakage fault diagnosis model is loaded. This model consists of three core sub-models: a leakage current feature map library, a time-series diagnostic network based on a long short-term memory network, and a fault prediction model.

[0067] S200, the original data is cleaned, feature extracted and compressed by the edge computing front end to obtain a processed high-efficiency data packet;

[0068] S300, the high-efficiency data packet is uploaded to the cloud data platform, and the cloud data platform associates the high-efficiency data packet with the vehicle identification code to form a vehicle leakage current health record;

[0069] Specifically, the cloud data platform receives high-efficiency data packets uploaded from different edge computing front-ends and parses the vehicle identification code in the high-efficiency data packets; the cloud data platform associates the historical high-efficiency data packets of the same vehicle with the vehicle identification code in a time sequence to construct the leakage current detection time-series data of the vehicle; the cloud data platform binds the leakage current detection time-series data with the vehicle's basic information to form the vehicle's leakage current health record.

[0070] The cloud-based data platform is responsible for data aggregation and management: the cloud receives data packets uploaded from different devices and repair shops, and associates them uniformly according to the vehicle identification number to form the vehicle's leakage current health record.

[0071] S400, by comparing the data in the vehicle's leakage health record with the leakage feature map in the leakage feature map library through the leakage fault intelligent diagnosis model, the fault type is identified, the fault point is located and the severity of leakage is quantified, and a diagnosis result is obtained.

[0072] S500, the diagnostic results are sent to the multi-application layer, which generates a diagnostic report and outputs warning information based on the diagnostic results.

[0073] This invention provides a method and system for processing vehicle leakage current data. It preprocesses raw detection data locally through an edge computing front-end, effectively reducing data transmission volume and cloud storage costs. It aggregates and correlates multi-source data through a cloud data platform to construct a vehicle leakage current health record, solving the problem of data silos. It achieves automatic identification and location of faults through an intelligent leakage current fault diagnosis model, reducing reliance on human experience. It empowers diagnostic capabilities to different user groups through a multi-application layer, forming a complete service loop.

[0074] The core of this embodiment lies in the edge-cloud collaborative vehicle leakage data processing architecture, which realizes intelligent processing of vehicle leakage data through three levels: edge-side intelligent processing, cloud-based model diagnosis, and multi-terminal application empowerment.

[0075] The overall system architecture is as follows Figure 2 As shown, it comprises four core components:

[0076] Data acquisition and edge computing layer (intelligent detection terminal): executed by the uninterruptible leakage current detection device described in this invention or other compatible intelligent detection terminals.

[0077] Cloud-based data platform: responsible for data access, storage, analysis, and model management.

[0078] Intelligent diagnostic model layer: includes a leakage current feature map library, a time-series diagnostic network based on long short-term memory network, and a fault prediction model.

[0079] Multi-application layer: includes maintenance terminal application / personal computer, car owner terminal mini-program, enterprise management backend and application programming interface open platform.

[0080] In some embodiments, in S200, the step of performing data cleaning, feature extraction, and data compression on the original data through an edge computing front-end to obtain a processed high-efficiency data packet includes:

[0081] S210, the edge computing front end performs data cleaning on the original data, filtering out environmental noise and occasional spike interference to obtain cleaned data;

[0082] S220, the edge computing front end performs feature extraction on the cleaned data, identifies and marks key events, including current spikes at the moment of locking the car, the stepped platforms of each electronic control unit going offline, the time point of entering stable sleep mode, and the statistical characteristics of sleep current, to obtain data after feature extraction;

[0083] S230, the edge computing front end packages and compresses the data after feature extraction, and encapsulates the original key waveforms and event labels of the key events to obtain the high-efficiency data packet.

[0084] In some embodiments, in S220, the edge computing front end performs feature extraction on the cleaned data, identifies and labels key events, including:

[0085] The edge computing front end performs time-series analysis on the cleaned data and detects abrupt changes in the current waveform as the current spike at the moment of locking the car.

[0086] The edge computing front end tracks the stepped platforms during the current decrease process and marks each stepped platform as the offline event of the corresponding electronic control unit.

[0087] The edge computing front end monitors the current fluctuation amplitude. When the current fluctuation amplitude is continuously lower than a preset threshold, the corresponding time point is marked as the time point for entering stable sleep mode.

[0088] The edge computing front end calculates the mean current and variance of the current during the stable sleep phase, which are used as statistical characteristics of the sleep current.

