Remote vehicle fault diagnosis methods, devices, and storage media based on big data

Abstract

This application provides a method, device, and storage medium for remote vehicle fault diagnosis based on big data. The method includes: determining a target vehicle requiring fault diagnosis; acquiring target vehicle fault data; inputting the target vehicle fault data into a vehicle fault diagnosis model to obtain a target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results from multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model using the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data and one fault diagnosis result. The target fault diagnosis result is then pushed to the vehicle diagnostic device. This method enables remote diagnosis of faulty vehicles when a fault occurs during operation, improving the accuracy and speed of vehicle fault diagnosis.

Description

Technical Field

[0001] This application relates to the field of vehicle fault diagnosis technology, and in particular to a method, device and storage medium for remote vehicle fault diagnosis based on big data. Background Technology

[0002] With the rapid development of new energy vehicles, the electronic control systems of modern vehicles have become increasingly complex, making traditional fault diagnosis methods ineffective in addressing current vehicle faults. Traditional remote vehicle fault diagnosis methods primarily rely on data acquisition modules to upload vehicle operating status information to a remote platform for analysis. While traditional methods offer relatively good accuracy in fault diagnosis, the precision of the diagnostic process is limited, and the system's response speed is slow, especially when facing complex and ever-changing fault scenarios, making it difficult to provide timely and effective solutions.

[0003] Therefore, improving the accuracy and speed of vehicle fault diagnosis is an urgent issue that needs to be addressed. Summary of the Invention

[0004] This application provides a method, device, and storage medium for remote vehicle fault diagnosis based on big data, which enables improved remote diagnosis of faulty vehicles when a fault occurs during vehicle operation, thereby improving the accuracy and speed of vehicle fault diagnosis.

[0005] In a first aspect, embodiments of this application provide a remote vehicle fault diagnosis method based on big data, applied to a server, wherein the server is communicatively connected to a vehicle diagnostic device, and the method includes:

[0006] Identify the target vehicle that requires fault diagnosis;

[0007] Obtain the target vehicle fault data;

[0008] The target vehicle fault data is input into the vehicle fault diagnosis model to obtain the target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results of multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model through the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data and one fault diagnosis result.

[0009] The target fault diagnosis results are pushed to the vehicle diagnostic equipment.

[0010] Secondly, embodiments of this application provide a vehicle fault remote diagnostic device based on big data, applied to a server, wherein the server is communicatively connected to vehicle diagnostic equipment, and the device includes a determining unit, an acquiring unit, a calculating unit, and a controlling unit, wherein:

[0011] The determining unit is used to determine the target vehicle that needs fault diagnosis;

[0012] The acquisition unit is used to acquire the target vehicle fault data of the target vehicle;

[0013] The computing unit is used to input the target vehicle fault data into the vehicle fault diagnosis model to obtain the target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results of multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model through the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data and one fault diagnosis result.

[0014] The control unit is used to push the target fault diagnosis result to the vehicle diagnostic equipment.

[0015] Thirdly, embodiments of this application provide a server, including a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the programs include instructions for performing steps in any method of the first aspect of this application.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in any method of the first aspect of this application.

[0017] Fifthly, embodiments of this application provide a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in any method of the first aspect of this application. The computer program product may be a software installation package.

[0018] By implementing the big data-based remote vehicle fault diagnosis method in this application embodiment, the following beneficial effects are achieved:

[0019] The process involves identifying the target vehicle requiring fault diagnosis; acquiring the target vehicle's fault data; inputting the target vehicle's fault data into a vehicle fault diagnosis model to obtain the target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results from multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model using the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data point and one fault diagnosis result. The target fault diagnosis result is then pushed to the vehicle diagnostic device. On one hand, by utilizing a big data platform to acquire large-scale sample vehicle fault data and fault diagnosis results, machine learning methods are used to establish and train a vehicle fault diagnosis model based on the large-scale vehicle data acquired through the big data platform. This allows for remote fault diagnosis of the target vehicle, thereby improving the accuracy of vehicle fault diagnosis. On the other hand, by acquiring the target vehicle's fault data and outputting the fault diagnosis result through the vehicle fault diagnosis model, the powerful computing capabilities of the remote diagnostic platform are utilized, improving the response speed of fault diagnosis and thus enhancing the speed of vehicle fault diagnosis. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is an architecture diagram of a vehicle fault remote diagnosis system based on big data, provided in an embodiment of this application.

