Fault tree-based data stream analysis method, apparatus and device, and storage medium

Abstract

The invention discloses a data flow analysis method, device and equipment based on a fault tree and a storage medium, and the method comprises the steps: obtaining the fault data of a current vehicle, decomposing the fault data, and obtaining a data flow vector; establishing a fault tree according to the data flow vector, and associating the fault tree with a vehicle log to obtain an association result of each link; and the association result of each link is uploaded to the cloud for AI analysis, and a data stream analysis result is obtained, so that the fault association accuracy can be improved, the vehicle fault diagnosis time consumption is greatly reduced, the model iteration period is shortened, and the vehicle cost is finally reduced by reducing the cloud hardware resource requirement and the vehicle end node performance requirement during vehicle cloud communication. And the speed and efficiency of data stream analysis based on the fault tree are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle fault diagnosis, and in particular to a data flow analysis method, device, equipment and storage medium based on a fault tree. Background Art

[0002] The existing vehicle fault diagnosis method, taking CN115373369A - A vehicle fault diagnosis system and method as an example, the system includes: an on-board terminal, used to obtain at least one vehicle fault data of the vehicle, and if the preset fault cause analysis conditions are met, determine the first root cause of the vehicle fault data, and send the first root cause to the cloud; if the fault cause analysis conditions are not met, send the vehicle fault data to the cloud; the cloud, used to match the first fault diagnosis plan corresponding to the first root cause if the first root cause is received, and send the first fault diagnosis plan to the diagnostic device; if the vehicle fault data is received, obtain the second fault diagnosis plan corresponding to the vehicle fault data, and send the second fault diagnosis plan to the diagnostic device; the diagnostic device, used to diagnose the vehicle according to the first fault diagnosis plan or the second fault diagnosis plan, thereby achieving faster and more effective vehicle fault diagnosis.

[0003] However, the existing technology has the following technical defects: big data analysis is performed based on full data upload, and data flow analysis methods are not used to address specific issues, resulting in a high false alarm rate.

[0004] General software fault analysis uses one or more inputs to generate one or more outputs, dividing the system into independent modules to simplify the process. However, this method can only identify the fault in a certain module and is not easy to reflect the weight of the fault caused by the combined action of multiple templates. The Bayesian algorithm reflects the weight of each module and each input factor (sub-fault), but cannot reflect the weight of the factor that causes the fault due to integral accumulation (that is, it cannot monitor the data flow to ensure that each step meets the definition but ultimately leads to errors due to accumulated deviations), making it impossible to optimize the key items of the problem. Summary of the Invention

[0005] The main purpose of the present invention is to provide a data flow analysis method, device, equipment and storage medium based on a fault tree, aiming to solve the technical problems in the prior art that vehicle fault diagnosis cannot monitor every step of the data flow, relies on big data analysis, has serious data discreteness, has a high vehicle fault false alarm rate, takes a long time to diagnose faults, and has low fault analysis efficiency.

[0006] In a first aspect, the present invention provides a data flow analysis method based on a fault tree, the data flow analysis method based on a fault tree comprising the following steps: Acquire fault data of the current vehicle, decompose the fault data, and obtain a data flow vector; Establish a fault tree based on the data flow vector, and associate the fault tree with the vehicle log to obtain association results of each link; Upload the correlation results of each link to the cloud for AI analysis and obtain data flow analysis results.

[0007] Optionally, the acquiring fault data of the current vehicle, decomposing the fault data, and obtaining a data flow vector includes: Obtain the current vehicle fault data and use power-on, bootloader loading, kernel startup, environment initialization, executable file loading, variable initialization, application execution, driver, and output as data decomposition nodes; The fault data is decomposed according to the data decomposition nodes to obtain data flow vectors corresponding to the respective data decomposition nodes.

[0008] Optionally, establishing a fault tree according to the data flow vector and associating the fault tree with the vehicle log to obtain association results of each link includes: Obtaining a data flow path and key nodes corresponding to the data flow vector; Obtaining current fault manifestations, analyzing the current fault manifestations, and obtaining fault logical relationships corresponding to each data flow path and each key node; Establishing a fault tree according to the fault logic relationship; The fault tree is associated with the vehicle log to obtain the association results of each link.

[0009] Optionally, establishing a fault tree according to the fault logical relationship includes: Calculating the failure probability and impact level of each key node in the fault logic relationship based on the historical data vector; A weighted directed fault tree is constructed according to the fault probability, the impact level, the data flow path and the key nodes.

[0010] Optionally, after establishing the fault tree according to the fault logical relationship, the fault tree-based data flow analysis method further includes: When it is detected that the data flow vector deviates from the preset training set, a fault tree reconstruction mechanism is automatically triggered; Converting the fault tree into an equivalent Bayesian network, where nodes correspond to events and edges correspond to conditional dependencies; Inputting the abnormal dimension of the data flow vector into the Bayesian network as evidence, and updating the posterior probability of each node through a preset belief propagation algorithm; When the posterior probability of the node is less than a first preset probability threshold, removing the corresponding node from the fault tree; When the posterior probability of the node is greater than a second preset probability threshold, adding an associated child node; After adding and deleting nodes in the fault tree, logic gates are adjusted to form an updated fault tree.

[0011] Optionally, associating the fault tree with the vehicle log to obtain association results of each link includes: Establishing an associative mapping relationship between each event node in the fault tree and the corresponding data field in the vehicle log; Cross-validate the logical deduction results of the fault tree with the time series data in the vehicle log to obtain verified log data; The association weights of the event nodes of the fault tree are dynamically adjusted according to the log data to obtain the association results of each link.

