Vehicle equipment diagnosis method and system based on large model, and storage medium

Through the large-model-based vehicle equipment diagnosis method, historical fault data and vehicle operation data are used to extract and predict features, and automatic fault diagnosis is achieved, solving the problem of low diagnostic efficiency of vehicle equipment in the existing technology and improving diagnostic efficiency and accuracy.

CN119087987APending Publication Date: 2024-12-06BEIJING UNISOUND INFORMATION TECH CO LTD

Patent Information

Application Number
CN202411272842.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-11
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The diagnostic efficiency of existing vehicle and machine equipment is low and requires professionals to make judgments, resulting in low diagnostic efficiency.

Method used

The vehicle-machinery equipment diagnosis method is used to build model training samples by obtaining historical fault data and vehicle operation data, perform feature extraction and fault prediction, and use parameter gradient backpropagation to converge the large model to achieve automatic fault diagnosis.

Benefits of technology

No manual judgment is required, which improves the diagnostic efficiency of vehicle and machine equipment, quickly identify fault types and causes, reduces human judgment errors, early warning of potential faults, and reduces maintenance costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119087987A_ABST
    Figure CN119087987A_ABST
Patent Text Reader

Abstract

The invention provides a vehicle equipment diagnosis method and system based on a large model and a storage medium, and the method comprises the steps: inputting a model training sample into the large model for feature extraction, obtaining diagnosis features, and carrying out the fault prediction of the diagnosis features, and obtaining a predicted fault; calculating model loss according to the predicted fault and the fault label of the model training sample, and calculating a parameter gradient according to the model loss; performing back propagation in the large model according to the parameter gradient until the large model converges; and acquiring vehicle-mounted information acquisition information of a target vehicle, and inputting the vehicle-mounted information acquisition information into the converged large model for fault diagnosis to obtain a fault diagnosis result. According to the embodiment of the invention, on the basis of the converged large model, the vehicle-mounted equipment diagnosis can be automatically carried out on the target vehicle, the vehicle-mounted equipment diagnosis does not need to be carried out manually, and the vehicle-mounted equipment diagnosis efficiency is effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to a vehicle equipment diagnosis method, system and storage medium based on a large model. Background Art

[0002] With the continuous development of automobile technology, vehicle systems are becoming increasingly complex, and the difficulty of fault diagnosis is also increasing. Vehicles are generally equipped with an on-board diagnostic system (OBD) interface. When vehicle equipment diagnosis is required, the vehicle status information is obtained through the OBD interface, and the vehicle equipment diagnosis is performed based on the obtained status information.

[0003] In the existing vehicle equipment diagnosis process, professionals are required to make judgments based on their experience, which leads to low efficiency in vehicle equipment diagnosis. Summary of the invention

[0004] The purpose of the embodiments of the present invention is to provide a vehicle equipment diagnosis method, system and storage medium based on a large model to solve the problem of low efficiency of vehicle equipment diagnosis in the prior art.

[0005] The embodiment of the present invention is implemented as follows: a vehicle equipment diagnosis method based on a large model, the method comprising:

[0006] Acquire historical fault data and vehicle operation data, and construct model training samples based on the historical fault data and the vehicle operation data;

[0007] Inputting the model training samples into the large model to extract features to obtain diagnostic features, and performing fault prediction on the diagnostic features to obtain predicted faults;

[0008] Calculating a model loss according to the predicted fault and the fault label of the model training sample, and calculating a parameter gradient according to the model loss;

[0009] Performing back propagation in the large model according to the parameter gradient until the large model converges;

[0010] The vehicle computer collected information of the target vehicle is obtained, and the vehicle computer collected information is input into the converged large model for fault diagnosis to obtain a fault diagnosis result.

[0011] Preferably, the model training samples are input into a large model for feature extraction to obtain diagnostic features, including:

[0012] Extracting trend features, periodic features and mutation features in the model training samples according to the large model, and generating time series features according to the trend features, the periodic features and the mutation features;

[0013] Extracting position information and structural information from the model training sample according to the large model, and generating spatial features according to the position information and the structural information;

[0014] Extracting a parameter mean, a parameter variance, a parameter skewness, and a parameter peak value in the model training sample according to the large model, and generating statistical features according to the parameter mean, the parameter variance, the parameter skewness, and the parameter peak value;

[0015] Wherein, the diagnostic features include the time series features, the spatial features and the statistical features.