[0089] The edge computing front-end data processing flow is as follows: First, the device filters the collected raw current, voltage, and temperature data to remove high-frequency noise and occasional interference signals. Then, it performs time-series feature analysis to identify key event nodes in the current waveform: detecting current abrupt changes to identify current spikes at the moment of vehicle locking; tracking stepped plateaus during current decline to mark the offline events of each electronic control unit; monitoring current fluctuation amplitude to determine the time point for entering a stable sleep state; and calculating the current statistical characteristics during the stable sleep phase. Finally, the processed key waveform data and event tags are compressed and packaged into a high-efficiency data package and uploaded to the cloud.

[0090] In some embodiments, S400, the step of comparing the data in the vehicle's leakage current health record with the leakage current feature map in the leakage current feature map library through the leakage current fault intelligent diagnosis model to identify the fault type, locate the fault point, and quantify the severity of the leakage current, and obtain a diagnostic result, includes:

[0091] S410, input the leakage detection time series data in the vehicle leakage health record into the time series diagnostic network based on the long short-term memory network. The time series diagnostic network models the long-term dependency relationship in the leakage detection time series data through a gating mechanism and outputs the fault type identification result and the corresponding confidence level.

[0092] S420, perform pattern matching between the fault type identification result and the typical current-time waveform patterns and characteristic parameters corresponding to each vehicle model and each fault cause stored in the leakage current feature map library to locate the fault point and quantify the severity of leakage current.

[0093] S430, the fault type identification result, the fault location result, and the leakage severity quantification result are integrated to obtain the diagnostic result.

[0094] The core intelligent diagnostic model is built on big data and machine learning:

[0095] 1. Leakage Current Feature Map Library: Utilizing historically confirmed leakage current case data, a feature map is trained and established. The map contains typical current-time waveform patterns and feature values ​​(such as the sleep current reference value, descent slope, etc.) corresponding to different vehicle models and different fault causes (such as gateway failure to sleep, leakage due to added recorder).

[0096] 2. Temporal Diagnostic Network Based on Long Short-Term Memory (LSTM): A complete time-series data segment of leakage current detection to be diagnosed is input into a trained LSM network. This network can automatically learn long-term dependencies in the data and output diagnostic results, such as: Fault type: Intermittent wake-up of the controller area network, confidence level: 95%.

[0097] The temporal diagnostic network based on long short-term memory network includes three core gating structures: input gate, forget gate, and output gate.

[0098] In some embodiments, in S410, the time-series diagnostic network models the long-term dependencies in the leakage current detection time-series data through a gating mechanism, and outputs fault type identification results and corresponding confidence scores, including:

[0099] The forget gate in the time-series diagnostic network performs a linear transformation on the hidden state of the previous time step and the input data of the current time step, and then applies a Sigmoid activation function to generate a forget gate vector. The forget gate vector determines the proportion of information retained in the hidden state of the previous time step.

[0100] The input gate in the time-series diagnostic network performs a linear transformation on the hidden state of the previous time step and the input data of the current time step, and then applies the Sigmoid activation function and the hyperbolic tangent activation function respectively to generate the input gating vector and the candidate memory vector.

[0101] The temporal diagnostic network multiplies the forgetting gate vector element-wise with the cell state of the previous time step, and multiplies the input gate vector element-wise with the candidate memory vector and then adds them together to update the cell state at the current time step.

[0102] The output gate in the time-series diagnostic network applies a hyperbolic tangent activation function to the cell state at the current time, and then multiplies element-wise with the output gating vector obtained by linearly transforming the hidden state at the previous time and the input data at the current time and applying a sigmoid activation function to generate the hidden state at the current time.

[0103] The time-series diagnostic network inputs the hidden state at the last moment into the fully connected layer for Softmax normalization classification, and outputs the fault type identification result and corresponding confidence level.

[0104] Specifically, at each time step, the forget gate performs a linear transformation on the hidden state from the previous time step and the input data from the current time step, then applies a Sigmoid activation function to generate a forgetting gate vector, which determines the proportion of information retained in the hidden state from the previous time step. The input gate performs a linear transformation on the hidden state from the previous time step and the input data from the current time step, then applies a Sigmoid activation function and a hyperbolic tangent activation function to generate an input gate vector and a candidate memory vector. The forgetting gate vector is multiplied element-wise with the cell state from the previous time step, and the input gate vector is multiplied element-wise with the candidate memory vector and then summed to update the cell state at the current time step. The output gate applies a hyperbolic tangent activation function to the cell state at the current time step, then multiplies element-wise with the output gate vector obtained by performing a linear transformation on the hidden state from the previous time step and applying a Sigmoid activation function to the current time step, generating the hidden state at the current time step. Finally, the hidden state at the last time step is classified by a fully connected layer and Softmax normalization, outputting the fault type identification result and the corresponding confidence score.