[0012] Optionally, uploading the correlation results of each link to the cloud for AI analysis and obtaining data flow analysis results includes: Upload the correlation results of each link to the cloud for AI analysis to obtain the current fault type; When the fault type is abnormal power input, the vehicle power input voltage and fuse status are reported to the vehicle main control through the power management chip PMIC, and the input voltage value is recorded in the PMIC register. After the current vehicle is powered on again, the PMIC is used to power each module to obtain the power data flow analysis result; When the fault type is a display backlight circuit failure, obtaining a display data flow analysis result from the backlight control signal to the boost circuit and then to the light emitting diode (LED) backlight strip; When the fault type is abnormal LVDS / eDP signal transmission, obtain the analysis results of the drive data flow from the main control system-level chip SoC output display signal to the flexible printed circuit board FPC connection line and then to the display driver board; When the fault type is an in-vehicle network communication fault, obtain an in-vehicle network data flow analysis result of the Ethernet / CAN bus transmitting the display control signal to the gateway and then to the vehicle main control; When the fault type is an AutoSAR display service configuration error, obtaining a display service configuration data flow analysis result of a resolution / refresh rate parameter error in the ARXML configuration file of the DisplayManager; When the fault type is display driver initialization failure, obtaining a display driver initialization data flow analysis result of a function call chain and a register state of a main control SoC display controller; When the fault type is an Ethernet / CAN / LIN communication protocol stack abnormality, the bus data is acquired and sent to the controller, and then the protocol frame is parsed, and then the communication protocol stack data flow analysis results triggered by the display event are displayed; When the fault type is an operating system fault, obtain the operating system data flow analysis results from the startup of the vehicle operating system OS to the display service, then to the graphics rendering pipeline, and then to the human-machine interface HMI application; When the fault type is an application software fault, the analysis result of the application software data flow from the start of the HMI application to the calling of the graphics application program interface API, to the submission of the rendering instruction, and then to the display of the synthesized application software is obtained.

[0013] In a second aspect, to achieve the above-mentioned purpose, the present invention further proposes a data flow analysis device based on a fault tree, the data flow analysis device based on a fault tree comprising: A data decomposition module is used to obtain fault data of the current vehicle, decompose the fault data, and obtain a data flow vector; an association module, configured to establish a fault tree based on the data flow vector, and associate the fault tree with the vehicle log to obtain association results of each link; The analysis module is used to upload the correlation results of each link to the cloud for AI analysis and obtain data flow analysis results.

[0014] In a third aspect, to achieve the above-mentioned purpose, the present invention also proposes a fault tree-based data flow analysis device, wherein the fault tree-based data flow analysis device comprises: a memory, a processor, and a fault tree-based data flow analysis program stored in the memory and executable on the processor, wherein the fault tree-based data flow analysis program is configured to implement the steps of the fault tree-based data flow analysis method as described above.

[0015] In a fourth aspect, in order to achieve the above-mentioned purpose, the present invention also proposes a storage medium, on which a fault tree-based data flow analysis program is stored. When the fault tree-based data flow analysis program is executed by a processor, the steps of the fault tree-based data flow analysis method described above are implemented.

[0016] The fault tree-based data flow analysis method proposed in the present invention obtains the fault data of the current vehicle, decomposes the fault data, and obtains data flow vectors; establishes a fault tree based on the data flow vectors, and associates the fault tree with the vehicle log to obtain the association results of each link; uploads the association results of each link to the cloud for AI analysis, and obtains data flow analysis results, which can improve the accuracy of fault association, significantly reduce the time consumption of vehicle fault diagnosis, shorten the model iteration cycle, and ultimately reduce the cost of the entire vehicle by reducing the cloud hardware resource requirements and the vehicle-side node performance requirements during vehicle-cloud communication, and improve the speed and efficiency of the fault tree-based data flow analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic diagram of the device structure of the hardware operating environment involved in the embodiment of the present invention; Figure 2 This is a flow chart of a first embodiment of a data flow analysis method based on a fault tree according to the present invention; Figure 3 This is a flow chart of a second embodiment of a data flow analysis method based on a fault tree according to the present invention; Figure 4 This is a flow chart of a third embodiment of a data flow analysis method based on a fault tree according to the present invention; Figure 5 This is a functional module diagram of the first embodiment of the data flow analysis device based on the fault tree of the present invention.

[0018] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0019] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] The solution of the embodiment of the present invention is mainly: by obtaining the fault data of the current vehicle, decomposing the fault data, and obtaining a data flow vector; establishing a fault tree according to the data flow vector, and associating the fault tree with the vehicle log to obtain the association results of each link; uploading the association results of each link to the cloud for AI analysis, and obtaining data flow analysis results, which can improve the accuracy of fault association, greatly reduce the time consumption of vehicle fault diagnosis, shorten the model iteration cycle, and ultimately reduce the cost of the whole vehicle by reducing the cloud hardware resource requirements and the vehicle-side node performance requirements during vehicle-cloud communication, and improve the speed and efficiency of data flow analysis based on the fault tree, and solve the technical problems in the existing technology that vehicle fault diagnosis cannot monitor every step of the data flow, relies on big data analysis, has serious data discreteness, has a high vehicle fault false alarm rate, consumes a long time for fault diagnosis, and has low fault analysis efficiency.

[0021] Reference Figure 1 , Figure 1 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiment of the present invention.

[0022] like Figure 1 As shown, the device may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to implement communication between these components. The user interface 1003 may include a display and an input unit such as a keyboard. Optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory, such as a disk storage. The memory 1005 may also be a storage device independent of the processor 1001.

[0023] Those skilled in the art will understand that Figure 1 The device structure shown in the figure does not constitute a limitation of the device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0024] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating device, a network communication module, a user interface module, and a data flow analysis program based on a fault tree.

[0025] The device of the present invention calls the fault tree-based data flow analysis program stored in the memory 1005 through the processor 1001 and performs the following operations: Acquire fault data of the current vehicle, decompose the fault data, and obtain a data flow vector; Establish a fault tree based on the data flow vector, and associate the fault tree with the vehicle log to obtain association results of each link; Upload the correlation results of each link to the cloud for AI analysis and obtain data flow analysis results.

[0026] The device of the present invention calls the fault tree-based data flow analysis program stored in the memory 1005 through the processor 1001, and further performs the following operations: Obtain the current vehicle fault data and use power-on, bootloader loading, kernel startup, environment initialization, executable file loading, variable initialization, application execution, driver, and output as data decomposition nodes; The fault data is decomposed according to the data decomposition nodes to obtain data flow vectors corresponding to the respective data decomposition nodes.

[0027] The device of the present invention calls the fault tree-based data flow analysis program stored in the memory 1005 through the processor 1001, and further performs the following operations: Obtaining a data flow path and key nodes corresponding to the data flow vector; Obtaining current fault manifestations, analyzing the current fault manifestations, and obtaining fault logical relationships corresponding to each data flow path and each key node; Establishing a fault tree according to the fault logic relationship; The fault tree is associated with the vehicle log to obtain the association results of each link.