[0016] Preferably, extracting trend features, periodic features and mutation features in the model training samples according to the large model includes:

[0017] Extracting fuel consumption information and emission information from the model training sample according to the large model, and constructing a fuel consumption trend curve and an emission trend curve according to the fuel consumption information and the emission information to obtain the trend feature;

[0018] Extracting wear information from the model training sample according to the large model, and constructing a wear curve according to the wear information to obtain the periodic feature, wherein the wear information includes tire wear and brake pad wear;

[0019] The vehicle parameter values ​​in the model training samples are obtained according to the large model, parameter mutation points are determined according to the vehicle parameter values, and the mutation features are constructed according to the parameter mutation points.

[0020] Preferably, extracting the position information and structure information in the model training sample according to the large model includes:

[0021] Extracting the vehicle position in the model training sample according to the large model to obtain the position information;

[0022] The component positions of the vehicle components in the model training sample are extracted according to the large model, and the relative positions between different vehicle components are determined according to the component positions to obtain the structural information.

[0023] Preferably, the model training samples are input into a large model for feature extraction to obtain diagnostic features, including:

[0024] Extracting fault codes and vehicle parameter values ​​in the model training samples according to the large model, and determining abnormal parts according to the vehicle parameter values;

[0025] Feature mapping is performed on the fault code and the abnormal part to obtain a first mapping feature and a second mapping feature, and a feature matrix is ​​constructed according to the first mapping feature and the second mapping feature to obtain the diagnostic feature.

[0026] Preferably, the information collected by the vehicle computer is input into the converged large model for fault diagnosis to obtain a fault diagnosis result, including:

[0027] Performing fault prediction on the information collected by the vehicle computer according to the converged large model to obtain the fault location and fault type, and determining the fault degree corresponding to the fault location according to the fault type;

[0028] Determining a maintenance suggestion according to the fault location, the fault type and the fault degree, and storing the corresponding fault type, the fault degree and the maintenance suggestion for the fault location to obtain a fault diagnosis entry;

[0029] The fault diagnosis items are sorted according to the fault degree to obtain the fault diagnosis result.

[0030] Preferably, the information collected by the vehicle computer is input into the converged large model for fault diagnosis, and after the fault diagnosis result is obtained, the method further includes:

[0031] Determine a maintenance plan according to the fault location and fault type in the fault diagnosis result, and obtain the current position of the target vehicle;

[0032] A maintenance store is determined according to the current location and the maintenance plan, and store information of the maintenance store is sent to the owner of the target vehicle.

[0033] Another object of an embodiment of the present invention is to provide a vehicle equipment diagnosis system based on a large model, the system comprising:

[0034] A sample construction module, used to obtain historical fault data and vehicle operation data, and construct a model training sample based on the historical fault data and the vehicle operation data;

[0035] A model training module is used to input the model training samples into a large model for feature extraction to obtain diagnostic features, and to perform fault prediction on the diagnostic features to obtain predicted faults;

[0036] Calculating a model loss according to the predicted fault and the fault label of the model training sample, and calculating a parameter gradient according to the model loss;

[0037] Performing back propagation in the large model according to the parameter gradient until the large model converges;

[0038] The fault diagnosis module is used to obtain the vehicle computer collected information of the target vehicle, and input the vehicle computer collected information into the converged large model for fault diagnosis to obtain the fault diagnosis result.

[0039] Preferably, the model training module is also used for:

[0040] Extracting trend features, periodic features and mutation features in the model training samples according to the large model, and generating time series features according to the trend features, the periodic features and the mutation features;

[0041] Extracting position information and structural information from the model training sample according to the large model, and generating spatial features according to the position information and the structural information;

[0042] Extracting a parameter mean, a parameter variance, a parameter skewness, and a parameter peak value in the model training sample according to the large model, and generating statistical features according to the parameter mean, the parameter variance, the parameter skewness, and the parameter peak value;

[0043] Wherein, the diagnostic features include the time series features, the spatial features and the statistical features.

