Automotive Fault Diagnosis Method and Related Equipment Based on Dual Quantization Logic Tree
By constructing a dual-quantification logic tree based on the number of fault codes and the number of suspicious parts, the problem of lack of basis for fault diagnosis path selection and incomplete investigation of non-component faults in the existing technology is solved, thereby improving the accuracy and standardization of fault diagnosis.
Patent Information
- Application Number
- CN202611494307.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-23
- Publication Date
- 2026-11-17
AI Technical Summary
Existing automotive fault diagnosis technologies fail to construct a classification decision framework using both the number of fault codes and the number of suspicious parts as dual quantification axes. They lack a priority investigation mechanism for non-part faults, do not embed the principle of detection priority ranking into the diagnostic logic, and lack standardized steps for graded processing of detection results.
By acquiring fault codes and fault symptoms, a dual-quantization logic tree is constructed. The number of fault codes and the number of suspicious parts are cross-matched to generate detection sorting rules. The measured values are compared with standard values to execute hierarchical processing instructions. Pre-fault judgment is introduced to filter out non-hardware damage factors, avoiding disassembly and inspection without verification.
It improved the accuracy of fault diagnosis and the standardization of business operations, significantly reduced unnecessary mechanical disassembly work, and improved the troubleshooting efficiency of the intelligent diagnostic system.
Smart Images

Figure CN123113833A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of automotive fault diagnosis technology, and in particular to an automotive fault diagnosis method and related equipment based on a dual quantization logic tree. Background Technology
[0002] Currently, the main technical approaches in automotive fault diagnosis include the following: First, data-driven diagnostic methods, which focus on using historical fault data for model training and rely on the quantity and quality of the dataset; second, fault tree analysis based on physical models, which expands logic gates from top to bottom to build a complete physical model; third, single-dimensional diagnostics, which use the number or type of fault codes as a single criterion for judgment; and fourth, heuristic diagnostics based on expert experience, which relies on the personal experience of repair technicians. Existing technologies have the following shortcomings: they fail to simultaneously use both "number of fault codes" and "number of suspicious parts" as dual quantification axes to construct a classification decision framework, resulting in a lack of basis for path selection in complex fault scenarios; they lack a priority investigation mechanism for non-part faults, confusing programming and coding issues with hardware problems; they fail to embed the principle of detection priority ranking into the diagnostic logic; and they lack standardized diagnostic judgment steps for comparing and classifying detection results with standard values. Summary of the Invention
[0003] Based on this, it is necessary to address the technical problems in existing technologies, such as the lack of basis for diagnostic path selection, the failure to pre-investigate non-component faults, and the lack of closed-loop judgment mechanisms. Therefore, a vehicle fault diagnosis method and related equipment based on dual quantization logic trees are proposed.
[0004] Firstly, a vehicle fault diagnosis method based on a dual-quantization logic tree is provided, the method comprising: The vehicle diagnostic interface and sensors are used to obtain the current vehicle's fault code set and fault phenomenon set. The fault code set and the fault phenomenon set are matched to obtain the suspected component set. The fault code set and the suspected component set are quantified to obtain the initial number of fault codes and the initial number of suspected components, respectively. Based on the fault code set, preliminary fault judgment is performed to obtain the updated number of fault codes and the updated number of suspicious parts; Based on the updated number of fault codes and the updated number of suspicious parts, cross-matching is performed in a preset two-dimensional four-quadrant classification matrix to obtain the target diagnostic logic tree type corresponding to the current fault. The target diagnostic logic tree type includes a first type with a single fault code and a single suspicious part, a second type with a single fault code and multiple suspicious parts, a third type with multiple fault codes and a single suspicious part, and a fourth type with multiple fault codes and multiple suspicious parts. The two-dimensional four-quadrant classification matrix is constructed based on the fault code quantity dimension and the suspicious part quantity dimension. Based on the target diagnostic logic tree type, the corresponding component detection sorting rules are obtained to determine the detection order of each suspicious component; According to the detection order, the target suspicious parts in the suspicious parts set are functionally measured in sequence to obtain the measured values. The measured values are compared with the preset design standard values to obtain the target deviation degree. The corresponding processing instructions are executed in sequence according to the priority order of the quantitative grading interval to which the target deviation degree belongs, until the fault is confirmed to be eliminated.
[0005] Secondly, a vehicle fault diagnosis device based on a dual quantization logic tree is provided, the device comprising: The information acquisition and quantification module is used to acquire the current vehicle's fault code set and fault phenomenon set through the vehicle diagnostic interface and sensors, match the fault code set and the fault phenomenon set to obtain a set of suspected parts, and quantify the fault code set and the set of suspected parts to obtain the initial number of fault codes and the initial number of suspected parts, respectively. The pre-judgment module is used to perform pre-fault judgment based on the fault code set to obtain the updated number of fault codes and the updated number of suspicious parts; The classification and matching module is used to perform cross-matching in a preset two-dimensional four-quadrant classification matrix based on the updated number of fault codes and the updated number of suspicious parts to obtain the target diagnostic logic tree type corresponding to the current fault. The sorting rule determination module is used to obtain the corresponding component detection sorting rule based on the target diagnostic logic tree type to determine the detection order of each suspicious component; The closed-loop judgment and execution module is used to perform functional measurements on the target suspicious parts in the suspicious parts set in the order of detection to obtain measured values, compare the measured values with the preset design standard values to obtain the target deviation degree, and execute the corresponding processing instructions in the order of priority corresponding to the quantitative grading interval to which the target deviation degree belongs, until the fault is confirmed to be eliminated.
[0006] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described vehicle fault diagnosis method based on a dual-quantization logic tree.
[0007] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described vehicle fault diagnosis method based on a dual-quantization logic tree.