[0028] The device of the present invention calls the fault tree-based data flow analysis program stored in the memory 1005 through the processor 1001, and further performs the following operations: Calculating the failure probability and impact level of each key node in the fault logic relationship based on the historical data vector; A weighted directed fault tree is constructed according to the fault probability, the impact level, the data flow path and the key nodes.

[0029] The device of the present invention calls the fault tree-based data flow analysis program stored in the memory 1005 through the processor 1001, and further performs the following operations: When it is detected that the data flow vector deviates from the preset training set, a fault tree reconstruction mechanism is automatically triggered; Converting the fault tree into an equivalent Bayesian network, where nodes correspond to events and edges correspond to conditional dependencies; Inputting the abnormal dimension of the data flow vector into the Bayesian network as evidence, and updating the posterior probability of each node through a preset belief propagation algorithm; When the posterior probability of the node is less than a first preset probability threshold, removing the corresponding node from the fault tree; When the posterior probability of the node is greater than a second preset probability threshold, adding an associated child node; After adding and deleting nodes in the fault tree, logic gates are adjusted to form an updated fault tree.

[0030] The device of the present invention calls the fault tree-based data flow analysis program stored in the memory 1005 through the processor 1001, and further performs the following operations: Establishing an associative mapping relationship between each event node in the fault tree and the corresponding data field in the vehicle log; Cross-validate the logical deduction results of the fault tree with the time series data in the vehicle log to obtain verified log data; The association weights of the event nodes of the fault tree are dynamically adjusted according to the log data to obtain the association results of each link.

[0031] The device of the present invention calls the fault tree-based data flow analysis program stored in the memory 1005 through the processor 1001, and further performs the following operations: Upload the correlation results of each link to the cloud for AI analysis to obtain the current fault type; When the fault type is abnormal power input, the vehicle power input voltage and fuse status are reported to the vehicle main control through the power management chip PMIC, and the input voltage value is recorded in the PMIC register. After the current vehicle is powered on again, the PMIC is used to power each module to obtain the power data flow analysis result; When the fault type is a display backlight circuit failure, obtaining a display data flow analysis result from the backlight control signal to the boost circuit and then to the light emitting diode (LED) backlight strip; When the fault type is abnormal LVDS / eDP signal transmission, obtain the analysis results of the drive data flow from the main control system-level chip SoC output display signal to the flexible printed circuit board FPC connection line and then to the display driver board; When the fault type is an in-vehicle network communication fault, obtain an in-vehicle network data flow analysis result of the Ethernet / CAN bus transmitting the display control signal to the gateway and then to the vehicle main control; When the fault type is an AutoSAR display service configuration error, obtaining a display service configuration data flow analysis result of a resolution / refresh rate parameter error in the ARXML configuration file of the DisplayManager; When the fault type is display driver initialization failure, obtaining a display driver initialization data flow analysis result of a function call chain and a register state of a main control SoC display controller; When the fault type is an Ethernet / CAN / LIN communication protocol stack abnormality, the bus data is acquired and sent to the controller, and then the protocol frame is parsed, and then the communication protocol stack data flow analysis results triggered by the display event are displayed; When the fault type is an operating system fault, obtain the operating system data flow analysis results from the startup of the vehicle operating system OS to the display service, then to the graphics rendering pipeline, and then to the human-machine interface HMI application; When the fault type is an application software fault, the analysis result of the application software data flow from the start of the HMI application to the calling of the graphics application program interface API, to the submission of the rendering instruction, and then to the display of the synthesized application software is obtained.

[0032] Through the above-mentioned solution, this embodiment obtains the fault data of the current vehicle, decomposes the fault data, and obtains data flow vectors; establishes a fault tree based on the data flow vectors, and associates the fault tree with the vehicle log to obtain the association results of each link; uploads the association results of each link to the cloud for AI analysis, and obtains data flow analysis results. This can improve the accuracy of fault association, significantly reduce the time consumption of vehicle fault diagnosis, shorten the model iteration cycle, and ultimately reduce the cost of the entire vehicle by reducing the cloud hardware resource requirements and the vehicle-side node performance requirements during vehicle-to-cloud communication, while improving the speed and efficiency of data flow analysis based on the fault tree.

[0033] Based on the above hardware structure, an embodiment of a data flow analysis method based on a fault tree of the present invention is proposed.

[0034] Reference Figure 2 , Figure 2 This is a flow chart of the first embodiment of the data flow analysis method based on the fault tree of the present invention.

[0035] In a first embodiment, the data flow analysis method based on a fault tree includes the following steps: Step S10: Acquire fault data of the current vehicle, decompose the fault data, and obtain a data flow vector.

[0036] It should be noted that after obtaining the fault data of the current vehicle, the fault data can be decomposed to obtain data flow vectors of different processes.

[0037] Step S20: Establish a fault tree based on the data flow vector, and associate the fault tree with the vehicle log to obtain association results of each link.

[0038] It should be understood that a fault tree can be established based on the data flow vector and associated with the log, that is, the fault tree is associated with the vehicle log to obtain the association results of each link.

[0039] In specific implementations, vehicle logs can only be viewed by the root user, and cannot be modified or deleted. Root is the superuser in the operating system and has the highest permissions: it can read and modify any file (including system core files); can control all processes, hardware devices, and network configurations; can install / uninstall software and modify user permissions.

[0040] Step S30: Upload the correlation results of each link to the cloud for AI analysis and obtain the data flow analysis results.

[0041] It is understandable that after uploading the correlation results of each link to the cloud for artificial intelligence (AI) analysis, the analysis conclusion, that is, the data flow analysis result, can be obtained.