[0044] The embodiments of the present invention can effectively construct model training samples by acquiring historical fault data and vehicle operation data. By inputting the model training samples into a large model for feature extraction, the diagnostic features in the model training samples can be effectively extracted. The large model can effectively perform fault prediction based on the diagnostic features to obtain predicted faults. By performing back propagation of parameter gradients in the large model, the large model after convergence can automatically perform vehicle equipment diagnosis on the target vehicle. There is no need to perform vehicle equipment diagnosis manually, thereby effectively improving the efficiency of vehicle equipment diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flow chart of a vehicle equipment diagnosis method based on a large model provided by a first embodiment of the present invention;

[0046] Figure 2 is a structural schematic diagram of a vehicle equipment diagnostic system based on a large model provided by a second embodiment of the present invention;

[0047] Figure 3 It is a schematic diagram of the structure of a terminal device provided in the third embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. 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.

[0049] In order to illustrate the technical solution of the present invention, a specific embodiment is provided below for illustration.

[0050] Embodiment 1

[0051] See also Figure 1 , is a flow chart of a vehicle equipment diagnostic method based on a large model provided by a first embodiment of the present invention. The vehicle equipment diagnostic method based on a large model can be applied to any device or system. The vehicle equipment diagnostic method based on a large model includes the following steps:

[0052] Step S10, acquiring historical fault data and vehicle operation data, and constructing a model training sample according to the historical fault data and the vehicle operation data;

[0053] Among them, historical fault data and vehicle operation data within a preset time range are obtained, and pre-processing operations such as data cleaning and formatting are performed on the acquired historical fault data and vehicle operation data to remove noise and redundant information and improve data quality. The pre-processed historical fault data and vehicle operation data are data labeled to obtain model training samples.

[0054] Step S20, inputting the model training samples into the large model to extract features to obtain diagnostic features, and performing fault prediction on the diagnostic features to obtain predicted faults;

[0055] Among them, by inputting the model training samples into the big model for feature extraction, the diagnostic features in the model training samples can be effectively extracted, and the big model based on the diagnostic features can effectively predict faults and obtain predicted faults. The predicted faults include information such as fault type, fault location, fault severity, and recommended maintenance measures.

[0056] Optionally, the model training samples are input into a large model for feature extraction to obtain diagnostic features, including:

[0057] Extracting trend features, periodic features and mutation features in the model training samples according to the large model, and generating time series features according to the trend features, the periodic features and the mutation features; wherein, by extracting trend features, periodic features and mutation features, time series features can be effectively generated based on trend features, periodic features and mutation features, and time series-related feature information in the model training samples can be effectively represented based on the time series features;

[0058] Extracting the position information and the structure information in the model training sample according to the large model, and generating the spatial features according to the position information and the structure information; wherein, by extracting the position information and the structure information, the spatial features can be effectively generated, and based on the spatial features, the spatially related feature information in the model training sample can be effectively represented;

[0059] Extracting the parameter mean, parameter variance, parameter skewness and parameter peak value in the model training sample according to the large model, and generating statistical features according to the parameter mean, parameter variance, parameter skewness and parameter peak value; wherein, by obtaining the parameter mean, parameter variance, parameter skewness and parameter peak value, statistical features can be effectively generated, and feature information related to parameter values ​​in the model training sample can be effectively represented based on the statistical features;

[0060] Among them, diagnostic features include time series features, spatial features and statistical features.

[0061] Furthermore, extracting trend features, periodic features and mutation features in the model training samples according to the large model includes:

[0062] Extracting fuel consumption information and emission information from the model training sample according to the large model, and constructing a fuel consumption trend curve and an emission trend curve according to the fuel consumption information and the emission information to obtain the trend feature; wherein, by constructing the fuel consumption trend curve and the emission trend curve, the changes in the fuel consumption and emission values ​​of the vehicle can be effectively represented, and optionally, a curve of the signal value change at any part of the target vehicle can also be generated;

[0063] Extracting wear information from the model training sample according to the large model, and constructing a wear curve according to the wear information to obtain the periodic feature; wherein the wear information includes tire wear and brake pad wear, and optionally, may also be the wear value of any part of the target vehicle;

[0064] The vehicle parameter values ​​in the model training samples are obtained according to the large model, the parameter mutation points are determined according to the vehicle parameter values, and the mutation features are constructed according to the parameter mutation points; wherein, if it is detected that the change value of any vehicle parameter value compared with the parameter value at the previous moment is greater than a change threshold, the vehicle parameter value is set as a mutation point, and the change threshold can be set according to demand.