[0008] Compared with existing technologies, this invention quantifies fault phenomena and fault codes by acquiring them. Based on this, it filters non-hardware damage factors through pre-fault judgment, and then uses a two-dimensional matrix cross-matching to determine the diagnostic logic tree type. Subsequently, it generates corresponding detection sorting rules and compares the measured values with standard values according to this order to execute hierarchical processing instructions. This invention introduces a pre-filtering and screening mechanism to avoid disassembly and inspection without verifying software or maintenance issues. By constructing a two-dimensional classification matrix composed of the number of fault codes and the number of suspicious components, it divides multi-state concurrent faults into a standard processing framework. Then, it generates a detection order based on classification differences, and finally executes a hierarchical closed loop of repair, adjustment, or replacement based on standard value deviation, thereby improving the accuracy of fault diagnosis and location and the standardization of business execution. Attached Figure Description
[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0010] in: Figure 1 This is an application environment diagram of an automotive fault diagnosis method based on a dual-quantization logic tree in one embodiment; Figure 2 Here is a flowchart of a diagnostic method in one embodiment; Figure 3 This is a structural block diagram of a diagnostic device in one embodiment; Figure 4 This is a structural block diagram of a computer device in one embodiment. Detailed Implementation
[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0012] The vehicle fault diagnosis method based on dual quantization logic tree provided in this invention can be applied to, for example... Figure 1The application environment includes an automotive diagnostic terminal and a server, which communicate via a network. The automotive diagnostic terminal establishes a low-level data connection with the vehicle's various electronic control units through a diagnostic physical interface (such as OBD / CAN bus). In this application environment, the system automatically collects the vehicle's fault codes and symptoms, and coordinates with the server's business database to match a set of suspicious components. Subsequently, the system's underlying layer performs quantitative sorting calculations based on a logic tree classification mechanism, directly generating and issuing a priority-ordered detection execution sequence. Further, the system automatically reads the sensor measured values of each component according to the execution sequence, quantitatively compares them with the preset design standard values in the database, and then automatically generates and outputs targeted hierarchical processing instructions, forming a system-level measurement and verification closed loop. This solution uses computer equipment as the execution subject, and through program-driven pre-filtering and multi-dimensional matrix matching algorithms, the system automatically locks the components with high weights for fixed-point detection and positioning, completely eliminating reliance on manual experience, significantly reducing unnecessary mechanical disassembly workload, and improving the troubleshooting efficiency of the intelligent diagnostic system.
[0013] Please see Figure 2 As shown, Figure 2 A flowchart illustrating the closed-loop diagnosis method for automotive faults provided in this embodiment of the invention includes the following steps: Step S1: Obtain the current vehicle's fault code set and fault phenomenon set through the vehicle diagnostic interface and sensors. Match the fault code set and the fault phenomenon set to obtain a set of suspected components. Quantize the fault code set and the set of suspected components to obtain the initial number of fault codes and the initial number of suspected components, respectively.
[0014] Specifically, before performing the matching operation, a structured business data association model is pre-established in a cloud-based or local AI corpus. This business data systematically integrates D01 fault code business data and D03 fault symptom business data, and records specific fault codes, fault code definitions, fault code triggering conditions, and the association mapping relationship between fault codes and components. During specific diagnosis, the system automatically reads the currently triggered fault code set of the entire vehicle and system through on-board diagnostic equipment (such as OBD / CAN bus interface) and combines it with the fault phenomenon set collected by sensors. Subsequently, using the acquired fault codes as search indexes, the system automatically calls and traverses the D01 fault code business data and D03 fault symptom business data in the AI corpus. By comparing fault codes and triggering conditions, it accurately filters and extracts records that meet the current triggering conditions, thereby automatically matching and outputting the corresponding set of suspicious components.
[0015] For example, when a multi-cylinder misfire fault code is read through the interface, the system automatically compares the trigger conditions, matches and locks a set of suspected components from the database, including spark plugs, ignition coils, and fuel injectors. After completing the above matching, the system statistically quantifies the number of elements in the acquired fault code set and the number of elements in the suspected component set, and records the number of fault codes counted as the initial fault code count. The number of suspicious parts counted is recorded as the initial number of suspicious parts. This step, through standardized corpus queries and dynamic statistics, abstracts scattered fault features into two computable quantitative dimensions, laying a complete data foundation for the subsequent matching logic tree.
[0016] Step S2: Perform preliminary fault judgment based on the fault code set to obtain the updated number of fault codes and the updated number of suspicious parts.
[0017] Specifically, based on the fault code set and the preset rules of the business database, it is determined whether the current fault is a physical damage fault that requires direct replacement of parts, such as an open circuit, short circuit, or physical collision and breakage.
[0018] If the fault is due to physical damage, the initial quantified data is used directly and sent to subsequent steps. If not, the vehicle's current status parameters (such as battery voltage) are read to determine if the diagnostic environment data meets the preset conditions. Provided the diagnostic environment conditions are met and no suspicious parts are replaced, the following steps are performed sequentially according to priority: programming operation verification and repair for the fault code set, code adaptive value calibration and reset, and basic maintenance operations.
[0019] After the above operations are completed, the vehicle status data is collected again to obtain an updated fault code set. If the updated fault code set is not empty, it indicates that non-hardware damage factors such as software abnormalities or improper routine maintenance have been ruled out, and the actual fault still exists. At this time, the updated set is quantified a second time to obtain the updated number of fault codes and the updated number of suspected parts.
[0020] This invention, through conditional branching and a specific execution order, logically prevents disassembly and inspection operations caused by misjudging software coding problems as hardware damage.
[0021] Step S3: Based on the updated number of fault codes and the updated number of suspicious parts, perform cross-matching in a preset two-dimensional four-quadrant classification matrix to obtain the target diagnostic logic tree type corresponding to the current fault.
[0022] Specifically, a two-dimensional four-quadrant classification matrix is pre-constructed using the number of fault codes as the first quantization axis and the number of suspicious parts as the second quantization axis, with logical boundaries of a quantity value of 1 and a quantity value greater than 1. After obtaining the updated data output from step S2, it is input into this matrix as two-dimensional coordinate points for cross-matching.
[0023] If the updated fault code count is 1 and the updated suspected component count is 1, the current fault is matched to the first type (D-type logic tree); if the updated fault code count is 1 and the updated suspected component count is greater than 1, the current fault is matched to the second type (B-type logic tree); if the updated fault code count is greater than 1 and the updated suspected component count is 1, the current fault is matched to the third type (C-type logic tree); if the updated fault code count is greater than 1 and the updated suspected component count is greater than 1, the current fault is matched to the fourth type (A-type logic tree). This matching process directly uses the integer parameters collected at the diagnostic site as dynamic mapping coordinates to fall into the corresponding quadrant, providing a structured classification entry point for subsequently generating differentiated component detection and sorting rules.
[0024] Step S4: Based on the target diagnostic logic tree type, obtain the corresponding component detection sorting rules to determine the detection order of each suspicious component.
[0025] Specifically, based on the target diagnostic logic tree type output in step S3, the corresponding sorting rules are executed to determine the detection order of each suspicious component. If it is the first type, a single fault code is directly associated with a single fault phenomenon, and the unique suspicious component is identified as the target suspicious component. If it is the third type, the common power supply or signal path of multiple fault codes is extracted as a common association point, and the uniquely locked suspicious component is identified as the target suspicious component. If it is the second type, the correlation degree between each suspicious component and the corresponding fault phenomenon is obtained, and the cross-influence coefficient is obtained by statistically analyzing the in-degree in the constructed component cross-influence directed graph. The influence weight is calculated by combining the correlation degree and the cross-influence coefficient, and then the detection order of each suspicious component is determined according to the influence weight from high to low. If it is the fourth type, the system establishes a ternary correlation matrix of fault codes, fault phenomena, and components, obtains the component's safety correlation degree, core function influence degree, and secondary fault risk degree weight from the business database, and calculates a comprehensive score by combining the extracted dynamic adjustment coefficient. Finally, the detection order of each suspicious component is determined by arranging them in descending order of the comprehensive score. This step matches the corresponding quantization ranking model for faults of different complexity types, transforming experience-based troubleshooting into ordered detection based on weighted data.