[0042] Furthermore, the step S30 specifically includes the following steps: Upload the correlation results of each link to the cloud for AI analysis to obtain the current fault type; When the fault type is abnormal power input, the vehicle power input voltage and fuse status are reported to the vehicle main control through the power management chip PMIC, and the input voltage value is recorded in the PMIC register. After the current vehicle is powered on again, the PMIC is used to power each module to obtain the power data flow analysis result; When the fault type is a display backlight circuit failure, obtaining a display data flow analysis result from the backlight control signal to the boost circuit and then to the light emitting diode (LED) backlight strip; When the fault type is abnormal LVDS / eDP signal transmission, obtain the analysis results of the drive data flow from the main control system-level chip SoC output display signal to the flexible printed circuit board FPC connection line and then to the display driver board; When the fault type is an in-vehicle network communication fault, obtain an in-vehicle network data flow analysis result of the Ethernet / CAN bus transmitting the display control signal to the gateway and then to the vehicle main control; When the fault type is an AutoSAR display service configuration error, obtaining a display service configuration data flow analysis result of a resolution / refresh rate parameter error in the ARXML configuration file of the DisplayManager; When the fault type is display driver initialization failure, obtaining a display driver initialization data flow analysis result of a function call chain and a register state of a main control SoC display controller; When the fault type is an Ethernet / CAN / LIN communication protocol stack abnormality, the bus data is acquired and sent to the controller, and then the protocol frame is parsed, and then the communication protocol stack data flow analysis results triggered by the display event are displayed; When the fault type is an operating system fault, obtain the operating system data flow analysis results from the startup of the vehicle operating system OS to the display service, then to the graphics rendering pipeline, and then to the human-machine interface HMI application; When the fault type is an application software fault, the analysis result of the application software data flow from the start of the HMI application to the calling of the graphics application program interface API, to the submission of the rendering instruction, and then to the display of the synthesized application software is obtained.

[0043] In the specific implementation, take the black screen of the car computer as an example to describe:

[0044] Data flow tracing and fault point analysis 1 Hardware failure Data flow path: Vehicle power supply → Vehicle power management module → Display backlight circuit → Low-voltage differential signaling (LVDS) / (embedded DisplayPort, eDP) signal transmission → Display panel → Touch controller (if integrated) Fault point analysis: Abnormal power input Collect data: Measure the vehicle power input voltage (usually 12V).

[0045] Check the fuse status.

[0046] Drive logic: The Power Management Integrated Circuit (PMIC) communicates with the main control unit through the Inter-Integrated Circuit (I2C) / System Management Bus (SMBus) to report the voltage status.

[0047] Intermediate value storage: Input voltage value recorded in the PMIC register.

[0048] Sequential logic: After the vehicle is powered on, the PMIC must complete power supply to each module (such as the main control, display, and audio) within 200ms.

[0049] Display backlight circuit failure Data flow: Backlight control signal → boost circuit → light-emitting diode (LED) backlight strip.

[0050] LVDS / eDP signal transmission is abnormal Data flow: The main control system-on-a-chip (SoC) outputs the display signal → Flexible Printed Circuit (FPC) connection line → Display driver board.

[0051] In-vehicle network communication failure Data flow: Ethernet / CAN bus transmission display control signal → gateway → vehicle main control.

[0052] 2. Low-level firmware / driver failure Data flow path: Automotive Open System Architecture (AutoSAR) service startup → Display driver initialization → Ethernet / CAN / Local Interconnect Network (LIN) communication protocol stack loading → Hardware Abstraction Layer (HAL) interaction Fault point analysis: AutoSAR shows service configuration error Key configuration: The resolution / refresh rate parameters in the DisplayManager's AutoSar Extensible Markup Language (ARXML) configuration file are incorrect. (ARXML is an XML (Extensible Markup Language)-based format used to describe system design, configuration, and interface information in AutoSar (Automotive Open System Architecture) projects.)

[0053] Display driver initialization failed Function call chain (taking QNX as an example): / / Display driver loading process devc-display -d fb0 -m lvds / / Start the display driver io-blk -d / dev / fb0 / / frame buffer mapping Key registers: The main SoC displays the register status of the controller.

[0054] Ethernet / CAN / LIN communication protocol stack abnormality Data flow: Bus data → Controller → Protocol stack analysis → Display event triggering.

[0055] Intermediate value storage: Receive data frames in the buffer.

[0056] 3 Operating system failure Data flow path: In-vehicle OS startup → Display service → Graphics rendering pipeline → Human-Machine Interface (HMI) application Fault point analysis: Display service crash Key processes: Android Automotive: surfaceflinger process status.

[0057] QNX: screen service status.

[0058] File system corruption Key documents: / system / bin / surfaceflinger (Android display service binary).

[0059] / etc / display.conf (display configuration file).

[0060] 4 Application software failure Data flow path: HMI application starts → calls the graphics application programming interface (API) → submits rendering instructions → displays the composite Fault point analysis: HMI application crashes Function call trace: Android: Stack trace when launching the HMI application with am start.

[0061] Intermediate value storage: Apply rendered frame data from memory.

[0062] Resource usage conflicts Typical scenarios: Navigation and multimedia applications simultaneously occupy the graphics processing unit (GPU) resources, causing video memory overflow.

[0063] It should be noted that the above is a fault detection process between modules and within modules for a single part. However, they are all discrete detection and process records, and there is no necessary connection between them. Moreover, the detected fault is not necessarily strongly correlated with the fault that can ultimately be presented to the customer. When a customer-perceivable fault occurs, each module uploads the internal detection results for the N min time period before and after the fault in the form of a log report (encrypted storage, root privileges can only be viewed, not modified or deleted).

[0064] The cloud uses the various parameters, interruptions, function calls, etc. recorded in the above process logs (these contents can be presented in a fixed format) as all related items on the vertical axis and time as the horizontal axis. The cloud AI can extract relevant factors in the logs based on the customer-perceivable faults, and use the fault time that the end customer can perceive as a reference point to analyze what happened in the entire vehicle data flow system at that point in time.

[0065] Through the above-mentioned solution, this embodiment obtains the fault data of the current vehicle, decomposes the fault data, and obtains data flow vectors; establishes a fault tree based on the data flow vectors, and associates the fault tree with the vehicle log to obtain the association results of each link; uploads the association results of each link to the cloud for AI analysis, and obtains data flow analysis results. This can improve the accuracy of fault association, significantly reduce the time consumption of vehicle fault diagnosis, shorten the model iteration cycle, and ultimately reduce the cost of the entire vehicle by reducing the cloud hardware resource requirements and the vehicle-side node performance requirements during vehicle-to-cloud communication, while improving the speed and efficiency of data flow analysis based on the fault tree.

