Abnormality analysis method and system and storage medium

Abstract

The invention provides an exception analysis method and system and a storage medium, and the method comprises the steps: constructing an exception large model and an exception knowledge base in a cloud in advance, receiving exception data uploaded by a vehicle terminal in real time, and when exception type information exists in the exception data, matching the exception data according to the exception knowledge base, obtaining a final abnormal conclusion; when the abnormal type information does not exist, inputting the abnormal data into an abnormal large model for reasoning, and matching the abnormal data by adopting the abnormal knowledge base based on a reasoning result to obtain a final abnormal conclusion; and issuing the final abnormal conclusion to the vehicle terminal. According to the method provided by the invention, the technical problems of slow system operation, large resource occupation and dead halt caused by abnormal data of the vehicle-mounted terminal are effectively solved, and the safety and reliability of the vehicle-mounted terminal system are effectively improved.

Description

Technical Field

[0001] This application belongs to the technical field of vehicle fault diagnosis, and particularly relates to an abnormal analysis method, system and storage medium. Background Art

[0002] Existing vehicle infotainment systems usually integrate the control and operation display of all applications on the vehicle. The control of the vehicle infotainment screen has gradually replaced traditional physical buttons and become the core carrier of vehicle function control and information interaction. However, this highly integrated design has a series of technical problems.

[0003] With the development demand of vehicle intelligence, there are more and more application programs in the vehicle terminal system. Such as navigation, hotel, music, communication, etc. Due to the uneven quality of products from each supplier, abnormal conditions may occur during operation, resulting in slow operation of the vehicle infotainment system, large resource occupation, phenomena such as crashing and freezing, affecting the normal use of functions. The common way to handle vehicle faults is to flash the firmware, that is, to flash a usable version. However, due to factors such as vehicle hardware and vehicle usage environment, even for the same version of the system, it cannot run properly on some vehicles or some abnormal faults will still occur after running for a period of time, and the abnormal problems cannot be eradicated. Summary of the Invention

[0004] To solve the above technical problems, this application proposes an abnormal analysis method, system and storage medium that combine an abnormal knowledge base and an abnormal large model to analyze abnormal data.

[0005] Specifically, for the abnormal analysis method proposed in this application, an abnormal large model and an abnormal knowledge base are pre-constructed in the cloud. The abnormal analysis method includes:

[0006] Receiving abnormal data uploaded by the vehicle infotainment terminal in real time.

[0007] Judging whether there is abnormal type information in the abnormal data. If so, matching the abnormal data according to the abnormal knowledge base to obtain a final abnormal conclusion.

[0008] Otherwise, inputting the abnormal data into the abnormal large model for reasoning, and matching the abnormal data according to the reasoning result using the abnormal knowledge base to obtain a final abnormal conclusion.

[0009] And sending the final abnormal conclusion to the vehicle infotainment terminal.

[0010] In the above technical solution, the constructed abnormal large model and abnormal knowledge base are used to analyze abnormal data, thereby effectively improving the accuracy of the abnormal conclusions obtained from the analysis of abnormal data. The constructed abnormal knowledge base can quickly obtain abnormal conclusions, that is, the reasons for the occurrence of abnormalities and the methods for handling abnormalities, improving the efficiency of abnormal data processing. The establishment of the abnormal knowledge base provides effective data support for the subsequent service upgrade of suppliers and the upgrade of software and hardware. By combining the abnormal conclusions obtained from the abnormal knowledge base and the abnormal large model, the reliability of the final abnormal conclusions obtained is improved.

[0011] As an implementation method, the abnormal large model and abnormal knowledge base are pre-constructed in the cloud, and it further includes:

[0012] Obtain a number of historical abnormal data uploaded by the in-vehicle terminal; perform manual analysis on the historical abnormal data to obtain historical abnormal conclusions.