[0026] Step S5: Perform functional measurements on the target suspicious components in the suspicious component set in the order of detection to obtain measured values. Compare the measured values with the preset design standard values to obtain the target deviation degree. Execute the corresponding processing instructions in sequence according to the priority order of the quantitative grading interval to which the target deviation degree belongs, until the fault is confirmed to be eliminated.
[0027] Specifically, following the aforementioned detection sequence, the system reads the current data stream and status of the sensors, actuators, or controllers corresponding to the suspected target component through the on-board diagnostic interface to obtain the measured values. Subsequently, the system uses the component's identifier as an index to extract the design standard value and tolerance range of the suspected target component from the D02 standard data table pre-stored in the artificial intelligence corpus. The system compares the obtained measured values with the design standard value and the tolerance range. When the measured value exceeds the tolerance range, the system calculates the absolute difference between the measured value and the design standard value and divides it by the design standard value to obtain the percentage deviation, thereby determining the degree of deviation from the design standard value.
[0028] This invention quantifies fault phenomena and fault codes by acquiring them. Based on this, it filters out non-hardware damage factors through pre-fault judgment, and then uses a two-dimensional matrix cross-matching to determine the diagnostic logic tree type. Subsequently, it generates corresponding detection sorting rules and compares the measured values with standard values according to this order to execute hierarchical processing instructions. This invention introduces a pre-filtering and screening mechanism to avoid disassembly and inspection without verifying software or maintenance issues. By constructing a two-dimensional classification matrix composed of the number of fault codes and the number of suspicious components, it divides multi-state concurrent faults into a standard processing framework. Then, it generates a detection order based on classification differences, and finally executes a hierarchical closed loop of repair, adjustment, or replacement based on the standard value deviation, thereby improving the accuracy of fault diagnosis and location and the standardization of business execution.
[0029] Furthermore, in some embodiments, the step of performing preliminary fault judgment based on the fault code set to obtain the updated number of fault codes and the updated number of suspicious components specifically includes: If the current fault is determined to be a fault that requires direct replacement of parts, the initial number of fault codes and the initial number of suspected parts are used as the updated number of fault codes and the updated number of suspected parts. If it is determined that the current fault does not belong to the fault that requires direct replacement of parts, and the current diagnostic environment data is found to meet the preset conditions, without replacing the suspected parts, the programming operation verification and repair, code adaptive value calibration and reset, and basic maintenance operations for the fault code set are executed in order of priority. By re-collecting data after the programming operation verification and repair, code adaptive value calibration and reset, and basic maintenance operations are completed, an updated set of fault codes is obtained. When it is determined that the updated fault code set is not empty, the updated fault code set and the re-matched suspicious parts set are quantified to obtain the updated fault code number and the updated suspicious parts number.
[0030] Specifically, before initiating specific component disassembly and testing operations, a pre-fault judgment mechanism for non-component faults is executed. First, based on the fault code set and preset rules of the business database, it is determined whether the current fault belongs to a physical damage fault that requires direct replacement of components (such as physical breakage of sensors caused by collisions or clear open or short circuits in the wiring).
[0031] If the current fault is determined to be one that requires direct replacement of parts, then the non-parts-related troubleshooting actions such as programming operation verification and repair, code adaptive value calibration and reset, and basic maintenance are skipped. At this point, the initial fault code count and the initial suspected part count obtained from the initial quantization are directly used as the updated fault code count and the updated suspected part count, and are directly sent to subsequent steps for cross-matching in the two-dimensional four-quadrant classification matrix.
[0032] If the current fault is determined not to be one that requires direct replacement of parts, then the diagnostic environment data is further assessed based on the current status parameters of the sensors, actuators, and controllers (such as current battery voltage and interface software version number) read by the diagnostic equipment to determine whether the current diagnostic environment data meets the preset programming or coding calibration prerequisites.
[0033] When the current diagnostic environment data is found to meet the preset conditions, without replacing the suspected components, the system will sequentially guide the execution of programming operation verification and repair for the fault code set, code setting or adaptive value calibration reset, and basic maintenance operations such as replacing battery coolant, according to priority.
[0034] Conversely, if the current diagnostic environment data does not meet the preset programming or coding calibration prerequisites when the pre-fault judgment mechanism is executed (for example, the current battery voltage is detected to be lower than the 12V voltage threshold required for programming), the blocking mechanism will be triggered, outputting a warning message indicating that the environment does not meet the requirements, and prohibiting the execution of the corresponding operation until the technician intervenes to restore the environment conditions to the required state.
[0035] After the programming, coding, and reset operations are completed, the vehicle status data is re-collected to obtain an updated fault code set. If the updated fault code set is empty, it is determined that the non-hardware damage fault has been resolved, a conclusion is output, and the diagnosis is terminated. If the updated fault code set is not empty, it indicates that the actual hardware fault still exists. At this time, the updated fault code set and the re-matched suspicious component set are quantified to obtain the updated number of fault codes and the updated number of suspicious components, and these are input into the next matrix matching stage.
[0036] It should be noted that the specific execution process for the programming operation verification and repair, and the coding adaptive value calibration and reset operation in the above-mentioned preliminary fault diagnosis steps is as follows: Programming operation usually refers to writing or updating the control software program to the vehicle's specific electronic control unit (ECU) through diagnostic equipment to repair known software control logic defects; coding operation usually refers to modifying, writing or activating the relevant function codes and operating parameters inside the electronic control unit according to the specific configuration of the vehicle (such as different optional configurations); Reset operation (such as throttle adaptive learning value reset) refers to clearing the adaptive learning value generated by the electronic control unit due to mechanical wear or environmental changes during long-term operation, restoring it to the initial factory setting state.
[0037] The above programming, coding, and adaptive value reset operations can all be completed using existing standard vehicle diagnostic communication protocols and conventional automotive diagnostic equipment. These are common knowledge to those skilled in the art, so we will not elaborate further on their underlying communication principles, communication message structures, and code burning process here.
[0038] If it is determined that the updated fault code set is not empty, then the updated fault code set and the re-matched suspicious parts set are subjected to secondary quantization to obtain the updated fault code number and the updated suspicious parts number, and then input into the next matrix matching stage.