[0105] 3. Fault prediction model: Combining historical vehicle leakage data, battery health, and user driving habits, a model is built to predict the probability of starting failure due to leakage within a certain period of time (e.g., 7 days), thus achieving proactive early warning.

[0106] In some embodiments, S400, the method further includes:

[0107] The fault prediction model acquires historical vehicle leakage data, battery health data, and user vehicle usage habit data.

[0108] The fault prediction model performs multi-source feature fusion on the vehicle's historical leakage data, the battery health status, and the user's vehicle usage habit data to construct a vehicle electrical health status vector.

[0109] The fault prediction model inputs the vehicle electrical health status vector into the prediction network, calculates the probability of the vehicle failing to start due to leakage within a preset time period, and generates proactive warning information.

[0110] In some embodiments, the multi-application layer empowers intelligent diagnostic capabilities to various users such as repair shops, insurance companies, and car owners, providing value-added services such as inspection reports, health warnings, and damage assessment basis.

[0111] Specifically, the multi-application layer includes the following service ports:

[0112] On the repair end: Repair technicians click "Smart Diagnosis" on the application, and the system automatically provides fault conclusions, repair suggestions, and relevant technical cases. After the test is completed, a professional portable document report containing vehicle information, test waveforms, diagnostic conclusions, and repair suggestions is generated with one click and can be sent to the customer for confirmation via WeChat.

[0113] For car owners: Car owners can view their vehicle's health report on the mini-program, or receive a proactive push notification during maintenance stating, "Your vehicle has a potential electrical leakage risk; we recommend you go for inspection."

[0114] Enterprise-level / Application Programming Interface (API) Open Platform: Provides a management backend for chain repair shops, allowing them to view inspection data and technician efficiency at each store. Provides an API for insurance companies, enabling them to query vehicle leakage history records during underwriting or damage assessment, using this as a risk assessment factor.

[0115] This invention also provides a vehicle leakage current data processing system, comprising:

[0116] At least one processor;

[0117] At least one memory for storing at least one program;

[0118] When the at least one program is executed by the at least one processor, the at least one processor performs the method described above.

[0119] The content of the above method embodiments is applicable to this embodiment. The specific functions implemented in this embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments. Therefore, they will not be repeated here.

[0120] This invention also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described above. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0121] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0122] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0123] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0124] This invention also provides a computer program product, including a computer program or computer instructions, which are stored in a memory. A processor of a computer device reads the computer program or computer instructions from the memory and executes the computer program or computer instructions, causing the computer device to perform the above-described method.

[0125] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0126] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0127] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0128] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A method for processing vehicle leakage current data, characterized in that, The method includes the following steps: S100: Obtain the raw data of vehicle leakage detection and the trained intelligent diagnostic model for leakage fault. The intelligent diagnostic model for leakage fault includes a leakage feature map library, a time-series diagnostic network based on a long short-term memory network, and a fault prediction model. S200, the original data is cleaned, feature extracted and compressed by the edge computing front end to obtain a processed high-efficiency data packet; S300, the high-efficiency data packet is uploaded to the cloud data platform, and the cloud data platform associates the high-efficiency data packet with the vehicle identification code to form a vehicle leakage current health record; S400, by comparing the data in the vehicle's leakage health record with the leakage feature map in the leakage feature map library through the leakage fault intelligent diagnosis model, the fault type is identified, the fault point is located and the severity of leakage is quantified, and a diagnosis result is obtained. S500, the diagnostic results are sent to the multi-application layer, which generates a diagnostic report and outputs warning information based on the diagnostic results.

2. The method according to claim 1, characterized in that, In S200, the step of performing data cleaning, feature extraction, and data compression on the original data through an edge computing front-end to obtain a processed, high-efficiency data packet includes: S210, the edge computing front end performs data cleaning on the original data, filtering out environmental noise and occasional spike interference to obtain cleaned data; S220, the edge computing front end performs feature extraction on the cleaned data, identifies and marks key events, including current spikes at the moment of locking the car, the stepped platforms of each electronic control unit going offline, the time point of entering stable sleep mode, and the statistical characteristics of sleep current, to obtain data after feature extraction; S230, the edge computing front end packages and compresses the data after feature extraction, and encapsulates the original key waveforms and event labels of the key events to obtain the high-efficiency data packet.