[0013] Train the abnormal large model based on the historical abnormal data and historical abnormal conclusions, and store the historical abnormal data and historical abnormal conclusions in the abnormal knowledge base in a fixed file.

[0014] Training the abnormal large model with a large amount of historical abnormal data and historical abnormal conclusions continuously optimizes and improves the inference accuracy of the abnormal large model. Storing a large amount of historical abnormal data and historical abnormal conclusions in the abnormal knowledge base in a fixed file provides effective data support for the subsequent matching of abnormal conclusions, enabling subsequent abnormal data to quickly match the corresponding abnormal conclusions, thereby improving the efficiency of abnormal processing.

[0015] Further, the matching of the abnormal data includes at least: general matching and keyword matching.

[0016] Based on the abnormal type information, determine whether keyword matching can be performed. If so, perform keyword matching on the abnormal data; otherwise, perform general matching on the abnormal data.

[0017] By using two matching methods, general matching and keyword matching, to match the abnormal conclusions corresponding to the abnormal data, the defects of a single matching method are made up for, and the efficiency and accuracy of abnormal analysis are improved.

[0018] Further, the general matching includes:

[0019] Obtain multiple abnormal handling methods from the abnormal knowledge base based on the abnormal data.

[0020] Match the abnormal data with the multiple abnormal handling methods in sequence to obtain the final abnormal conclusion.

[0021] By sequentially matching the abnormal data with multiple abnormal handling methods, the best abnormal handling method can be identified, effectively improving the accuracy of abnormal analysis. By obtaining multiple abnormal handling methods from the abnormal knowledge base and sequentially matching these abnormal handling methods with the abnormal data alone, automatic identification and handling of abnormal situations can be achieved, reducing the need for manual intervention. By matching the abnormal data with multiple handling methods one by one, a comprehensive analysis of the abnormal data can be ensured.

[0022] Further, the keyword matching includes:

[0023] Obtaining multiple abnormal keywords based on the abnormal data; matching the abnormal keywords with a pre-configured keyword list in the abnormal knowledge base to obtain the final abnormal conclusion.

[0024] By matching the obtained abnormal keywords with the pre-configured keyword list in the abnormal knowledge base, the abnormal conclusion corresponding to the abnormal data can be accurately matched, effectively reducing the probability of misidentification, thereby improving the system's accurate response to abnormal situations. By matching the abnormal keywords with the keyword list, the system can quickly find a suitable abnormal handling method, thus accelerating the decision-making process of abnormal handling.

[0025] Further, after obtaining the final abnormal conclusion, it further includes:

[0026] Recording the intermediate data generated during the matching process of the abnormal data to form associated data; optimizing the final abnormal conclusion based on the associated data.

[0027] By recording the intermediate data generated during the matching process of the abnormal data, the entire abnormal matching process can be traced, and the final abnormal conclusion can be carefully optimized through the associated data, thereby improving the accuracy and reliability of the final abnormal conclusion.

[0028] Further, after sending the final abnormal conclusion to the in-vehicle terminal, it further includes:

[0029] Updating the abnormal knowledge base based on the abnormal data and the final abnormal conclusion, and optimizing the training of the abnormal large model.

[0030] Updating the exception knowledge base with exception data and final exception conclusions ensures that the exception knowledge base contains the latest exception situations and coping strategies, improves the timeliness and accuracy of the exception knowledge base, and expands the data reserve in the exception knowledge base. Optimizing and training the exception large model with exception data and final exception conclusions enables the system to continuously learn and improve, effectively enhancing the reasoning ability of the exception large model for exception data. Over time, the exception large model will become more and more mature, capable of accurately identifying and handling diverse exception situations, thereby improving the stability and reliability of the system.

[0031] Based on the same inventive concept, the present application also proposes an exception analysis system, which includes:

[0032] A data acquisition module for receiving exception data uploaded by the vehicle head unit terminal in real time.

[0033] A type judgment module for judging whether the exception data has exception type information. If so, the exception data is input into the knowledge base matching module; otherwise, the exception data is input into the exception large model reasoning module.