[0022] Figure 2 This is a flowchart illustrating a remote vehicle fault diagnosis method based on big data, provided in an embodiment of this application.

[0023] Figure 3 This is a functional unit block diagram of a vehicle fault remote diagnosis device based on big data provided in an embodiment of this application;

[0024] Figure 4 This is a schematic diagram of the structure of a server provided in an embodiment of this application. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0026] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0027] It should be understood that the term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this document indicates that the preceding and following related objects are in an "or" relationship. In the embodiments of this application, "multiple" refers to two or more.

[0028] In the embodiments of this application, "at least one item" or its similar expression refers to any combination of these items, including any combination of a single item or a plurality of items. "One or more" means one or more, while "multiple" means two or more. For example, "at least one item" of a, b, or c can represent the following seven cases: a, b, c; a and b; a and c; b and c; a, b, and c. Each of a, b, and c can be an element or a set containing one or more elements.

[0029] In this application, the term "connection" refers to various connection methods, such as direct connection or indirect connection, to achieve communication between devices. This application does not impose any limitations on this.

[0030] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0031] Current remote vehicle fault diagnosis methods rely on data acquisition modules to upload vehicle operating status information to a remote platform for analysis and to match the vehicle's operating status with a fault diagnosis database. However, the accuracy of the diagnosis process is limited, and the system's response speed is slow. Especially when faced with complex and ever-changing fault scenarios, it is difficult to provide timely and effective solutions.

[0032] To address the aforementioned issues, this application provides a method, apparatus, and storage medium for remote vehicle fault diagnosis based on big data. Applied to a server, the method first identifies the target vehicle requiring fault diagnosis and acquires its fault data. Then, the fault data is input into a vehicle fault diagnosis model to obtain a target fault diagnosis result. This model is obtained by acquiring multiple fault data points and diagnosis results from multiple sample vehicles through a big data platform, and training an initial model using the fault data and diagnosis results. Each sample vehicle corresponds to one fault data point and one diagnosis result. Finally, the target fault diagnosis result is pushed to the vehicle diagnostic device, thereby improving the accuracy and speed of vehicle fault diagnosis.

[0033] The following is combined Figure 1 The architecture of a vehicle fault remote diagnosis system based on big data, as described in this application embodiment, is as follows: Figure 1 This is an architecture diagram of a vehicle fault remote diagnosis system based on big data provided in an embodiment of this application. The vehicle fault remote diagnosis system 100 based on big data includes a fault diagnosis server 110 and a vehicle diagnostic device 120, which transmit data to each other through a wireless communication connection. The fault diagnosis server 110 includes a big data platform 111, a data processing module 112, and a fault diagnosis module 113; the vehicle diagnostic device 120 includes a data acquisition module 121 and a fault display module 122.

[0034] The fault diagnosis server 110 is used to build a vehicle fault diagnosis model, acquire and process the fault data of the target vehicle, output the diagnosis results of the target vehicle through the vehicle fault diagnosis model, and finally push the diagnosis results to the vehicle diagnosis device 120.

[0035] In one possible embodiment, the fault diagnosis server 110 acquires the fault data of the target vehicle that needs to be diagnosed through a wireless communication network. Then, it preprocesses the fault data of the target vehicle to make it conform to the data input format of the vehicle fault diagnosis model. Then, it inputs the preprocessed fault data of the target vehicle into the vehicle fault diagnosis model to obtain the diagnosis result of the target vehicle. Finally, it sends the diagnosis result of the target vehicle to the vehicle diagnosis device through the wireless communication network, and the user obtains the vehicle's fault information through the vehicle diagnosis device.

[0036] Among them, the big data platform 111 is used to provide centralized data processing, including data storage and management, data analysis and mining, and collaborative support functions. The big data platform 111 collects the results of large-scale vehicle faults and diagnoses from the Internet, analyzes and mines them, and sends the collected and analyzed data to the data processing module 112. The data processing module 112 cleans, transforms and standardizes the data from the big data platform 111, and then uses the preprocessed data as input to the fault diagnosis module 113 to train the fault diagnosis model in the fault diagnosis module 113.