[0066] Furthermore, Figure 3 FIG. 1 is a flow chart of a second embodiment of a data flow analysis method based on a fault tree according to the present invention. Figure 3 As shown, a second embodiment of the data flow analysis method based on a fault tree of the present invention is proposed based on the first embodiment. In this embodiment, step S10 specifically includes the following steps: Step S11, obtain the fault data of the current vehicle, and use power-on, bootloader loading, kernel startup, environment initialization, executable file loading, variable initialization, application execution, driving and output as data decomposition nodes.

[0067] It should be noted that after obtaining the fault data of the current vehicle, power-on, loading the bootloader, starting the kernel, initializing the environment, loading the executable file, initializing the variables, executing the application, driving and outputting can be used as data decomposition nodes.

[0068] Step S12: Decompose the fault data according to the data decomposition nodes to obtain data flow vectors corresponding to the data decomposition nodes.

[0069] It can be understood that the fault data is decomposed according to the data decomposition nodes, thereby obtaining the data flow vector corresponding to each data decomposition node.

[0070] In the specific implementation, all steps can be decomposed and common points can be extracted with data as a reference. Each module contains all or part of the process: power-on - loading the bootloader - starting the kernel - initializing the environment - loading the executable file - initializing variables - executing the application (function call & stack management & interrupt processing & releasing resources, etc.) - driving - output. Based on the data flow vector, the input and output of each step, as well as the function interface called in the middle, the processed events, the stack detection, etc. are recorded, and then the results of each link are uploaded to the cloud for AI analysis and conclusions; thereby, the accuracy can be further improved and the requirements for cloud hardware resources and the performance requirements of the vehicle-side nodes during vehicle-cloud communication can be reduced.

[0071] Through the above scheme, this embodiment obtains the fault data of the current vehicle and uses power-on, bootloader loading, kernel startup, environment initialization, executable file loading, variable initialization, application execution, driver, and output as data decomposition nodes; decomposes the fault data according to the data decomposition nodes to obtain data flow vectors corresponding to each data decomposition node. This can quickly obtain the data flow vectors of each node, improve the fault correlation accuracy, significantly reduce the time consumption of vehicle fault diagnosis, shorten the model iteration cycle, and improve the speed and efficiency of fault tree-based data flow analysis.

[0072] Furthermore, Figure 4 FIG. 1 is a flow chart of a third embodiment of a data flow analysis method based on a fault tree according to the present invention. Figure 4 As shown, a third embodiment of the data flow analysis method based on a fault tree of the present invention is proposed based on the first embodiment. In this embodiment, step S20 specifically includes the following steps: Step S21: Obtain the data flow path and key nodes corresponding to the data flow vector.

[0073] It should be noted that the data flow vectors correspond to corresponding data flow paths and key nodes.

[0074] Step S22: Obtain current fault manifestations, analyze the current fault manifestations, and obtain fault logical relationships corresponding to each data flow path and each key node.

[0075] It is understandable that after obtaining the current fault manifestation, the current fault manifestation can be used as the top event for corresponding analysis, thereby obtaining the fault logical relationship corresponding to each data flow path and each key node.

[0076] Step S23: Establish a fault tree according to the fault logic relationship.

[0077] It should be understood that after obtaining the logical relationship of the fault, a corresponding fault tree can be constructed. A fault tree is a tree structure that connects fault events through logical symbols (AND gates, OR gates, etc.) and is used to infer the underlying causes from the top event (such as system failure).

[0078] Furthermore, the step S23 specifically includes the following steps: Calculating the failure probability and impact level of each key node in the fault logic relationship based on the historical data vector; A weighted directed fault tree is constructed according to the fault probability, the impact level, the data flow path and the key nodes.

[0079] It is understandable that the failure probability and impact level of each key node in the fault logic relationship can be calculated based on the historical data vector; and then a weighted directed fault tree can be constructed based on the failure probability, the impact level, the data flow path and the key nodes.

[0080] In specific implementations, a fault tree can be established based on data flow vectors. Data features need to be mapped into a logical structure to construct a causal chain from top events to bottom events. That is, key parameters of the data flow (such as voltage, frequency, and packet loss rate) and their threshold ranges are extracted to form a multidimensional vector space. The data flow vector features are defined. Based on historical vector data, the conditional probability of each node can be calculated (such as P (transport layer fault | vector mutation) = 0.68), forming a weighted directed graph, that is, building a weighted directed fault tree.

[0081] Furthermore, after step S23, the data flow analysis method based on the fault tree further includes the following steps: When it is detected that the data flow vector deviates from the preset training set, a fault tree reconstruction mechanism is automatically triggered; Converting the fault tree into an equivalent Bayesian network, where nodes correspond to events and edges correspond to conditional dependencies; Inputting the abnormal dimension of the data flow vector into the Bayesian network as evidence, and updating the posterior probability of each node through a preset belief propagation algorithm; When the posterior probability of the node is less than a first preset probability threshold, removing the corresponding node from the fault tree; When the posterior probability of the node is greater than a second preset probability threshold, adding an associated child node; After adding and deleting nodes in the fault tree, logic gates are adjusted to form an updated fault tree.

[0082] In the specific implementation, when the real-time data flow vector deviates from the training set, the fault tree reconstruction will be automatically triggered. The node probability will be corrected through the Bayesian network, and the fault tree will be converted into an equivalent Bayesian network (BN). The nodes correspond to events and the edges represent conditional dependencies. When a vector anomaly is detected, the anomaly dimension is input into the BN as evidence E (such as E={x1>3σ}). The posterior probability of each node is updated through the belief propagation algorithm. If the posterior probability of a node P (H|E) is less than 0.05, it is temporarily removed from the fault tree; if P (H|E) is greater than 0.7, a new associated child node is added. After the node addition and deletion is completed, the logic gate can be adjusted. That is, according to the conditional independence test, the original "AND gate" is replaced by the original "AND gate". Adjust to an "OR gate" (if multiple independent fault paths are found). Furthermore, the Expectation-Maximization (EM) algorithm can be used to iteratively optimize the BN parameters, that is, the EM algorithm can be used to iteratively optimize the node probability parameters. At the same time, incremental learning architecture and edge computing are deployed to achieve real-time response, and the dual-tree verification mechanism is used to ensure diagnostic reliability. Ultimately, an adaptive diagnostic model is formed to improve the accuracy and efficiency of complex system fault diagnosis.