[0034] A knowledge base matching module for matching the exception data according to the exception knowledge base to obtain a final exception conclusion.

[0035] An exception large model reasoning module for reasoning the exception data according to the exception large model and inputting the obtained reasoning result into the knowledge base matching module.

[0036] And a conclusion distribution module for distributing the final exception conclusion to the vehicle head unit terminal.

[0037] Further, the knowledge base matching module further includes:

[0038] A general matching module for obtaining multiple exception handling methods from the exception knowledge base based on the exception data; sequentially matching the exception data with the multiple exception handling methods to obtain the final exception conclusion.

[0039] A keyword matching module for obtaining multiple exception keywords based on the exception data; performing matching based on the exception keywords and a pre-configured keyword list in the exception knowledge base to obtain the final exception conclusion.

[0040] Based on the same inventive concept, the present application also proposes a computer-readable storage medium storing computer-executable instructions, and when the computer-executable instructions are executed by a control processor, the exception analysis method is implemented.

[0041] Compared with the prior art, the present application has at least the following beneficial effects:

[0042] The abnormal analysis method proposed in this application effectively solves the technical problems of slow system operation, high resource occupancy, and system freeze and lag caused by abnormal vehicle machine data, and effectively improves the security and reliability of the vehicle machine terminal system. By analyzing abnormal data through the constructed abnormal large model and abnormal knowledge base, the misjudgment and missed judgment of abnormal data are effectively reduced, and the accuracy of abnormal recognition is improved. Through the constructed abnormal knowledge base, abnormal conclusions can be quickly obtained, thereby improving the efficiency of abnormal data analysis. The establishment of the abnormal knowledge base provides effective data support for the subsequent service upgrade of suppliers and the upgrade of software and hardware. By combining the abnormal conclusions obtained from the abnormal knowledge base and the abnormal large model, the reliability of the final abnormal conclusion obtained is improved. Description of the Drawings

[0043] Figure 1 It is a flowchart of the abnormal analysis method shown in the embodiments of this application.

[0044] Figure 2 It is a schematic diagram of the training of the abnormal large model shown in the embodiments of this application.

[0045] Figure 3 It is a schematic diagram of the abnormal analysis process shown in the embodiments of this application.

[0046] Figure 4 It is a schematic diagram of the supplier updating the vehicle machine terminal system and applications shown in the embodiments of this application.

[0047] Figure 5 It is a schematic diagram of the abnormal analysis system shown in the embodiments of this application. Detailed Embodiments

[0048] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0049] It should be noted that the terms "first", "second", etc. in the specification, claims and above-mentioned drawings of this application are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server comprising a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0050] Embodiment 1:

[0051] Please refer to Figure 1 , and this anomaly analysis method mainly includes steps S100 to S400.

[0052] Among them, step S100 includes: receiving abnormal data uploaded by the vehicle-mounted terminal in real time. Various abnormal data are collected through the data collection service in the vehicle-mounted terminal when an anomaly occurs. For example, abnormal data in various dimensions such as the memory, CPU, disk, IO, and application running logs of the vehicle-mounted terminal application and system are uploaded to the cloud.

[0053] Step S200 includes: determining whether there is abnormal type information in the abnormal data. If so, the abnormal data is matched according to the abnormal knowledge base to obtain a final abnormal conclusion. The abnormal knowledge base is a database pre-constructed in the cloud, which stores a large amount of historical abnormal data and abnormal conclusions. Through the abnormal knowledge base, the abnormal conclusion corresponding to the abnormal data can be quickly and efficiently matched.

[0054] Step S300 includes: otherwise, inputting the abnormal data into an abnormal large model for inference, and matching the abnormal data based on the inference result using the abnormal knowledge base to obtain a final abnormal conclusion. The abnormal large model can mainly be an AI large model pre-constructed in the cloud and trained with a large amount of historical abnormal data and abnormal conclusions. Through the trained abnormal large model, the inference result corresponding to the abnormal data can be accurately obtained.