[0065] Furthermore, extracting the position information and structure information in the model training sample according to the large model includes:

[0066] Extracting the vehicle position in the model training sample according to the large model to obtain the position information; wherein, by obtaining the position information, the large model can effectively analyze whether there is a relationship between the vehicle position and the vehicle equipment failure;

[0067] The component positions of the vehicle components in the model training samples are extracted according to the large model, and the relative positions between different vehicle components are determined according to the component positions to obtain the structural information; wherein, by determining the relative positions between different vehicle components, it is possible to effectively determine whether positional displacement has occurred in each vehicle component.

[0068] Preferably, the model training samples are input into a large model for feature extraction to obtain diagnostic features, including:

[0069] Extracting fault codes and vehicle parameter values ​​in the model training samples according to the large model, and determining abnormal parts according to the vehicle parameter values; wherein, obtaining abnormal values ​​in the vehicle parameter values, and determining the vehicle parts corresponding to the abnormal values ​​as abnormal parts, in this step, obtaining the standard range corresponding to the vehicle parts, comparing the vehicle parameter values ​​of the vehicle parts with the standard range, and if the vehicle parameter values ​​are not within the standard range, setting the vehicle parameter values ​​as abnormal values;

[0070] Feature mapping is performed on the fault code and the abnormal part to obtain a first mapping feature and a second mapping feature, and a feature matrix is ​​constructed based on the first mapping feature and the second mapping feature to obtain the diagnostic feature; wherein, feature mapping is performed on the fault code and the abnormal part according to a preset mapping relationship, which can effectively convert the fault code and the abnormal part into information of a preset dimension, effectively facilitating the construction of the feature matrix, and the preset mapping relationship can be set according to needs.

[0071] Step S30, calculating a model loss according to the predicted fault and the fault label of the model training sample, and calculating a parameter gradient according to the model loss;

[0072] Among them, by calculating the difference between the predicted fault and the standard fault corresponding to the fault label of the model training sample, the model loss can be effectively obtained, and the chain rule is used to calculate the gradient of the loss function with respect to each parameter to obtain the parameter gradient.

[0073] Step S40, performing back propagation in the large model according to the parameter gradient until the large model converges;

[0074] Among them, the parameter gradient is propagated from the output layer of the large model to the input layer through the back propagation algorithm. According to the parameter gradient, the optimization algorithm (such as gradient descent, etc.) is used to update the weights and bias parameters of the neural network in the large model to reduce the loss function value.

[0075] Step S50, obtaining the vehicle computer collected information of the target vehicle, and inputting the vehicle computer collected information into the converged large model to perform fault diagnosis to obtain a fault diagnosis result;

[0076] Among them, through the OBD interface of the target vehicle, multiple key parameter data of the vehicle's engine, transmission system, emission system, etc. are collected in real time to obtain vehicle computer collection information. By inputting the vehicle computer collection information into the converged large model for fault diagnosis, the vehicle computer equipment on the target vehicle can be automatically diagnosed to obtain fault diagnosis results.

[0077] Optionally, the information collected by the vehicle computer is input into the converged large model for fault diagnosis to obtain a fault diagnosis result, including:

[0078] According to the converged large model, fault prediction is performed on the information collected by the vehicle computer to obtain the fault location and fault type, and the fault degree corresponding to the fault location is determined according to the fault type; wherein the fault location and the fault type are combined to obtain combined information, and the combined information is matched with the degree query table to obtain the fault degree, and the degree query table stores the corresponding relationship between different combined information and corresponding fault degrees;

[0079] Determining a maintenance suggestion according to the fault location, the fault type and the fault degree, and storing the corresponding fault type, the fault degree and the maintenance suggestion for the fault location to obtain a fault diagnosis entry;

[0080] The fault diagnosis items are sorted according to the fault degree to obtain the fault diagnosis result; wherein, sorting the fault diagnosis items according to the fault degree effectively facilitates the vehicle owner to view the most serious fault part on the target vehicle.

[0081] Furthermore, after the vehicle computer collected information is input into the converged large model for fault diagnosis and the fault diagnosis result is obtained, the method further includes:

[0082] Determine a maintenance plan according to the fault location and fault type in the fault diagnosis result, and obtain the current position of the target vehicle; wherein the combination information is matched with a solution query table to obtain a maintenance plan, and the solution query table stores a corresponding relationship between different combination information and corresponding maintenance plans;

[0083] A maintenance store is determined according to the current location and the maintenance plan, and store information of the maintenance store is sent to the owner of the target vehicle.