[0039] Furthermore, in some embodiments, the step of performing cross-matching in a preset two-dimensional four-quadrant classification matrix based on the updated number of fault codes and the updated number of suspicious components to obtain the target diagnostic logic tree type corresponding to the current fault specifically includes: If the number of updated fault codes is 1 and the number of updated suspicious parts is 1, the current fault is matched to the first type in the two-dimensional four-quadrant classification matrix, and the first type is determined as the target diagnostic logic tree type. If the number of updated fault codes is 1 and the number of updated suspicious components is greater than 1, the current fault is matched to the second type in the two-dimensional four-quadrant classification matrix, and the second type is determined as the target diagnostic logic tree type. If the number of updated fault codes is greater than 1 and the number of updated suspicious parts is 1, the current fault is matched to the third type in the two-dimensional four-quadrant classification matrix, and the third type is determined as the target diagnostic logic tree type. If the number of updated fault codes is greater than 1 and the number of updated suspicious components is greater than 1, the current fault is matched to the fourth type in the two-dimensional four-quadrant classification matrix, and the fourth type is determined as the target diagnostic logic tree type.
[0040] Specifically, before performing cross-matching or during the diagnostic initialization phase, the two-dimensional four-quadrant classification matrix is pre-constructed. The specific construction process is as follows: a two-dimensional planar analysis matrix is constructed using the number of fault codes as the first-dimensional quantization axis and the number of suspicious parts as the second-dimensional quantization axis.
[0041] On the first-dimensional quantization coordinate axis, the first logical boundary is defined as a quantity value of 1 (single fault code) and a quantity value greater than 1 (multiple fault codes); on the second-dimensional quantization coordinate axis, the second logical boundary is defined as a quantity value of 1 (single component) and a quantity value greater than 1 (multiple components).
[0042] By orthogonally intersecting the first and second logical boundary boundaries in the two-dimensional plane analysis matrix, the matrix space is divided into four independent quadrant regions. Each of the four quadrant regions is then mapped to one of four preset diagnostic logic trees. Specifically, this includes: The first quadrant region where the first dimension is 1 and the second dimension is 1 is intersected is preset to the first type (i.e., single fault code, single component abnormality type). The second quadrant region where the first dimension is 1 and the second dimension is greater than 1 intersects is pre-configured as the second type (i.e., single fault code, multi-component abnormality type). The third quadrant region where the first dimension is greater than 1 and the second dimension is 1 intersects is pre-configured as the third type (i.e., multiple fault codes, single component abnormality type). The fourth quadrant region where the first dimension is greater than 1 and the second dimension is greater than 1 is pre-configured as the fourth type (i.e., multiple fault codes, multiple component abnormality type).
[0043] Through the above construction process, the classification matrix of this invention does not rely on model training and feature extraction based on historical fault datasets. Instead, it directly uses two quantifiable integer parameters (i.e., the number of fault codes and the number of suspicious parts) collected in real time at the diagnostic site as dynamic mapping coordinates. After obtaining the updated number of fault codes and the updated number of suspicious parts, these are directly substituted into the two-dimensional four-quadrant classification matrix as coordinate points, which can quickly and uniquely fall into the corresponding quadrant region. This achieves structured dimensionality reduction classification of complex concurrent fault scenarios and provides a standardized execution entry point for subsequently generating differentiated component detection and ranking rules.
[0044] If the number of updated fault codes is 1 and the number of updated suspicious parts is 1, the current fault is matched to the first type (i.e., a D-type logic tree with a single fault code and a single part abnormality) in the two-dimensional four-quadrant classification matrix, and the first type is determined as the target diagnostic logic tree type. If the number of updated fault codes is 1 and the number of updated suspicious components is greater than 1, the current fault is matched to the second type (i.e., a B-type logic tree with a single fault code and multiple component anomalies) in the two-dimensional four-quadrant classification matrix, and the second type is determined as the target diagnostic logic tree type. If the number of updated fault codes is greater than 1 and the number of updated suspicious parts is 1, the current fault is matched to the third type (i.e., a C-type logic tree with multiple fault codes and single part abnormalities) in the two-dimensional four-quadrant classification matrix, and the third type is determined as the target diagnostic logic tree type. If the number of updated fault codes is greater than 1 and the number of updated suspicious components is greater than 1, the current fault is matched to the fourth type (i.e., the most complex A-type logic tree with multiple fault codes and multiple component anomalies) in the two-dimensional four-quadrant classification matrix, and the fourth type is determined as the target diagnostic logic tree type.
[0045] By establishing this two-dimensional four-quadrant matrix structure, a definite mapping classification can be output for fault scenarios involving both single-concurrency and multi-concurrency scenarios, thus avoiding the blindness of experience-based diagnosis.
[0046] Furthermore, in some embodiments, obtaining the corresponding component detection sorting rules based on the target diagnostic logic tree type to determine the detection order of each suspicious component specifically includes: If it is the first type, directly associate a single fault code with a single fault phenomenon, and identify the unique suspected component as the target suspected component; If it is the second type, the cross-influence coefficient is obtained by statistically analyzing the in-degree of the directed graph in the set of suspicious parts, and the influence weight is calculated in combination with the cross-influence coefficient. The detection order of each suspicious part is determined according to the influence weight from high to low. If it is the third type, extract the common correlation points of multiple fault codes, and identify the uniquely locked suspicious component as the target suspicious component; If it is the fourth type, a ternary correlation matrix of fault codes, fault phenomena and components is established to obtain the safety correlation degree, core function impact degree and secondary fault risk degree, and a comprehensive score is calculated in combination with the dynamic adjustment coefficient. The components are then arranged in descending order of the comprehensive score to determine the detection order of each suspicious component.
[0047] Specifically, for the first type, a single fault code is directly associated with a single fault symptom, identifying the only suspect component as the target suspect component. For the third type, common correlations among multiple fault codes are extracted, identifying the only identified suspect component as the target suspect component. Since the first and third types ultimately associate only one suspect component, their identification mechanism simplifies the scope of the investigation.
[0048] If it is the second type, the cross-influence coefficient is obtained by statistically analyzing the in-degree of the directed graph of the suspected component set, and the influence weight is calculated based on the cross-influence coefficient. The detection order of each suspected component is determined according to the influence weight from high to low. Specifically, the correlation between each suspected component and the corresponding fault phenomenon is obtained; this is calculated using the formula... Where W_inf is the influence weight, P_i is the correlation degree, R_i is the cross-influence coefficient, and N_phenomena is the number of fault phenomena, which is automatically determined by constructing a directed graph network of cross-influences between components and using reachability traversal or in-degree statistics. A higher value is taken when the in-degree is 0 (not affected by other components). N_phenomena is the total number of fault phenomena currently counted. This method quantifies the cascading influence between components, allowing root cause components to be prioritized for investigation.