[0037] The vehicle diagnostic device 120, used to acquire data from the target vehicle and interact with the user, includes a data acquisition module 121 and a fault display module 122. The data acquisition module 121 collects vehicle operating status data and connects to various vehicle subsystems via a Controller Area Network (CAN) bus, collecting multi-dimensional data on the vehicle's operating status in real time, including information such as temperature, pressure, speed, vibration, and current, which are not limited here. The vehicle diagnostic device 120 wirelessly transmits the vehicle fault data acquired by the data acquisition module 121 to a fault diagnosis server 110. The fault diagnosis server 110 remotely diagnoses the fault data through the fault diagnosis module 113 to obtain the fault diagnosis result of the target vehicle and displays the fault diagnosis result on the fault display module 122 to the user.

[0038] As can be seen, the above system architecture can improve the accuracy and speed of vehicle fault diagnosis.

[0039] After understanding the software and hardware architecture of this application, the following will be combined with... Figure 2 This application describes a remote vehicle fault diagnosis method based on big data in its embodiments. Figure 2 A flowchart illustrating a remote vehicle fault diagnosis method based on big data, provided in this application embodiment, specifically includes the following steps:

[0040] Step S210: Determine the target vehicle for which fault diagnosis is required.

[0041] The target vehicle refers to a vehicle that has experienced a malfunction and requires remote diagnosis. By comprehensively analyzing the target vehicle's operating status, historical maintenance records, and current environmental conditions, it is determined whether the vehicle meets the triggering conditions for fault diagnosis.

[0042] Specifically, the process of identifying target vehicles includes the following aspects: Monitoring operational status involves real-time collection of vehicle operating data using vehicle diagnostic equipment, including engine speed, fuel consumption, braking system status, temperature, current, and voltage. If certain parameters deviate from their normal range, the vehicle is deemed to have a potential fault risk. Alternatively, historical data analysis of the target vehicle's maintenance records and fault frequency data reveals that certain components have frequently failed recently, prioritizing these vehicles as target vehicles.

[0043] It should be noted that the determination of the target vehicle does not rely solely on a single data source, but rather on a more comprehensive judgment obtained through multi-dimensional data fusion and analysis. Furthermore, when multiple vehicles simultaneously meet the trigger conditions for fault diagnosis, the system prioritizes the vehicles according to the severity of the fault. For example, vehicles with obvious sensor malfunctions will be diagnosed first, while vehicles requiring regular maintenance will be addressed in subsequent diagnoses.

[0044] Step S220: Obtain the target vehicle fault data of the target vehicle.

[0045] The target vehicle fault data refers to relevant data that characterizes the fault state of the target vehicle, including but not limited to sensor data, diagnostic codes (DTCs), operating parameters, and historical fault records. This data is collected in real-time by the data acquisition module of the vehicle diagnostic equipment, preprocessed locally, and then uploaded to the fault diagnosis server via a wireless network.

[0046] Specifically, acquiring fault data for the target vehicle mainly involves the following steps: First, real-time data acquisition is performed by monitoring the operating status of key components in real time using in-vehicle sensors (such as temperature sensors, pressure sensors, and current sensors). Examples include engine operating temperature, battery voltage and current, and brake system hydraulic pressure. When certain sensors detect abnormal values, the data is recorded and marked as potential fault data. Next, diagnostic codes are extracted using the vehicle's on-board diagnostics (OBD) system or other fault diagnosis software interfaces to obtain the vehicle's fault codes (DTCs). These DTCs contain specific fault information identified during the vehicle's self-test, such as sensor malfunction or electronic control module abnormality. Then, historical data is correlated by retrieving historical fault records from a big data platform and comparing them with the real-time acquired data to identify the frequency, scope, and possible causes of the faults. For example, if a sensor frequently generates fault signals, it can be further verified whether the sensor is aging or has a poor connection. Finally, the multidimensional data is fused by combining real-time data, diagnostic codes, and historical data. The analysis function of the big data platform is used to perform multidimensional data fusion, clean up invalid or redundant information, and generate comprehensive target vehicle fault data.