[0084] In this embodiment, the fault diagnosis results can be displayed to the car owner or maintenance personnel through various methods such as the car display screen, mobile phone APP, etc. The car owner can take corresponding measures in time according to the prompts, and the maintenance personnel can quickly locate and solve the problem according to the fault diagnosis results.

[0085] In this embodiment, by acquiring historical fault data and vehicle operation data, the model training samples can be effectively constructed. By inputting the model training samples into the large model for feature extraction, the diagnostic features in the model training samples can be effectively extracted. Based on the diagnostic features, the large model can effectively predict faults and obtain predicted faults. The parameter gradient is back-propagated in the large model, so that the converged large model can automatically perform vehicle equipment diagnosis on the target vehicle, without the need to use manual methods to perform vehicle equipment diagnosis, effectively improving the efficiency of vehicle equipment diagnosis. In this embodiment, the large model can quickly identify the type and cause of vehicle faults and reduce human judgment errors. By early warning of potential faults, car owners can take timely measures to avoid the expansion of faults and reduce maintenance costs. Car owners can obtain vehicle fault information through convenient methods such as mobile phone APPs, and can understand the condition of the vehicle without going to a repair shop.

[0086] Embodiment 2

[0087] See also Figure 2 , is a schematic diagram of the structure of a vehicle equipment diagnostic system 100 based on a large model provided in a second embodiment of the present invention, including:

[0088] The sample construction module 10 is used to obtain historical fault data and vehicle operation data, and to construct a model training sample according to the historical fault data and the vehicle operation data.

[0089] The model training module 11 is used to input the model training samples into the large model for feature extraction to obtain diagnostic features, and perform fault prediction on the diagnostic features to obtain predicted faults;

[0090] Calculating a model loss according to the predicted fault and the fault label of the model training sample, and calculating a parameter gradient according to the model loss;

[0091] Back propagation is performed in the large model according to the parameter gradient until the large model converges.

[0092] Optionally, the model training module 11 is further used to: extract trend features, periodic features and mutation features in the model training samples according to the large model, and generate time series features according to the trend features, the periodic features and the mutation features;

[0093] Extracting position information and structural information from the model training sample according to the large model, and generating spatial features according to the position information and the structural information;

[0094] Extracting a parameter mean, a parameter variance, a parameter skewness, and a parameter peak value in the model training sample according to the large model, and generating statistical features according to the parameter mean, the parameter variance, the parameter skewness, and the parameter peak value;

[0095] Wherein, the diagnostic features include the time series features, the spatial features and the statistical features.

[0096] Furthermore, the model training module 11 is also used to: extract the fuel consumption information and emission information in the model training sample according to the large model, and construct a fuel consumption trend curve and an emission trend curve according to the fuel consumption information and the emission information to obtain the trend feature;

[0097] Extracting wear information from the model training sample according to the large model, and constructing a wear curve according to the wear information to obtain the periodic feature, wherein the wear information includes tire wear and brake pad wear;

[0098] The vehicle parameter values ​​in the model training samples are obtained according to the large model, parameter mutation points are determined according to the vehicle parameter values, and the mutation features are constructed according to the parameter mutation points.

[0099] Furthermore, the model training module 11 is also used to: extract the vehicle position in the model training sample according to the large model to obtain the position information;

[0100] The component positions of the vehicle components in the model training sample are extracted according to the large model, and the relative positions between different vehicle components are determined according to the component positions to obtain the structural information.

[0101] Preferably, the model training module 11 is further used to: extract the fault codes and vehicle parameter values ​​in the model training samples according to the large model, and determine the abnormal parts according to the vehicle parameter values;

[0102] Feature mapping is performed on the fault code and the abnormal part to obtain a first mapping feature and a second mapping feature, and a feature matrix is ​​constructed according to the first mapping feature and the second mapping feature to obtain the diagnostic feature.

[0103] The fault diagnosis module 12 is used to obtain the vehicle computer collected information of the target vehicle, and input the vehicle computer collected information into the converged large model to perform fault diagnosis to obtain a fault diagnosis result.