[0083] Step S24: Correlate the fault tree with the vehicle log to obtain correlation results of each link.

[0084] It should be understood that after associating the fault tree with the vehicle log, the management results of each link can be obtained.

[0085] Furthermore, the step S24 specifically includes the following steps: Establishing an associative mapping relationship between each event node in the fault tree and the corresponding data field in the vehicle log; Cross-validate the logical deduction results of the fault tree with the time series data in the vehicle log to obtain verified log data; The association weights of the event nodes of the fault tree are dynamically adjusted according to the log data to obtain the association results of each link.

[0086] It can be understood that an association mapping relationship is established between each event node in the fault tree and the corresponding data field in the vehicle log, and then the logical deduction result of the fault tree is cross-validated with the time series data in the vehicle log to obtain the verified log data; the association weights of each event node in the fault tree are dynamically adjusted according to the log data to obtain the association results of each link.

[0087] In a specific implementation, a mapping relationship can be established between each event node in the fault tree (such as "sensor data anomaly" and "Controller Area Network (CAN) bus communication failure") and the corresponding data field in the vehicle log (such as on-board diagnostic system (OBD) fault code, (electronic control unit, ECU) operating parameters, CAN bus error frame count). For example, when the "engine speed sensor failure" node in the fault tree is activated, the system automatically retrieves data such as the speed signal jump record and the number of checksum errors in the engine control unit log as supporting evidence. Secondly, by using timestamp alignment and multi-source data fusion technology, the logical deduction results of the fault tree can be cross-validated with the time series data in the log. For example, by analyzing the voltage and temperature curves in the battery management system log, the triggering probability of the "power battery thermal runaway risk" node in the fault tree can be verified.

[0088] In addition, by combining Bayesian networks with machine learning algorithms, historical log data can be trained to dynamically adjust the association weights of each node in the fault tree. For example, when a certain ECU restart record frequently appears in the log, the diagnostic priority of the "controller program crash" node in the fault tree is automatically increased.

[0089] Ultimately, the visual interface displays the correlation results of each link of the fault tree and log data in real time, helping technicians quickly locate the root cause of the fault and provide maintenance decision-making suggestions, which can significantly shorten the troubleshooting time and improve vehicle maintenance efficiency and safety.

[0090] This embodiment adopts the above scheme, by obtaining the data flow path and key nodes corresponding to the data flow vector; obtaining the current fault manifestation, analyzing the current fault manifestation, and obtaining the fault logical relationship corresponding to each data flow path and each key node; establishing a fault tree according to the fault logical relationship; correlating the fault tree with the vehicle log to obtain the correlation results of each link; it can improve the fault correlation accuracy, significantly reduce the time consumption of vehicle fault diagnosis, shorten the model iteration cycle, and ultimately reduce the cost of the entire vehicle by reducing the cloud hardware resource requirements and the vehicle-side node performance requirements during vehicle-cloud communication, and improve the speed and efficiency of data flow analysis based on the fault tree.

[0091] Accordingly, the present invention further provides a data flow analysis device based on a fault tree.

[0092] Reference Figure 5 , Figure 5 This is a functional module diagram of the first embodiment of the data flow analysis device based on the fault tree of the present invention.

[0093] In a first embodiment of the data flow analysis device based on a fault tree of the present invention, the data flow analysis device based on a fault tree comprises: The data decomposition module 10 is used to obtain fault data of the current vehicle, decompose the fault data, and obtain a data flow vector.

[0094] The association module 20 is used to establish a fault tree according to the data flow vector, and associate the fault tree with the vehicle log to obtain association results of each link.

[0095] The analysis module 30 is used to upload the correlation results of each link to the cloud for AI analysis and obtain data flow analysis results.

[0096] The data decomposition module 10 is also used to obtain the fault data of the current vehicle, and uses power-on, bootloader loading, kernel startup, environment initialization, executable file loading, variable initialization, application execution, driving and output as data decomposition nodes; the fault data is decomposed according to the data decomposition nodes to obtain the data flow vector corresponding to each data decomposition node.

[0097] The association module 20 is further configured to obtain the data flow paths and key nodes corresponding to the data flow vector; obtain the current fault manifestation, analyze the current fault manifestation, and obtain the fault logical relationship corresponding to each data flow path and each key node; establish a fault tree based on the fault logical relationship; and associate the fault tree with the vehicle log to obtain the association results of each link.

[0098] The association module 20 is further configured to calculate the failure probability and impact level of each key node in the fault logic relationship based on the historical data vector; and construct a weighted directed fault tree based on the failure probability, the impact level, the data flow path, and the key nodes.

[0099] The association module 20 is further configured to automatically trigger a fault tree reconstruction mechanism when it is detected that the data flow vector deviates from a preset training set; convert the fault tree into an equivalent Bayesian network, where nodes correspond to events and edges correspond to conditional dependencies; input the abnormal dimension of the data flow vector as evidence into the Bayesian network, and update the posterior probability of each node through a preset belief propagation algorithm; remove the corresponding node from the fault tree when the posterior probability of the node is less than a first preset probability threshold; add an associated child node when the posterior probability of the node is greater than a second preset probability threshold; and perform logic gate adjustment after the addition and deletion of nodes in the fault tree are completed to form an updated fault tree.

[0100] The association module 20 is further configured to establish an association mapping relationship between each event node in the fault tree and a corresponding data field in the vehicle log; cross-validate the logical deduction results of the fault tree with the time series data in the vehicle log to obtain verified log data; and dynamically adjust the association weights of each event node in the fault tree according to the log data to obtain association results for each link.