[0047] It should be noted that the collection and processing of target vehicle fault data is not limited to data sources inside the vehicle; it can also be combined with environmental data (such as temperature, humidity, and smoke concentration) for comprehensive analysis. For example, by using a smoke sensor to obtain the concentration of smoke particles inside the vehicle and combining it with high-temperature data from a temperature sensor, it may be possible to more accurately determine whether the vehicle is at risk of fire.

[0048] Step S230: Input the target vehicle fault data into the vehicle fault diagnosis model to obtain the target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results of multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model through the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data and one fault diagnosis result.

[0049] The target vehicle fault data includes real-time operating parameters, DTCs (Distributed Troubleshooting), and sensor data, which characterize the target vehicle's current operating status and potential faults. This data is acquired through a large-scale vehicle data acquisition platform, undergoing data preprocessing. A machine learning-based fault diagnosis model is then trained based on the preprocessed data. After obtaining the trained model, data collected via the target vehicle's OBD (On-Board Diagnostics) system is used as input to diagnose the target vehicle's faults. Finally, the fault diagnosis results are sent to the vehicle diagnostic equipment for display, allowing users to understand the current fault information of the vehicle.

[0050] Specifically, the construction and training process of the vehicle fault diagnosis model mainly includes the following steps: First, sample data is collected through a big data platform, gathering a large amount of vehicle fault data and corresponding fault diagnosis results from multiple sample vehicles. Each sample vehicle's data includes fault-related operating parameters and DTCs. Next, a feature matrix is ​​constructed from the data obtained from the big data platform. Based on the fault codes and corresponding first fault frequencies of each sample vehicle, the correlation strength value between the fault codes and the first fault frequencies is determined, resulting in n correlation strength values. These strength values ​​represent the degree of correlation between the fault codes and frequencies. Combining multiple diagnostic result data from the sample vehicles, a first data matrix is ​​constructed. This matrix is ​​m*n dimensional and is used to extract feature data of vehicle faults. Then, feature selection is performed on the first data matrix to remove redundant information and extract key feature datasets (such as highly correlated or highly weighted fault parameters). These key features are used to further train the vehicle fault diagnosis model, improving the model's accuracy and generalization ability. Finally, based on the collected sample vehicle data, an initial vehicle fault diagnosis model is trained using machine learning algorithms (such as deep neural network models). This model can predict the fault category of input data by learning the mapping relationship between sample vehicle fault data and fault diagnosis results.

[0051] Optionally, the multiple vehicle fault data includes n fault codes and n first fault frequencies, where n is an integer greater than 1; the multiple fault diagnosis results include m diagnosis result data, where m is an integer greater than 1 and m is less than or equal to n; and the method further includes the following steps:

[0052] 21. Determine the correlation strength value between the fault code and the first fault frequency to obtain n correlation strength values, wherein the correlation strength value represents the degree of correlation between the fault code and the first fault frequency;

[0053] 22. Construct a first data matrix based on the m diagnostic results and the n correlation strength values, wherein the first data matrix is ​​an m*n dimensional matrix;

[0054] 23. Perform feature selection on the first data matrix to obtain a vehicle fault feature dataset;

[0055] 24. Train the initial vehicle fault diagnosis model based on the vehicle fault feature dataset to obtain the vehicle fault diagnosis model.