[0104] Optionally, the fault diagnosis module 12 is further used to: perform fault prediction on the information collected by the vehicle computer according to the converged large model, obtain the fault location and fault type, and determine the fault degree corresponding to the fault location according to the fault type;

[0105] Determining a maintenance suggestion according to the fault location, the fault type and the fault degree, and storing the corresponding fault type, the fault degree and the maintenance suggestion for the fault location to obtain a fault diagnosis entry;

[0106] The fault diagnosis items are sorted according to the fault degree to obtain the fault diagnosis result.

[0107] Furthermore, the fault diagnosis module 12 is also used to: determine a maintenance plan according to the fault location and fault type in the fault diagnosis result, and obtain the current position of the target vehicle;

[0108] A maintenance store is determined according to the current location and the maintenance plan, and store information of the maintenance store is sent to the owner of the target vehicle.

[0109] In this embodiment, by acquiring historical fault data and vehicle operation data, the model training samples can be effectively constructed. By inputting the model training samples into the large model for feature extraction, the diagnostic features in the model training samples can be effectively extracted. Based on the diagnostic features, the large model can effectively predict faults and obtain predicted faults. By backpropagating the parameter gradient in the large model, the converged large model can automatically perform on-board equipment diagnosis on the target vehicle. There is no need to use manual methods to diagnose the on-board equipment, which effectively improves the efficiency of on-board equipment diagnosis.

[0110] Embodiment 3

[0111] Figure 3 2 is a block diagram of a terminal device 2 provided in the third embodiment of the present application. Figure 3 As shown, the terminal device 2 of this embodiment includes: a processor 20, a memory 21, and a computer program 22 stored in the memory 21 and executable on the processor 20, such as a program of a vehicle equipment diagnostic method based on a large model. When the processor 20 executes the computer program 22, the steps in each embodiment of the vehicle equipment diagnostic method based on a large model are implemented.

[0112] Exemplarily, the computer program 22 may be divided into one or more modules, which are stored in the memory 21 and executed by the processor 20 to complete the present application. The one or more modules may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 22 in the terminal device 2. The terminal device may include, but is not limited to, a processor 20 and a memory 21.

[0113] The processor 20 may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0114] The memory 21 may be an internal storage unit of the terminal device 2, such as a hard disk or memory of the terminal device 2. The memory 21 may also be an external storage device of the terminal device 2, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal device 2. Further, the memory 21 may also include both an internal storage unit and an external storage device of the terminal device 2. The memory 21 is used to store the computer program and other programs and data required by the terminal device. The memory 21 may also be used to temporarily store data that has been output or is to be output.

[0115] In addition, each functional module in each embodiment of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional unit.

[0116] If the integrated module is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Among them, the computer-readable storage medium can be non-volatile or volatile. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable storage medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in computer-readable storage media can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable storage media do not include electrical carrier signals and telecommunication signals.

[0117] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A vehicle equipment diagnosis method based on a large model, characterized in that: The method comprises: Acquire historical fault data and vehicle operation data, and construct model training samples based on the historical fault data and the vehicle operation data; Inputting the model training samples into the large model to extract features to obtain diagnostic features, and performing fault prediction on the diagnostic features to obtain predicted faults; Calculating a model loss according to the predicted fault and the fault label of the model training sample, and calculating a parameter gradient according to the model loss; Performing back propagation in the large model according to the parameter gradient until the large model converges; The vehicle computer collected information of the target vehicle is obtained, and the vehicle computer collected information is input into the converged large model for fault diagnosis to obtain a fault diagnosis result.

2. The vehicle equipment diagnosis method based on a large model as claimed in claim 1, characterized in that: The model training samples are input into the large model for feature extraction to obtain diagnostic features, including: Extracting trend features, periodic features and mutation features in the model training samples according to the large model, and generating time series features according to the trend features, the periodic features and the mutation features; Extracting position information and structural information from the model training sample according to the large model, and generating spatial features according to the position information and the structural information; Extracting a parameter mean, a parameter variance, a parameter skewness, and a parameter peak value in the model training sample according to the large model, and generating statistical features according to the parameter mean, the parameter variance, the parameter skewness, and the parameter peak value; Wherein, the diagnostic features include the time series features, the spatial features and the statistical features.