[0055] And step S400 includes: sending the final abnormal conclusion to the vehicle-mounted terminal. The final abnormal conclusion obtained by combining the abnormal large model and the abnormal knowledge base can make the obtained abnormal conclusion more accurate, thus ensuring the accuracy and reliability of the obtained final abnormal conclusion.

[0056] In the specific implementation process, an abnormal large model and an abnormal knowledge base are pre-constructed in the cloud. The abnormal large model is trained with the obtained historical abnormal data and abnormal conclusions, and the historical abnormal data and historical conclusions are archived and stored in the abnormal knowledge base. In the in-vehicle terminal, various abnormal data are collected in real time through a data collection service and uploaded to the cloud. After the cloud receives the abnormal data, the abnormal type information in the abnormal data is obtained, and based on the abnormal type information, the received new abnormal data is matched with the historical abnormal data and historical abnormal conclusions in the abnormal knowledge base to obtain the final abnormal conclusion. When there is no abnormal type information in the received abnormal data, the cloud inputs the received abnormal data into the abnormal large model for reasoning, and based on the reasoning result, the abnormal knowledge base is used to match the abnormal data to obtain the final abnormal conclusion. The obtained final abnormal conclusion is sent to the in-vehicle terminal.

[0057] In some embodiments, pre-constructing an abnormal large model and an abnormal knowledge base in the cloud further includes:

[0058] Obtaining a number of historical abnormal data uploaded by the in-vehicle terminal; manually analyzing the historical abnormal data to obtain historical abnormal conclusions; training the abnormal large model based on the historical abnormal data and historical abnormal conclusions, and archiving and storing the historical abnormal data and historical abnormal conclusions in the abnormal knowledge base.

[0059] Please refer to Figure 2 , the historical abnormal data collected in real time through the in-vehicle terminal is manually analyzed to obtain the corresponding historical abnormal conclusions. The abnormal large model is trained with a large amount of obtained historical abnormal data and historical abnormal conclusions to improve the accuracy and reliability of the abnormal large model reasoning. And the historical abnormal data and historical abnormal conclusions are archived and stored in the abnormal knowledge base to provide data support for the matching of abnormal conclusions of subsequent abnormal data.

[0060] Optionally, matching the abnormal data includes at least: general matching and keyword matching.

[0061] Based on the abnormal type information, it is judged whether keyword matching can be performed. If so, keyword matching is performed on the abnormal data; otherwise, general matching is performed on the abnormal data.

[0062] Among them, a keyword list can be pre-configured in the exception knowledge base to obtain multiple exception keywords from the exception data, and match based on the exception keywords and the keyword list, so as to obtain the exception type of the exception data. For example, the keywords that can be included in the keyword list are IO logs, cpu / memory logs, running logs, network logs, etc. When there is no exception keyword corresponding to the exception data in the keyword list, general matching is performed on the exception data to obtain the final exception conclusion.

[0063] Optionally, the general matching includes:

[0064] Obtain multiple exception handling methods from the exception knowledge base based on the exception data.

[0065] Match the exception data with the multiple exception handling methods in sequence to obtain the final exception conclusion.

[0066] Among them, before performing general matching on the exception data through the exception knowledge base, preprocessing of the exception data can also be included, such as operations like cleaning, denoising, and text standardization of the exception data, to ensure the quality of the exception data and help extract effective feature information from the exception data. Query and obtain multiple exception handling methods related to it from the exception knowledge base according to the feature information extracted from the exception data, such as restarting the service, checking the network connection, checking the system load, clearing the cache, updating the driver, etc. Match the exception data with the multiple obtained exception handling methods in sequence, sort the multiple exception handling methods according to the matching results, and use the exception handling method that can efficiently solve the exception data as the exception conclusion of the exception data. After obtaining the final exception conclusion, input the exception data and the corresponding final exception conclusion into the exception large model and the exception knowledge base to optimize and train the exception large model and synchronously update the exception knowledge base.