[0056] The calculation of association strength values ​​can be achieved in various ways, such as using the BERT (Bidtrectional Encoder Representations from Transformers) model. Specifically, firstly, based on vehicle operating data, a relation extraction model is used to extract the relationship between specific fault codes and their corresponding frequencies. After obtaining the triples representing the association between fault codes and their association strength values, the BERT model outputs the predicted values ​​of the relationships between entity pairs. To illustrate this more clearly, an example is given: First, the fault diagnosis data is preprocessed, including word segmentation and stop word removal, and entities in the data are labeled, namely fault frequencies (e.g., "faults occur 3 times per month") and fault codes (e.g., "P0010"), while simultaneously determining the type of relationship between them, including positive correlation, negative correlation, and no significant correlation. Then, the preprocessed data is input into the BERT model for fine-tuning. During fine-tuning, the labeled entity pairs (fault frequencies and fault codes) and their relationships are used as labels. For example, for the sentence "When the failure frequency reaches 3 times per month, fault code P0010 frequently occurs," the entity pair (fault frequency - fault code) is labeled as (3 times per month - P0010), and the relation label is "positive correlation." Then, the model is trained to learn the features of the relationship between fault frequency and fault code. Finally, in the model's output layer, the fine-tuned BERT outputs a probability distribution regarding the relation type. Assuming the model output vector is [0.6, 0.1, 0.3], this means that for a given fault frequency and fault code entity pair, the model predicts a probability of "positive correlation" of 0.6, a probability of "negative correlation" of 0.1, and a probability of "no significant correlation" of 0.3. The probability of "positive correlation" in the output vector is then determined as the correlation strength value between the fault code and the fault frequency; that is, 0.6 is the correlation strength value corresponding to the fault frequency and fault code. The higher the correlation strength value, the stronger the correlation between the fault code and the first fault frequency.

[0057] Specifically, in the first data matrix, row m represents the fault diagnosis results, and column n represents the correlation strength value between the fault code and the frequency. Each element in the matrix can represent the relationship between a diagnosis result and a specific correlation strength value. To enhance the matrix's ability to express features, the data is standardized during construction to ensure uniform data distribution, thereby improving the feature learning capability based on the neural network model.

[0058] As can be seen, in this embodiment, the calculation and processing of sample data significantly improves the targeting and efficiency of feature selection, ensuring that the constructed fault feature dataset can more effectively characterize the potential characteristics of vehicle faults. Simultaneously, training the initial diagnostic model using the optimized feature dataset enables the model to possess excellent accuracy and robustness when handling complex vehicle faults, thus providing technical support for intelligent vehicle diagnosis and prediction. This allows the model to adapt to various fault scenarios, effectively reducing misdiagnosis and missed diagnosis rates, and improving the level of intelligent vehicle operation and maintenance.

[0059] Optionally, step 23, performing feature selection on the first data matrix to obtain the vehicle fault feature dataset, may include the following steps:

[0060] 231. By calculating the similarity between each pair of rows of data in the first data matrix, a first similarity matrix is ​​obtained. The first similarity matrix is ​​an m*m dimensional matrix.

[0061] 232. Perform sparsity processing on the first similarity matrix to obtain a second similarity matrix, specifically: when the element value of the target element in the first similarity matrix is ​​greater than or equal to a preset first threshold, the target element is retained; otherwise, when the element value of the target element is less than the first threshold, the target element is set to 0; the target element is any element in the first similarity matrix.

[0062] 233. Generate a vehicle fault feature dataset based on the second similarity matrix and the m diagnostic result data.

[0063] The similarity calculation can be performed using various methods, such as cosine similarity, Euclidean distance similarity, or Pearson correlation coefficient, etc., which are not limited here. Each element of the first similarity matrix represents the degree of similarity between two pairs of sample vehicles. The results of the similarity calculation can help identify potential connections between sample data.

[0064] The purpose of sparsity processing is to reduce redundant information in the similarity matrix and improve computational efficiency. The preset first threshold can be set according to actual needs, such as an empirical value determined based on historical data distribution, generally set to 0.3 or 0.5. Through sparsity processing, only highly similar sample pairs are retained, thereby highlighting the key relationships between data, reducing computational complexity, saving computing resources of the fault diagnosis server, and improving the speed of fault diagnosis.

[0065] Specifically, after sparsity processing, the second similarity matrix is ​​combined with the original diagnostic results data to further extract key features from the vehicle fault feature dataset. By comparing and analyzing the sparse data in the second similarity matrix with the diagnostic results, specific fault patterns associated with highly similar samples can be identified, and redundant information can be eliminated, thereby extracting the most valuable features for the diagnostic process. When generating the vehicle fault feature dataset, the non-zero elements in the second similarity matrix are first mapped to the original diagnostic results data to filter out key features corresponding to highly similar samples. Next, clustering algorithms (such as K-means clustering or hierarchical clustering) are used to group the features, further refining the feature distribution corresponding to different fault types. Then, prior information is added to the model; for example, by incorporating domain knowledge, features irrelevant to key fault patterns are removed to ensure the generated dataset has higher representativeness and interpretability.