[0101] The analysis module 30 is also used to upload the correlation results of each link to the cloud for AI analysis to obtain the current fault type; when the fault type is power input abnormality, the vehicle power input voltage and fuse status are reported to the vehicle main control through the power management chip PMIC, and the input voltage value is recorded in the PMIC register. After the current vehicle is powered on again, the PMIC is used to power each module to obtain the power data flow analysis result; when the fault type is display backlight circuit failure, the backlight control signal is obtained to the boost circuit, and then to the light emitting diode LED backlight strip display data flow analysis result; when the fault type is LVDS / eDP signal transmission abnormality, the main control system chip SoC output display signal to the flexible printed circuit board FPC connection line, and then to the display driver board drive data flow analysis result is obtained; when the fault type is vehicle network communication failure, the Ethernet / CAN bus is obtained to transmit the display control signal to the gateway, and then to the vehicle main control vehicle network data flow analysis result. When the fault type is an AutoSAR display service configuration error, the display service configuration data flow analysis result of the resolution / refresh rate parameter error in the ARXML configuration file of the DisplayManager is obtained; when the fault type is a display driver initialization failure, the display driver initialization data flow analysis result of the function call chain and the register status of the main SoC display controller is obtained; when the fault type is an Ethernet / CAN / LIN communication protocol stack exception, the communication protocol stack data flow analysis result of the bus data to the controller, the protocol frame parsing, and the display event triggering is obtained; when the fault type is an operating system fault, the operating system data flow analysis result of the vehicle operating system OS startup to the display service, the graphics rendering pipeline, and the human-machine interface HMI application is obtained; when the fault type is an application software fault, the application software data flow analysis result of the HMI application startup to the calling of the graphics application program interface API, the submission of rendering instructions, and the display synthesis is obtained.

[0102] The steps implemented by the functional modules of the fault tree-based data flow analysis device can refer to the various embodiments of the fault tree-based data flow analysis method of the present invention, and will not be described in detail here.

[0103] In addition, an embodiment of the present invention further provides a storage medium, on which a fault tree-based data flow analysis program is stored. When the fault tree-based data flow analysis program is executed by a processor, the following operations are implemented: Acquire fault data of the current vehicle, decompose the fault data, and obtain a data flow vector; Establish a fault tree based on the data flow vector, and associate the fault tree with the vehicle log to obtain association results of each link; Upload the correlation results of each link to the cloud for AI analysis and obtain data flow analysis results.

[0104] Furthermore, when the fault tree-based data flow analysis program is executed by the processor, the following operations are also implemented: Obtain the current vehicle fault data and use power-on, bootloader loading, kernel startup, environment initialization, executable file loading, variable initialization, application execution, driver, and output as data decomposition nodes; The fault data is decomposed according to the data decomposition nodes to obtain data flow vectors corresponding to the respective data decomposition nodes.

[0105] Furthermore, when the fault tree-based data flow analysis program is executed by the processor, the following operations are also implemented: Obtaining a data flow path and key nodes corresponding to the data flow vector; Obtaining current fault manifestations, analyzing the current fault manifestations, and obtaining fault logical relationships corresponding to each data flow path and each key node; Establishing a fault tree according to the fault logic relationship; The fault tree is associated with the vehicle log to obtain the association results of each link.

[0106] Furthermore, when the fault tree-based data flow analysis program is executed by the processor, the following operations are also implemented: Calculating the failure probability and impact level of each key node in the fault logic relationship based on the historical data vector; A weighted directed fault tree is constructed according to the fault probability, the impact level, the data flow path and the key nodes.

[0107] Furthermore, when the fault tree-based data flow analysis program is executed by the processor, the following operations are also implemented: When it is detected that the data flow vector deviates from the preset training set, a fault tree reconstruction mechanism is automatically triggered; Converting the fault tree into an equivalent Bayesian network, where nodes correspond to events and edges correspond to conditional dependencies; Inputting the abnormal dimension of the data flow vector into the Bayesian network as evidence, and updating the posterior probability of each node through a preset belief propagation algorithm; When the posterior probability of the node is less than a first preset probability threshold, removing the corresponding node from the fault tree; When the posterior probability of the node is greater than a second preset probability threshold, adding an associated child node; After adding and deleting nodes in the fault tree, logic gates are adjusted to form an updated fault tree.

[0108] Furthermore, when the fault tree-based data flow analysis program is executed by the processor, the following operations are also implemented: Establishing an associative mapping relationship between each event node in the fault tree and the corresponding data field in the vehicle log; Cross-validate the logical deduction results of the fault tree with the time series data in the vehicle log to obtain verified log data; The association weights of the event nodes of the fault tree are dynamically adjusted according to the log data to obtain the association results of each link.

[0109] Furthermore, when the fault tree-based data flow analysis program is executed by the processor, the following operations are also implemented: Upload the correlation results of each link to the cloud for AI analysis to obtain the current fault type; When the fault type is abnormal power input, the vehicle power input voltage and fuse status are reported to the vehicle main control through the power management chip PMIC, and the input voltage value is recorded in the PMIC register. After the current vehicle is powered on again, the PMIC is used to power each module to obtain the power data flow analysis result; When the fault type is a display backlight circuit failure, obtaining a display data flow analysis result from the backlight control signal to the boost circuit and then to the light emitting diode (LED) backlight strip; When the fault type is abnormal LVDS / eDP signal transmission, obtain the analysis results of the drive data flow from the main control system-level chip SoC output display signal to the flexible printed circuit board FPC connection line and then to the display driver board; When the fault type is an in-vehicle network communication fault, obtain an in-vehicle network data flow analysis result of the Ethernet / CAN bus transmitting the display control signal to the gateway and then to the vehicle main control; When the fault type is an AutoSAR display service configuration error, obtaining a display service configuration data flow analysis result of a resolution / refresh rate parameter error in the ARXML configuration file of the DisplayManager; When the fault type is display driver initialization failure, obtaining a display driver initialization data flow analysis result of a function call chain and a register state of a main control SoC display controller; When the fault type is an Ethernet / CAN / LIN communication protocol stack abnormality, the bus data is acquired and sent to the controller, and then the protocol frame is parsed, and then the communication protocol stack data flow analysis results triggered by the display event are displayed; When the fault type is an operating system fault, obtain the operating system data flow analysis results from the startup of the vehicle operating system OS to the display service, then to the graphics rendering pipeline, and then to the human-machine interface HMI application; When the fault type is an application software fault, the analysis result of the application software data flow from the start of the HMI application to the calling of the graphics application program interface API, to the submission of the rendering instruction, and then to the display of the synthesized application software is obtained.