[0067] Optionally, the keyword matching includes:

[0068] Obtain multiple exception keywords based on the exception data; match based on the exception keywords and the keyword list pre-configured in the exception knowledge base to obtain the final exception conclusion.

[0069] Since most of the abnormal data collected and uploaded by the in-vehicle terminal are log data, such as string data and text data, the abnormal keywords of the abnormal data can be obtained by configuring relevant query methods to determine the query range, query method, abnormal value size, etc. of the abnormal data. Match the obtained multiple abnormal keywords with the keyword list pre-configured in the abnormal knowledge base to obtain the corresponding abnormal causes and abnormal solutions of the abnormal data in the abnormal knowledge base. By matching multiple abnormal keywords and keyword lists in multiple dimensions and using a cross-matching method of querying and matching, verify whether the abnormal logs and files conform to the abnormal conclusions in the abnormal knowledge base, thereby reducing the possibility of misjudgment.

[0070] Optionally, after obtaining the final abnormal conclusion, it further includes:

[0071] Record the intermediate data generated during the matching process of the abnormal data to form associated data.

[0072] Optimize the final abnormal conclusion based on the associated data.

[0073] By recording the intermediate data generated during the abnormal data analysis process to form associated data for abnormal data analysis, and optimizing the final abnormal conclusion through the management data, the accuracy of the final abnormal conclusion can be effectively improved.

[0074] Optionally, after sending the final abnormal conclusion to the in-vehicle terminal, it further includes:

[0075] Update the abnormal knowledge base based on the abnormal data and the final abnormal conclusion, and optimize the training of the abnormal large model.

[0076] Please refer to Figure 3 , since the establishment of the abnormal knowledge base is a long-term process, by storing the obtained abnormal data and abnormal conclusions in the abnormal knowledge base each time, the abnormal knowledge base can be continuously expanded, providing sufficient data support for subsequent matching of abnormal data with abnormal conclusions. By inputting the obtained abnormal data and abnormal conclusions into the abnormal large model for optimization training each time, the abnormal large model becomes more and more mature, enabling the corresponding abnormal conclusions to be quickly inferred by the abnormal large model when subsequent in-vehicle terminals have abnormalities, effectively improving the reliability and accuracy of abnormal handling, and enabling users to timely master the "physical examination report" of the in-vehicle system.

[0077] Optionally, please refer to Figure 4, the supplier can capture the final abnormal conclusion corresponding to the abnormal data in real time through the abnormal knowledge base, obtain the inference result by inputting the abnormal data into the abnormal large model, and capture the abnormal conclusion of the abnormal data from the abnormal knowledge base based on the inference result as the final abnormal conclusion. By combining the inference of the abnormal large model and the matching of the abnormal knowledge base, a more accurate abnormal handling method can be obtained, thereby optimizing and updating the in-vehicle terminal system and applications, and ensuring the stability of the in-vehicle terminal system and application programs.

[0078] Embodiment 2:

[0079] Please refer to Figure 5 , this application also proposes a system adopting the abnormal analysis method described in Embodiment 1. The system includes: a data acquisition module, a type judgment module, a knowledge base matching module, an abnormal large model inference module, and a conclusion distribution module.

[0080] The data acquisition module is used to receive the abnormal data uploaded by the in-vehicle terminal in real time. When an abnormality occurs in the in-vehicle terminal, the data acquisition service module of the in-vehicle terminal collects the abnormal data in real time and uploads it to the cloud. The data acquisition module in the cloud receives the abnormal data in real time and transmits the abnormal data to the knowledge base matching module for abnormal matching.

[0081] The type judgment module is used to judge whether there is abnormal type information in the abnormal data. If so, the abnormal data is input into the knowledge base matching module; otherwise, the abnormal data is input into the abnormal large model inference module.