[0049] For the fourth type, a ternary correlation matrix is established between fault codes, fault phenomena, and components to obtain the safety correlation degree, core function impact degree, and secondary fault risk degree. A comprehensive score is then calculated based on this, and the components are arranged in descending order to determine the testing sequence. The calculation formula is: Comprehensive Score .in, This is a safety correlation weighting coefficient, used to measure the degree of threat that the failure of this component poses to vehicle driving safety; This is the core function impact weighting coefficient, used to measure the level of impact of the failure of this component on the vehicle's core power or control system. This is a secondary failure risk weighting coefficient, used to measure the probability that a failure of this component will cause damage to other components. The above... , , To obtain objective parameters that characterize the inherent properties of components, the system uses the identification information of suspicious components as a query index to perform matching and retrieval in a pre-established enterprise core database, and directly reads and extracts the corresponding preset values from the database (for example, the safety correlation weight of braking system components is usually preset to the highest value of 1.0, and the weight of comfort system components is preset to 0.3). , , This is a dynamic adjustment coefficient used to adjust the weight ratios of each component based on the specific physical architecture of the vehicle under test. It is obtained by employing a hybrid dynamic adjustment mechanism instead of using fixed preset values. The vehicle's powertrain type (including conventional gasoline vehicles, hybrid electric vehicles (HEVs), plug-in hybrid electric vehicles (PHEVs), and battery electric vehicles (BEVs)) is read through the vehicle diagnostic interface. Then, using the powertrain type as an index, a static lookup is performed in a preset baseline coefficient table to obtain the set of coefficients uniquely corresponding to that powertrain type. , , The value. For example, when matching a gasoline vehicle, the coefficient is extracted ( When matching a pure electric vehicle (BEV), the coefficients are extracted. ).
[0050] Specifically, after matching the fourth type of logic tree, the system automatically extracts the currently acquired multiple sets of fault codes, multiple sets of fault phenomena, and multiple initially identified suspicious components. Using fault codes as the first dimension, fault phenomena as the second dimension, and suspicious components as the third dimension, a ternary association matrix is initialized and established in memory. The data structure of this matrix is implemented using a multidimensional array or an association mapping table. The element values in the matrix represent the association strength. When the system determines that a certain fault code and a certain fault phenomenon can be physically and logically caused by specific components, the corresponding matrix element is activated and assigned the corresponding association mapping value. Through this process, the system spatially and structurally binds discrete fault manifestations to physical entities at the computer level.
[0051] After constructing the ternary correlation matrix, the parameter extraction operation is performed by traversing the matrix to obtain the security correlation degree. Impact of core functions and the risk of secondary failures .
[0052] Traverse the ternary correlation matrix Extract the unique identifiers (such as component IDs or hardware codes) of all suspicious components corresponding to the activated associated nodes.
[0053] The unique identifier is used as the primary key to initiate a query request to the enterprise's core business database, either in the cloud or locally. This database contains a pre-defined static weight mapping table based on the physical and safety attributes of all vehicle components.
[0054] By matching the primary key, the calibration values of the corresponding component in three dimensions are directly read and returned. For example, if the matrix lock node is a brake vacuum pump, the query will automatically map it to specific values and assign them (e.g., ...). ).
[0055] The above processing mechanism ensures that the sorting logic takes into account both safety priorities and dynamically adapts to the structural characteristics of different energy vehicle models.
[0056] Furthermore, in some embodiments, the step of sequentially performing functional measurements on target suspicious components in the suspicious component set according to the detection order to obtain measured values, and comparing the measured values with preset design standard values to obtain the target deviation degree, specifically includes: Read the current data stream and status of the corresponding sensor, actuator, or controller of the suspected target component to obtain the measured value; Extract the design standard values and tolerance ranges of the target suspected components that are reserved in the business database; The measured value is compared with the design standard value and the tolerance range, and when the measured value exceeds the tolerance range, the degree of deviation of the measured value from the design standard value is determined.
[0057] Specifically, data communication is established with the vehicle through an on-board diagnostic interface (such as an OBD interface or CAN bus), and the underlying diagnostic protocol is invoked to read the real-time dynamic data stream and operating status of the sensor, actuator, or controller corresponding to the first suspected target component in the current detection sequence. For example, when detecting an air flow meter or intake air temperature sensor, its current output voltage, air flow value, or temperature value is directly captured and used as the measured value characterizing the current functional status of the component.
[0058] Subsequently, using the identification information of the suspected component as a search index, targeted queries are performed in the business database (such as the D02 standard data table) pre-reserved in the artificial intelligence corpus. From this, design standard values and tolerance ranges matching the component and the current test environment conditions are extracted. For example, the standard output voltage center value of a sensor at idle speed and its allowable upper and lower limits (e.g., a standard value of 1.8-2.2V) can be extracted, or the specific standard operating range of the intake air volume (e.g., 4.5 to 8.5 gm / sec).
[0059] After obtaining the above data, the measured value is logically compared within the normal threshold range formed by the design standard value and the tolerance range. When it is determined that the measured value falls outside the boundary of the tolerance range, it indicates that the component has malfunctioned. Subsequently, deviation quantification calculation is initiated to calculate the specific magnitude of the measured value's deviation from the design standard value, thereby determining the target deviation degree.
[0060] The target deviation is calculated by dividing the absolute difference between the measured value and the design standard value by the design standard value to obtain the deviation percentage (for example, the measured value deviates from the standard baseline by 4%, 15%, or 25%, etc.).
[0061] This invention transforms the traditional diagnostic model that relies on manual experience to observe data flow into a mechanism that automatically pulls baseline parameters and performs percentage quantification calculations. By accurately capturing the degree to which measured data deviates from the standard, it provides accurate numerical boundary basis for subsequent objective judgment of whether the component only needs mechanical adjustment, deep repair, or must be completely scrapped and replaced.
[0062] Furthermore, in some embodiments, the step of sequentially executing corresponding processing instructions according to the priority order of the quantization grading interval to which the target deviation degree belongs, until the fault is confirmed to be eliminated, specifically includes: By determining whether the parameters reflected by the degree of target deviation exceed the standard value, the corresponding quantitative grading interval is determined; If the target deviation falls into the first quantification grading range that indicates an adjustment beyond the standard value, an adjustment instruction is generated and executed as the processing instruction. If the target deviation falls into the second quantitative grading range, which indicates that the deviation exceeds the standard value and is repairable, a repair instruction is generated and executed as the processing instruction. If the target deviation falls into the third quantitative grading range, which indicates that the target deviation exceeds the standard value and cannot be repaired, a replacement instruction is generated and executed as the processing instruction. The fault is determined to be eliminated when the component status data is reread after executing any of the above instructions and the measured value is confirmed to have recovered to the preset design standard value range.
[0063] Specifically, quantitative grading interval boundaries corresponding to the target deviation level are pre-defined in the business data. After obtaining the target deviation level, the specific circumstances under which the parameter reflected by the deviation level exceeds the standard value are determined, and the parameter is then classified into the corresponding quantitative grading interval.
[0064] If the target deviation falls within the first quantitative grading range, indicating that the deviation exceeds the standard value but is adjustable, the system generates and prompts for an adjustment command. For example, when the system detects that the front wheel toe-in value of the four-wheel alignment exceeds the design standard value, but the deviation is within the mechanically adjustable range, the system outputs a processing command to adjust the extension and retraction of the ball joints on both the left and right steering tie rods.