3. The method according to claim 2, characterized in that, In S220, the edge computing front-end performs feature extraction on the cleaned data, identifies and marks key events, including: The edge computing front end performs time-series analysis on the cleaned data and detects abrupt changes in the current waveform as the current spike at the moment of locking the car. The edge computing front end tracks the stepped platforms during the current decrease process and marks each stepped platform as the offline event of the corresponding electronic control unit. The edge computing front end monitors the current fluctuation amplitude. When the current fluctuation amplitude is continuously lower than a preset threshold, the corresponding time point is marked as the time point for entering stable sleep mode. The edge computing front end calculates the mean current and variance of the current during the stable sleep phase, which are used as statistical characteristics of the sleep current.

4. The method according to claim 1, characterized in that, In S300, the cloud data platform associates the high-efficiency data packets with the vehicle identification code to form a vehicle leakage current health record, including: The cloud-based data platform receives high-efficiency data packets uploaded from different edge computing front-ends and parses the vehicle identification codes in the high-efficiency data packets; The cloud-based data platform associates historical high-efficiency data packets of the same vehicle with the vehicle identification code in a time sequence to construct leakage current detection time-series data for the vehicle. The cloud-based data platform binds the leakage current detection time-series data with the vehicle's basic information to form the vehicle's leakage current health record.

5. The method according to claim 1, characterized in that, In S400, the process of comparing data from the vehicle's leakage current health record with leakage current feature maps in the leakage current feature map library using the intelligent leakage current fault diagnosis model to identify the fault type, locate the fault point, and quantify the severity of the leakage current, thereby obtaining a diagnostic result, including: S410, input the leakage detection time series data in the vehicle leakage health record into the time series diagnostic network based on the long short-term memory network. The time series diagnostic network models the long-term dependency relationship in the leakage detection time series data through a gating mechanism and outputs the fault type identification result and the corresponding confidence level. S420, perform pattern matching between the fault type identification result and the typical current-time waveform patterns and characteristic parameters corresponding to each vehicle model and each fault cause stored in the leakage current feature map library to locate the fault point and quantify the severity of leakage current. S430, the fault type identification result, the fault location result, and the leakage severity quantification result are integrated to obtain the diagnostic result.

6. The method according to claim 5, characterized in that, In S410, the time-series diagnostic network models the long-term dependencies in the leakage current detection time-series data through a gating mechanism, and outputs fault type identification results and corresponding confidence levels, including: The forget gate in the time-series diagnostic network performs a linear transformation on the hidden state of the previous time step and the input data of the current time step, and then applies a Sigmoid activation function to generate a forget gate vector. The forget gate vector determines the proportion of information retained in the hidden state of the previous time step. The input gate in the time-series diagnostic network performs a linear transformation on the hidden state of the previous time step and the input data of the current time step, and then applies the Sigmoid activation function and the hyperbolic tangent activation function respectively to generate the input gating vector and the candidate memory vector. The temporal diagnostic network multiplies the forgetting gate vector element-wise with the cell state of the previous time step, and multiplies the input gate vector element-wise with the candidate memory vector and then adds them together to update the cell state at the current time step. The output gate in the time-series diagnostic network applies a hyperbolic tangent activation function to the cell state at the current time, and then multiplies element-wise with the output gating vector obtained by linearly transforming the hidden state at the previous time and the input data at the current time and applying a sigmoid activation function to generate the hidden state at the current time. The time-series diagnostic network inputs the hidden state at the last moment into the fully connected layer for Softmax normalization classification, and outputs the fault type identification result and corresponding confidence level.

7. The method according to claim 1, characterized in that, In S400, the method further includes: The fault prediction model acquires historical vehicle leakage data, battery health data, and user vehicle usage habit data. The fault prediction model performs multi-source feature fusion on the vehicle's historical leakage data, the battery health status, and the user's vehicle usage habit data to construct a vehicle electrical health status vector. The fault prediction model inputs the vehicle electrical health status vector into the prediction network, calculates the probability of the vehicle failing to start due to leakage within a preset time period, and generates proactive warning information.

8. A vehicle leakage current data processing system, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor performs the method as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images