[0082] The knowledge base matching module is used to match the abnormal data according to the abnormal knowledge base to obtain the final abnormal conclusion. By matching the abnormal data through the pre-constructed abnormal knowledge base in the knowledge base matching module, the final abnormal conclusion corresponding to the abnormal data can be obtained. A large amount of abnormal data and corresponding final abnormal conclusions are stored in the abnormal knowledge base, so as to cope with various multi-dimensional abnormal situations and ensure the reliability of the obtained final abnormal conclusion.

[0083] The abnormal large model inference module is used to infer the abnormal data according to the abnormal large model, and input the obtained inference result into the knowledge base matching module. The abnormal large model can mainly adopt neural networks, convolutional neural networks, long short-term memory networks, etc., without limitation. After training the pre-constructed abnormal large model with a large amount of historical abnormal data and historical abnormal conclusions, the obtained abnormal large model becomes more mature and can accurately infer the inference result corresponding to the abnormal data.

[0084] And, a conclusion issuing module, configured to issue the final exception conclusion to the vehicle-mounted terminal. By issuing the final exception conclusion to the vehicle-mounted terminal, the conclusion issuing module enables the user to obtain the final exception conclusion corresponding to the exception of the vehicle-mounted terminal in a timely manner, so that the user can take corresponding exception handling based on the final exception conclusion, thereby ensuring the stability of the vehicle-mounted terminal system. After the issuance of the exception conclusion is completed, it further includes transmitting the obtained final exception conclusion to the knowledge base matching module and the exception large model inference module, thereby enriching the exception data and the corresponding exception conclusions in the exception knowledge base matching module, enabling the exception large model to be further strengthened in training, and improving the accuracy of subsequent exception data analysis.

[0085] Optionally, the knowledge base matching module further includes: a general matching module and a keyword matching module.

[0086] Among them, the general matching module is configured to obtain multiple exception handling methods from the exception knowledge base based on the exception data; sequentially match the exception data with the multiple exception handling methods to obtain the final exception conclusion.

[0087] The keyword matching module is configured to obtain multiple exception keywords based on the exception data; perform matching based on the exception keywords and a keyword list pre-configured in the exception knowledge base to obtain the final exception conclusion.

[0088] Embodiment 3:

[0089] The present application also proposes a computer-readable storage medium, which includes:

[0090] The computer-readable storage medium stores computer-executable instructions.

[0091] When the computer-executable instructions are executed by a control processor, the exception analysis method described in Embodiment 1 is implemented.

[0092] In the computer-readable storage medium, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid-state disk (SSD)).

[0093] In summary, the anomaly analysis method proposed in this application effectively solves the technical problems of slow system operation, high resource occupancy, and system freeze and lag caused by in-vehicle computer anomaly data, and effectively improves the security and reliability of the in-vehicle computer terminal system. By analyzing the anomaly data through the constructed anomaly large model and anomaly knowledge base, the misjudgment and missed judgment of the anomaly data are effectively reduced, and the accuracy of anomaly recognition is improved. Through the constructed anomaly knowledge base, the anomaly conclusion can be quickly obtained, thereby improving the efficiency of anomaly data analysis. The establishment of the anomaly knowledge base provides effective data support for the subsequent service upgrade of suppliers and the upgrade of software and hardware. By combining the anomaly conclusion obtained from the anomaly knowledge base and the anomaly large model, the reliability of the final obtained anomaly conclusion is improved.

[0094] In several embodiments provided in the present application, it can be understood that each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code includes one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the block may occur in a different order than marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved.

[0095] When the above-mentioned functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing an electronic device to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0096] The specific embodiments described above further elaborate on the purpose, technical solution, and beneficial effects of this application. It should be understood that the above description is only for the specific embodiments of this application and is not used to limit the protection scope of this application. In particular, it is pointed out that for those skilled in the art, any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of this application should be included in the protection scope of this application.