[0065] If the target deviation falls within the second quantitative grading range, indicating that the deviation exceeds the standard value but is repairable, the system generates and prompts for a repair instruction. For example, when a tire pressure below the standard value is detected, triggering an underpressure fault, and the assessed leakage deviation is within the repairable range, the system outputs a processing instruction to repair the tire to restore the standard tire pressure value.
[0066] If the target deviation falls into the third quantitative grading range, which indicates that the deviation exceeds the standard value and cannot be repaired, the system generates and prompts for a replacement instruction. For example, when the measured data of the intake air temperature sensor indicates that it has completely failed and the deviation indicates that it cannot provide effective intake air temperature feedback to the engine controller, the system directly outputs a processing instruction to replace the sensor.
[0067] The first quantitative grading interval is defined as follows: when the percentage of deviation calculated from the target deviation falls within the first quantitative grading interval that is greater than 0 and less than or equal to 5% (for example, the front wheel toe-in value parameter of the four-wheel alignment exceeds the design standard value but the deviation is extremely small and within the mechanically adjustable range), it is determined that the parameter exceeds the standard value but can be restored through mechanical structure adjustment, and then an adjustment command is automatically generated and prompted to be executed (such as adjusting the extension and retraction of the ball joints on the left and right sides of the steering tie rod).
[0068] The second quantitative grading interval (repair interval) is defined as follows: when the percentage of deviation corresponding to the target deviation falls within the second quantitative grading interval that is greater than 5% and less than or equal to 20% (e.g., tire leakage causing low tire pressure, deviation within the repairable or repairable range of local parts), it is determined that the deviation exceeds the standard value and can be restored through repair. Then, a repair instruction (such as repairing the tire to restore the standard tire pressure value) is automatically generated and prompted to be executed.
[0069] The third quantification grading interval is when the percentage of deviation corresponding to the target deviation falls into the third quantification grading interval greater than 20% (for example, the measured data shows that the internal circuit of the sensor is completely ineffective, or the deviation from the center value is too large, resulting in the inability to provide effective feedback). It is determined that it exceeds the standard value and cannot be restored by conventional adjustment or repair. Then, it automatically generates and prompts to execute a replacement instruction (such as directly replacing the intake air temperature sensor).
[0070] After actually executing the above adjustment, repair, or replacement instructions, a closed-loop judgment procedure is executed to forcibly reread the current data stream and status data of the suspected target component. A second comparison is performed, and if it is confirmed that the measured value has recovered to the preset design standard value range, the currently associated fault is determined to be eliminated, and the diagnostic process for this branch ends.
[0071] The above implementation transforms the maintenance judgment logic into execution rules constrained by threshold ranges. It limits the escalation of intervention methods based on data deviation quantification results and outputs instructions according to a gradient priority order of adjustment, repair, and replacement. This effectively avoids over-repair caused by indiscriminately recommending component replacement. Simultaneously, the post-processing verification mechanism ensures that every fault handling operation has objective data to corroborate the conclusions, achieving a system-level diagnostic verification closed loop and reducing the possibility of lingering potential faults.
[0072] Scenario 1: Application Example of Type 4 (Class A Logic Tree, Multiple Fault Codes + Multiple Component Anomalies) Taking a certain new energy vehicle as an example, the vehicle simultaneously exhibited multiple fault phenomena such as insufficient power, warning lights illuminating, and abnormal brake pedal. The diagnostic equipment read three fault codes: drive motor overheating (DTC-E001), battery management system communication abnormality (DTC-B012), and brake vacuum pump low pressure alarm (DTC-C007). The execution process is as follows: In stage S1, the initial N_dtc=3, and the initial N_comp=3 (drive motor assembly, battery BMS controller, brake vacuum pump) is determined through correlation analysis. In the S2 phase, the programming and coding status were checked and found to be normal. However, a review of the maintenance records revealed that the battery coolant had not been replaced on schedule, prompting a request to perform the basic maintenance operation of replacing the coolant. After the operation, a retest showed that DTC-B012 had disappeared, but E001 and C007 remained. In the S3 phase, after the update of N_dtc=2 and N_comp=2, the system will match the current fault to the fourth type (Class A logic tree). In the S4 stage, the comprehensive score is calculated based on the three-level weight, and the detection order is determined as follows: brake vacuum pump (highest safety correlation, priority for investigation) → drive motor cooling system (high impact of core functions, second highest). In the S5 stage, the brake vacuum pump power supply voltage was first measured to be 10.2V (standard value ≥11.5V). The deviation fell within the repairable range. The investigation revealed that the auxiliary battery was depleted. After charging (repair instruction) was executed, DTC-C007 was eliminated. Further inspection of the drive motor revealed that the cooling fan blades were damaged. After executing the replacement instruction, DTC-E001 was eliminated. All faults were confirmed to be eliminated, and the diagnostic loop was closed.
[0073] Scenario 2: Application Example of the Third Type (Class C Logic Tree, Multiple Fault Codes + Single Component Anomaly) Taking a gasoline vehicle as an example, the vehicle simultaneously triggers four fault codes: throttle position (DTC-P0121), idle speed control (DTC-P0505), lean fuel (DTC-P0171), and intake pressure (DTC-P0107), accompanied by idle speed vibration. The execution process is as follows: In stage S1, the initial N_dtc=4. Through common power / signal path tracing, it was determined that these 4 fault codes all pointed to a single intake air volume metering component, i.e., the initial N_comp=1 (air flow meter). In stage S2, it was found that the throttle position learning value had a large deviation. Throttle adaptive learning (reset operation) was performed without replacing the parts, but the fault codes were not cleared after retesting. In stage S3, the updated N_dtc=4 and N_comp=1 were matched to the third type (Class C logic tree). In stage S4, the air flow meter was directly identified as the single target detection component. In stage S5, the measured output voltage at idle was only 1.1V (standard value 1.8~2.2V), and the target deviation fell into the third quantization level range, triggering a replacement command. After replacement, all four fault codes were cleared, and the closed loop ended.
[0074] Scenario 3: Application Example of Type 2 (Class B Logic Tree, Single Fault Code + Multiple Component Anomaly) A vehicle triggers a single fault code P0300 (random multi-cylinder misfire), but this is accompanied by various other fault symptoms such as reduced power and vibration. The execution process is as follows: In the S1 phase, the initial N_dtc=1, and the matched suspicious parts include spark plugs, ignition coils, fuel injectors, etc., N_comp=4; In phase S2, the relevant adaptive value reset operation of the ignition system was checked and performed, but the fault was not eliminated; In the S3 phase, the updated N_dtc=1 and N_comp=4 match the second type (B-type logic tree). In the S4 stage, a cross-influence analysis was performed, and it was found that the influence weight of the fuel injector causing other phenomena was the highest, so it was ranked first in the detection order. During the S5 stage, the measured fuel injection quantity of cylinder 1 was significantly outside the tolerance range. The fault was eliminated after the replacement order was executed.