[0110] Those skilled in the art will understand that all or part of the steps in the above-mentioned implementation methods can be implemented by instructing related hardware through a program. The program is stored in a storage medium and includes a number of instructions for enabling a device (which may be a single-chip microcomputer, chip, etc.) or a processor to execute all or part of the steps of the method described in each embodiment of the present application; and the aforementioned storage medium is a computer-readable storage medium, including: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0111] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0112] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0113] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A data flow analysis method based on a fault tree, characterized in that: The data flow analysis method based on the fault tree includes: Acquire fault data of the current vehicle, decompose the fault data, and obtain a data flow vector; Establish a fault tree based on the data flow vector, and associate the fault tree with the vehicle log to obtain association results of each link; Upload the correlation results of each link to the cloud for AI analysis and obtain data flow analysis results.

2. The data flow analysis method based on the fault tree according to claim 1, characterized in that: The acquiring fault data of the current vehicle, decomposing the fault data, and obtaining a data flow vector includes: Obtain the current vehicle fault data and use power-on, bootloader loading, kernel startup, environment initialization, executable file loading, variable initialization, application execution, driver, and output as data decomposition nodes; The fault data is decomposed according to the data decomposition nodes to obtain data flow vectors corresponding to the respective data decomposition nodes.

3. The data flow analysis method based on fault tree according to claim 1, characterized in that: The step of establishing a fault tree based on the data flow vector and associating the fault tree with the vehicle log to obtain association results of each link includes: Obtaining a data flow path and key nodes corresponding to the data flow vector; Obtaining current fault manifestations, analyzing the current fault manifestations, and obtaining fault logical relationships corresponding to each data flow path and each key node; Establishing a fault tree according to the fault logic relationship; The fault tree is associated with the vehicle log to obtain the association results of each link.

4. The data flow analysis method based on fault tree according to claim 3, characterized in that: The establishing of a fault tree according to the fault logical relationship includes: Calculating the failure probability and impact level of each key node in the fault logic relationship based on the historical data vector; A weighted directed fault tree is constructed according to the fault probability, the impact level, the data flow path and the key nodes.

5. The data flow analysis method based on fault tree according to claim 3, characterized in that: After establishing the fault tree according to the fault logical relationship, the data flow analysis method based on the fault tree further includes: When it is detected that the data flow vector deviates from the preset training set, a fault tree reconstruction mechanism is automatically triggered; Converting the fault tree into an equivalent Bayesian network, where nodes correspond to events and edges correspond to conditional dependencies; Inputting the abnormal dimension of the data flow vector into the Bayesian network as evidence, and updating the posterior probability of each node through a preset belief propagation algorithm; When the posterior probability of the node is less than a first preset probability threshold, removing the corresponding node from the fault tree; When the posterior probability of the node is greater than a second preset probability threshold, adding an associated child node; After adding and deleting nodes in the fault tree, logic gates are adjusted to form an updated fault tree.

6. The data flow analysis method based on fault tree according to claim 3, characterized in that: The step of associating the fault tree with the vehicle log to obtain the association results of each link includes: Establishing an associative mapping relationship between each event node in the fault tree and the corresponding data field in the vehicle log; Cross-validate the logical deduction results of the fault tree with the time series data in the vehicle log to obtain verified log data; The association weights of the event nodes of the fault tree are dynamically adjusted according to the log data to obtain the association results of each link.

7. The data flow analysis method based on fault tree according to claim 1, characterized in that: The above mentioned process of uploading the correlation results of each link to the cloud for AI analysis and obtaining data flow analysis results includes: Upload the correlation results of each link to the cloud for AI analysis to obtain the current fault type; When the fault type is abnormal power input, the vehicle power input voltage and fuse status are reported to the vehicle main control through the power management chip PMIC, and the input voltage value is recorded in the PMIC register. After the current vehicle is powered on again, the PMIC is used to power each module to obtain the power data flow analysis result; When the fault type is a display backlight circuit failure, obtaining a display data flow analysis result from the backlight control signal to the boost circuit and then to the light emitting diode (LED) backlight strip; When the fault type is abnormal LVDS / eDP signal transmission, obtain the analysis results of the drive data flow from the main control system-level chip SoC output display signal to the flexible printed circuit board FPC connection line and then to the display driver board; When the fault type is an in-vehicle network communication fault, obtain an in-vehicle network data flow analysis result of the Ethernet / CAN bus transmitting the display control signal to the gateway and then to the vehicle main control; When the fault type is an AutoSAR display service configuration error, obtaining a display service configuration data flow analysis result of a resolution / refresh rate parameter error in the ARXML configuration file of the DisplayManager; When the fault type is display driver initialization failure, obtaining a display driver initialization data flow analysis result of a function call chain and a register state of a main control SoC display controller; When the fault type is an Ethernet / CAN / LIN communication protocol stack abnormality, the bus data is acquired and sent to the controller, and then the protocol frame is parsed, and then the communication protocol stack data flow analysis results triggered by the display event are displayed; When the fault type is an operating system fault, obtain the operating system data flow analysis results from the startup of the vehicle operating system OS to the display service, then to the graphics rendering pipeline, and then to the human-machine interface HMI application; When the fault type is an application software fault, the analysis result of the application software data flow from the start of the HMI application to the calling of the graphics application program interface API, to the submission of the rendering instruction, and then to the display of the synthesized application software is obtained.

8. A data flow analysis device based on a fault tree, characterized in that: The data flow analysis device based on the fault tree comprises: A data decomposition module is used to obtain fault data of the current vehicle, decompose the fault data, and obtain a data flow vector; an association module, configured to establish a fault tree based on the data flow vector, and associate the fault tree with the vehicle log to obtain association results of each link; The analysis module is used to upload the correlation results of each link to the cloud for AI analysis and obtain data flow analysis results.

9. A data flow analysis device based on a fault tree, characterized in that: The fault tree-based data flow analysis device includes: a memory, a processor, and a fault tree-based data flow analysis program stored in the memory and executable on the processor, wherein the fault tree-based data flow analysis program is configured to implement the steps of the fault tree-based data flow analysis method according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium stores a fault tree-based data flow analysis program, which, when executed by a processor, implements the steps of the fault tree-based data flow analysis method according to any one of claims 1 to 7.

Images