Claims

1. An abnormal analysis method, characterized in that An abnormal large model and an abnormal knowledge base are pre-constructed in the cloud. The abnormal analysis method includes: Receiving abnormal data uploaded by the in-vehicle terminal in real time; Judging whether there is abnormal type information in the abnormal data. If so, matching the abnormal data according to the abnormal knowledge base to obtain a final abnormal conclusion; Otherwise, inputting the abnormal data into the abnormal large model for reasoning, and matching the abnormal data according to the reasoning result using the abnormal knowledge base to obtain a final abnormal conclusion; And sending the final abnormal conclusion to the in-vehicle terminal.

2. The abnormal analysis method according to claim 1, characterized in that The pre-construction of the abnormal large model and the abnormal knowledge base in the cloud further includes: Obtaining a number of historical abnormal data uploaded by the in-vehicle terminal; Performing manual analysis on the historical abnormal data to obtain historical abnormal conclusions; Training the abnormal large model based on the historical abnormal data and historical abnormal conclusions, and storing the historical abnormal data and historical abnormal conclusions in the abnormal knowledge base in a fixed file.

3. The abnormal analysis method according to claim 1, wherein Matching the abnormal data at least includes: general matching and keyword matching; Judging whether keyword matching can be performed based on the abnormal type information. If so, performing keyword matching on the abnormal data; otherwise, performing general matching on the abnormal data.

4. The abnormal analysis method according to claim 3, wherein, The general matching includes: Obtaining multiple abnormal handling methods from the abnormal knowledge base based on the abnormal data; Sequentially matching the abnormal data with the multiple abnormal handling methods to obtain the final abnormal conclusion.

5. The abnormal analysis method according to claim 3, characterized in that, The keyword matching includes: Obtaining multiple abnormal keywords based on the abnormal data; Matching based on the abnormal keywords and a keyword list pre-configured in the abnormal knowledge base to obtain the final abnormal conclusion.

6. The abnormal analysis method according to claim 5, characterized in that, After obtaining the final abnormal conclusion, it further includes: recording intermediate data generated during the abnormal data matching process to form associated data; Performing optimization processing on the final abnormal conclusion based on the associated data.

7. The abnormal analysis method according to claim 6, wherein After sending the final abnormal conclusion to the in-vehicle terminal, it further includes: Updating the abnormal knowledge base based on the abnormal data and the final abnormal conclusion, and performing optimization training on the abnormal large model.

8. A system based on the anomaly analysis method according to any one of claims 1-7, characterized in that, The system includes: A data acquisition module for receiving abnormal data uploaded by the in-vehicle terminal in real time; A type judgment module for judging whether there is abnormal type information in the abnormal data. If so, inputting the abnormal data into the knowledge base matching module; otherwise, inputting the abnormal data into the abnormal large model reasoning module; A knowledge base matching module for matching the abnormal data according to the abnormal knowledge base to obtain a final abnormal conclusion; An abnormal large model reasoning module for reasoning the abnormal data according to the abnormal large model, and inputting the obtained reasoning result into the knowledge base matching module; And a conclusion sending module for sending the final abnormal conclusion to the in-vehicle terminal.

9. The system of the anomaly analysis method according to claim 8, wherein The knowledge base matching module further includes: a general matching module for obtaining multiple abnormal handling methods from the abnormal knowledge base based on the abnormal data; sequentially matching the abnormal data with the multiple abnormal handling methods to obtain the final abnormal conclusion; A keyword matching module, configured to obtain a plurality of abnormal keywords based on the abnormal data; match the abnormal keywords with a pre-configured keyword list in an abnormal knowledge base to obtain the final abnormal conclusion.

10. A computer-readable storage medium storing computer-executable instructions, characterized in that, When the computer-executable instructions are executed by a control processor, the abnormal analysis method according to any one of claims 1-7 is implemented.

Images