[0075] Scenario 4: Application Example of Type 1 (D-type logic tree, single fault code + single component anomaly) A vehicle triggered a single fault code P0420 (catalytic conversion efficiency below threshold). The execution process is as follows: In stage S1, N_dtc=1, and the positioning system N_comp=1 (three-way catalytic converter). During the S2 phase, the engine control unit coding status and external base connections were checked and found to be normal. In the S3 phase, matching is performed to the first type (D-type logic tree); In the S4 stage, the three-way catalytic converter is directly positioned; In the S5 stage, the signals from the front and rear oxygen sensors are detected to confirm the catalytic converter failure. The target deviation indicates the need for replacement, and the fault is eliminated after the replacement command is executed.
[0076] Please see Figure 3 As shown, in one embodiment, a vehicle fault diagnosis device based on a dual-quantization logic tree is provided, the device comprising: The information acquisition and quantification module 801 is used to acquire the current vehicle's fault code set and fault phenomenon set through the vehicle diagnostic interface and sensors, match the fault code set and the fault phenomenon set to obtain a suspicious component set, and quantify the fault code set and the suspicious component set to obtain the initial number of fault codes and the initial number of suspicious components, respectively. The pre-judgment module 802 is used to perform pre-fault judgment based on the fault code set to obtain the updated number of fault codes and the updated number of suspicious parts. The classification and matching module 803 is used to perform cross-matching in a preset two-dimensional four-quadrant classification matrix based on the updated number of fault codes and the updated number of suspicious parts to obtain the target diagnostic logic tree type corresponding to the current fault. The target diagnostic logic tree type includes a first type of single fault code and single suspicious part, a second type of single fault code and multiple suspicious parts, a third type of multiple fault codes and single suspicious parts, and a fourth type of multiple fault codes and multiple suspicious parts. The sorting rule determination module 804 is used to obtain the corresponding component detection sorting rule according to the target diagnostic logic tree type to determine the detection order of each suspicious component; The closed-loop judgment and execution module 805 is used to perform functional measurements on the target suspicious parts in the suspicious parts set in the order of detection to obtain measured values, compare the measured values with the preset design standard values to obtain the target deviation degree, and execute the corresponding processing instructions in the order of priority corresponding to the quantitative grading interval to which the target deviation degree belongs, until the fault is confirmed to be eliminated.
[0077] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 4 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores operating systems and computer programs. The internal memory provides an environment for the operation of the operating systems and computer programs stored in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When executed by the processor, the computer program implements the functions or steps of a vehicle fault diagnosis method based on a dual-quantization logic tree.
[0078] In one embodiment, a computer device is provided, the internal structure of which can be as follows: Figure 4 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. When executed by the processor, the computer program implements the steps of the automotive fault diagnosis method based on a dual-quantization logic tree as described in the claims. This invention quantifies fault phenomena and fault codes by acquiring them. Based on this, it filters out non-hardware damage factors through pre-fault judgment, and then uses a two-dimensional matrix cross-matching to determine the diagnostic logic tree type. Subsequently, it generates corresponding detection sorting rules and compares the measured values with standard values according to this order to execute hierarchical processing instructions. This invention introduces a pre-filtering and screening mechanism to avoid disassembly and inspection without verifying software or maintenance issues. By constructing a two-dimensional classification matrix composed of the number of fault codes and the number of suspicious components, it divides multi-state concurrent faults into a standard processing framework. Then, it generates a detection order based on classification differences, and finally executes a hierarchical closed loop of repair, adjustment, or replacement based on the standard value deviation, thereby improving the accuracy of fault diagnosis and location and the standardization of business execution.
[0079] This invention quantifies fault phenomena and fault codes by acquiring them. Based on this, it filters out non-hardware damage factors through pre-fault judgment, and then uses a two-dimensional matrix cross-matching to determine the diagnostic logic tree type. Subsequently, it generates corresponding detection sorting rules and compares the measured values with standard values according to this order to execute hierarchical processing instructions. This invention introduces a pre-filtering and screening mechanism to avoid disassembly and inspection without verifying software or maintenance issues. By constructing a two-dimensional classification matrix composed of the number of fault codes and the number of suspicious components, it divides multi-state concurrent faults into a standard processing framework. Then, it generates a detection order based on classification differences, and finally executes a hierarchical closed loop of repair, adjustment, or replacement based on the standard value deviation, thereby improving the accuracy of fault diagnosis and location and the standardization of business execution.
[0080] In one embodiment, a computer-readable storage medium is provided, storing a computer program that, when executed by a processor, implements the steps of the automotive fault diagnosis method based on a dual-quantization logic tree as claimed. By executing this program, the system can be automated to perform non-component inspection, matrix matching classification, and closed-loop measurement verification.
[0081] This invention quantifies fault phenomena and fault codes by acquiring them. Based on this, it filters out non-hardware damage factors through pre-fault judgment, and then uses a two-dimensional matrix cross-matching to determine the diagnostic logic tree type. Subsequently, it generates corresponding detection sorting rules and compares the measured values with standard values according to this order to execute hierarchical processing instructions. This invention introduces a pre-filtering and screening mechanism to avoid disassembly and inspection without verifying software or maintenance issues. By constructing a two-dimensional classification matrix composed of the number of fault codes and the number of suspicious components, it divides multi-state concurrent faults into a standard processing framework. Then, it generates a detection order based on classification differences, and finally executes a hierarchical closed loop of repair, adjustment, or replacement based on the standard value deviation, thereby improving the accuracy of fault diagnosis and location and the standardization of business execution.
[0082] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.
[0083] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0084] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.
[0085] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.
Claims
1. A vehicle fault diagnosis method based on a dual-quantization logic tree, characterized in that, The term includes: The vehicle diagnostic interface and sensors are used to obtain the current vehicle's fault code set and fault phenomenon set. The fault code set and the fault phenomenon set are matched to obtain the suspected component set. The fault code set and the suspected component set are quantified to obtain the initial number of fault codes and the initial number of suspected components, respectively. Based on the fault code set, preliminary fault judgment is performed to obtain the updated number of fault codes and the updated number of suspicious parts; Based on the updated number of fault codes and the updated number of suspicious components, cross-matching is performed in a preset two-dimensional four-quadrant classification matrix to obtain the target diagnostic logic tree type corresponding to the current fault. Based on the target diagnostic logic tree type, the corresponding component detection sorting rules are obtained to determine the detection order of each suspicious component; According to the detection order, the target suspicious parts in the suspicious parts set are functionally measured in sequence to obtain the measured values. The measured values are compared with the preset design standard values to obtain the target deviation degree. The corresponding processing instructions are executed in sequence according to the priority order of the quantitative grading interval to which the target deviation degree belongs, until the fault is confirmed to be eliminated.