3. The vehicle equipment diagnosis method based on a large model as claimed in claim 2, characterized in that: Extracting trend features, periodic features and mutation features in the model training samples according to the large model includes: Extracting fuel consumption information and emission information from the model training sample according to the large model, and constructing a fuel consumption trend curve and an emission trend curve according to the fuel consumption information and the emission information to obtain the trend feature; Extracting wear information from the model training sample according to the large model, and constructing a wear curve according to the wear information to obtain the periodic feature, wherein the wear information includes tire wear and brake pad wear; The vehicle parameter values ​​in the model training samples are obtained according to the large model, parameter mutation points are determined according to the vehicle parameter values, and the mutation features are constructed according to the parameter mutation points.

4. The vehicle equipment diagnosis method based on a large model as claimed in claim 2, characterized in that: Extracting the location information and structure information in the model training sample according to the large model includes: Extracting the vehicle position in the model training sample according to the large model to obtain the position information; The component positions of the vehicle components in the model training sample are extracted according to the large model, and the relative positions between different vehicle components are determined according to the component positions to obtain the structural information.

5. The vehicle equipment diagnosis method based on a large model as claimed in claim 1, characterized in that: The model training samples are input into the large model for feature extraction to obtain diagnostic features, including: Extracting fault codes and vehicle parameter values ​​in the model training samples according to the large model, and determining abnormal parts according to the vehicle parameter values; Feature mapping is performed on the fault code and the abnormal part to obtain a first mapping feature and a second mapping feature, and a feature matrix is ​​constructed according to the first mapping feature and the second mapping feature to obtain the diagnostic feature.

6. The vehicle equipment diagnosis method based on a large model as claimed in claim 1, characterized in that: Inputting the vehicle computer collected information into the converged large model to perform fault diagnosis and obtain a fault diagnosis result, including: Performing fault prediction on the information collected by the vehicle computer according to the converged large model to obtain the fault location and fault type, and determining the fault degree corresponding to the fault location according to the fault type; Determining a maintenance suggestion according to the fault location, the fault type and the fault degree, and storing the corresponding fault type, the fault degree and the maintenance suggestion for the fault location to obtain a fault diagnosis entry; The fault diagnosis items are sorted according to the fault degree to obtain the fault diagnosis result.

7. The vehicle equipment diagnosis method based on a large model as claimed in claim 1, characterized in that: After inputting the vehicle computer collected information into the converged large model for fault diagnosis and obtaining the fault diagnosis result, the method further includes: Determine a maintenance plan according to the fault location and fault type in the fault diagnosis result, and obtain the current position of the target vehicle; A maintenance store is determined according to the current location and the maintenance plan, and store information of the maintenance store is sent to the owner of the target vehicle.

8. A vehicle equipment diagnostic system based on a large model, characterized in that: The system comprises: A sample construction module, used to obtain historical fault data and vehicle operation data, and construct a model training sample based on the historical fault data and the vehicle operation data; A model training module is used to input the model training samples into a large model for feature extraction to obtain diagnostic features, and to perform fault prediction on the diagnostic features to obtain predicted faults; Calculating a model loss according to the predicted fault and the fault label of the model training sample, and calculating a parameter gradient according to the model loss; Performing back propagation in the large model according to the parameter gradient until the large model converges; The fault diagnosis module is used to obtain the vehicle computer collected information of the target vehicle, and input the vehicle computer collected information into the converged large model for fault diagnosis to obtain the fault diagnosis result.

9. The vehicle equipment diagnostic system based on a large model as claimed in claim 8, characterized in that: The model training module is also used to: Extracting trend features, periodic features and mutation features in the model training samples according to the large model, and generating time series features according to the trend features, the periodic features and the mutation features; Extracting position information and structural information from the model training sample according to the large model, and generating spatial features according to the position information and the structural information; Extracting a parameter mean, a parameter variance, a parameter skewness, and a parameter peak value in the model training sample according to the large model, and generating statistical features according to the parameter mean, the parameter variance, the parameter skewness, and the parameter peak value; Wherein, the diagnostic features include the time series features, the spatial features and the statistical features.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Citation Information

Patent Citations

  • On-line diagnosis method of vehicle body assembly quality based on part posture relationship discrimination

    CN107745758A

  • Online monitoring and fault diagnosis system for automobile engine

    CN108303262A

  • Production site data online collection analysis and fault prediction system

    CN108873830A

  • Large-scale train shift fault detection method and system based on deep learning

    CN110264440A

  • Vehicle operation process fault diagnosis method and system based on multivariate time series data

    CN113485302A

Cited By

  • Intelligent equipment management system based on cloud platform

    CN119989117A