2. The method according to claim 1, characterized in that, The step of performing preliminary fault judgment based on the fault code set to obtain the updated number of fault codes and the updated number of suspicious components specifically includes: If the current fault is determined to be a fault that requires direct replacement of parts, the initial number of fault codes and the initial number of suspected parts are used as the updated number of fault codes and the updated number of suspected parts. If it is determined that the current fault does not belong to the fault that requires direct replacement of parts, and the current diagnostic environment data is found to meet the preset conditions, without replacing the suspected parts, the programming operation verification and repair, code adaptive value calibration and reset, and basic maintenance operations for the fault code set are executed in order of priority. By re-collecting data after the programming operation verification and repair, code adaptive value calibration and reset, and basic maintenance operations are completed, an updated set of fault codes is obtained. When it is determined that the updated fault code set is not empty, the updated fault code set and the re-matched suspicious parts set are quantified to obtain the updated fault code number and the updated suspicious parts number.
3. The method according to claim 1, characterized in that, The step involves performing cross-matching within a preset two-dimensional four-quadrant classification matrix based on the updated number of fault codes and the updated number of suspicious components to obtain the target diagnostic logic tree type corresponding to the current fault. Specifically, this includes: If the number of updated fault codes is 1 and the number of updated suspicious parts is 1, the current fault is matched to the first type in the two-dimensional four-quadrant classification matrix, and the first type is determined as the target diagnostic logic tree type. If the number of updated fault codes is 1 and the number of updated suspicious components is greater than 1, the current fault is matched to the second type in the two-dimensional four-quadrant classification matrix, and the second type is determined as the target diagnostic logic tree type. If the number of updated fault codes is greater than 1 and the number of updated suspicious parts is 1, the current fault is matched to the third type in the two-dimensional four-quadrant classification matrix, and the third type is determined as the target diagnostic logic tree type. If the number of updated fault codes is greater than 1 and the number of updated suspicious components is greater than 1, the current fault is matched to the fourth type in the two-dimensional four-quadrant classification matrix, and the fourth type is determined as the target diagnostic logic tree type.
4. The method according to claim 1, characterized in that, The step of obtaining the corresponding component detection sorting rules based on the target diagnostic logic tree type to determine the detection order of each suspicious component specifically includes: If it is the first type, directly associate a single fault code with a single fault phenomenon, and identify the unique suspected component as the target suspected component; If it is the second type, the cross-influence coefficient is obtained by statistically analyzing the in-degree of the directed graph in the set of suspicious parts, and the influence weight is calculated in combination with the cross-influence coefficient. The detection order of each suspicious part is determined according to the influence weight from high to low. If it is the third type, extract the common correlation points of multiple fault codes, and identify the uniquely locked suspicious component as the target suspicious component; If it is the fourth type, a ternary correlation matrix of fault codes, fault phenomena and components is established to obtain the safety correlation degree, core function impact degree and secondary fault risk degree, and a comprehensive score is calculated in combination with the dynamic adjustment coefficient. The components are then arranged in descending order of the comprehensive score to determine the detection order of each suspicious component.
5. The method according to claim 4, characterized in that, The method of obtaining the cross-influence coefficient by statistically analyzing the in-degree of the directed graph in the set of suspected parts, and calculating the influence weight by combining the cross-influence coefficient, specifically includes: Obtain the correlation between each suspicious component and the corresponding fault phenomenon; By combining the cross-influence coefficient, the correlation degree, and the influence calculation formula for the number of fault phenomena, the influence weight of each suspected component is obtained. The influence calculation formula is as follows: Wherein, W_inf is the influence weight, P_i is the correlation, R_i is the cross-influence coefficient, and N_phenomena is the number of fault phenomena.
6. The method according to claim 1, characterized in that, The step of performing functional measurements on target suspicious components in the suspicious component set according to the detection order to obtain measured values, and comparing the measured values with preset design standard values to obtain the degree of target deviation, specifically includes: Read the current data stream and status of the corresponding sensor, actuator, or controller of the suspected target component to obtain the measured value; Extract the design standard values and tolerance ranges of the target suspected components that are reserved in the business database; The measured value is compared with the design standard value and the tolerance range, and when the measured value exceeds the tolerance range, the degree of deviation of the measured value from the design standard value is determined.
7. The method according to claim 1, characterized in that, The step of executing corresponding processing instructions sequentially according to the priority order of the quantization grading interval to which the target deviation degree belongs, until the fault is confirmed to be eliminated, specifically includes: By determining whether the parameters reflected by the degree of target deviation exceed the standard value, the corresponding quantitative grading interval is determined; If the target deviation falls into the first quantification grading range that indicates an adjustment beyond the standard value, an adjustment instruction is generated and executed as the processing instruction. If the target deviation falls into the second quantitative grading range, which indicates that the deviation exceeds the standard value and is repairable, a repair instruction is generated and executed as the processing instruction. If the target deviation falls into the third quantitative grading range, which indicates that the target deviation exceeds the standard value and cannot be repaired, a replacement instruction is generated and executed as the processing instruction. The fault is determined to be eliminated when the component status data is reread after executing any of the above instructions and the measured value is confirmed to have recovered to the preset design standard value range.
8. A vehicle fault diagnosis device based on a dual-quantization logic tree, characterized in that, The device includes: The information acquisition and quantification module is used to acquire the current vehicle's fault code set and fault phenomenon set through the vehicle diagnostic interface and sensors, match the fault code set and the fault phenomenon set to obtain a set of suspected parts, and quantify the fault code set and the set of suspected parts to obtain the initial number of fault codes and the initial number of suspected parts, respectively. The pre-judgment module is used to perform pre-fault judgment based on the fault code set to obtain the updated number of fault codes and the updated number of suspicious parts; The classification and matching module is used to perform cross-matching in a preset two-dimensional four-quadrant classification matrix based on the updated number of fault codes and the updated number of suspicious parts to obtain the target diagnostic logic tree type corresponding to the current fault. The target diagnostic logic tree type includes a first type of single fault code and single suspicious part, a second type of single fault code and multiple suspicious parts, a third type of multiple fault codes and single suspicious parts, and a fourth type of multiple fault codes and multiple suspicious parts. The sorting rule determination module is used to obtain the corresponding component detection sorting rule based on the target diagnostic logic tree type to determine the detection order of each suspicious component; The closed-loop judgment and execution module is used to perform functional measurements on the target suspicious parts in the suspicious parts set in the order of detection to obtain measured values, compare the measured values with the preset design standard values to obtain the target deviation degree, and execute the corresponding processing instructions in the order of priority corresponding to the quantitative grading interval to which the target deviation degree belongs, until the fault is confirmed to be eliminated.
9. A computer device, characterized in that, The computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the vehicle fault diagnosis method based on a dual quantization logic tree as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the vehicle fault diagnosis method based on a dual-quantization logic tree as described in any one of claims 1 to 7.