[0066] As can be seen, by processing the feature selection process through three steps—similarity calculation, sparsity processing, and dataset generation—this embodiment can effectively extract key features of vehicle faults and avoid the waste of computing resources caused by redundant data. In particular, in the sparsity processing step, by setting an appropriate threshold, not only can low-correlation sample pairs be removed, but also high-correlation sample pairs can be accurately retained, thereby achieving optimization in balancing computational efficiency and diagnostic accuracy.

[0067] Optionally, step 233, generating the vehicle fault feature dataset based on the second similarity matrix and the m diagnostic result data, may include the following steps:

[0068] 2331. Construct a diagnostic result matrix based on the m diagnostic result data;

[0069] 2332. Determine the vehicle fault feature dataset based on the second similarity matrix and the diagnostic result matrix.

[0070] The diagnostic result matrix is ​​constructed based on the organization and summarization of historical vehicle diagnostic data. The matrix is ​​an m*p dimensional matrix, where m represents the category of the diagnostic result, p represents the diagnostic attribute description for each result, and each element in the matrix represents the performance value of a specific diagnostic result on a specific diagnostic attribute, such as the frequency of fault occurrence. The diagnostic result matrix can be constructed by directly collecting data records from vehicle diagnostic equipment.

[0071] Specifically, the numerical data in the diagnostic result matrix is ​​first normalized to ensure that diagnostic attributes with different dimensions can be analyzed uniformly, improving the comparability between features. Next, based on the highly similar samples selected from the second similarity matrix, features with low contribution to fault prediction or those that are redundant are removed from the diagnostic result matrix. Then, the feature dataset is further optimized by aggregating or calculating derived features (such as averages, weighted sums, etc.) on the selected diagnostic attributes. In the process of combining the second similarity matrix and the diagnostic result matrix, similarity information can be mapped to diagnostic attributes using matrix multiplication.

[0072] As can be seen, the feature dataset generation step in this embodiment not only improves the correlation between samples through similarity matrix filtering, but also effectively reduces data dimensionality through optimization of the diagnostic result matrix and feature filtering, thereby enhancing the vehicle fault diagnosis model's ability to express key features. This method can balance computational efficiency and diagnostic accuracy in practical applications, providing an efficient and reliable feature extraction method for large-scale vehicle diagnosis. Finally, the generated vehicle fault feature dataset can significantly improve the model's generalization performance during training, reducing errors caused by noisy data or redundant features, thus improving the accuracy and reliability of fault diagnosis.

[0073] Optionally, step 24, training the initial vehicle fault diagnosis model based on the vehicle fault feature dataset to obtain the vehicle fault diagnosis model, may include the following steps:

[0074] 241. Divide the vehicle fault feature dataset to obtain a training set and a test set;

[0075] 242. The initial vehicle fault diagnosis model is trained based on the training set to obtain the first vehicle fault diagnosis model;

[0076] 243. Evaluate the first vehicle fault diagnosis model based on the test set to obtain the model evaluation results;

[0077] 244. Adjust the parameters of the first vehicle fault diagnosis model according to the model evaluation results to obtain the vehicle fault diagnosis model.

[0078] Typically, the feature dataset is divided into training and testing sets according to a certain ratio, such as 70% for training and 30% for testing, or 80% for training and 20% for testing; no specific limit is set here. The training set is used to optimize the model parameters, while the testing set is used to evaluate the model's generalization ability.

[0079] The first vehicle fault diagnosis model is evaluated using a test set to obtain the model evaluation results. Evaluation metrics may include, but are not limited to, accuracy, recall, precision, F1-scores, and mean squared error. For example, for a classification model, the classification performance can be analyzed using a confusion matrix.

[0080] The model evaluation results are used to adjust the parameters of the first vehicle fault diagnosis model to optimize the final vehicle fault diagnosis model. Specific adjustments include tuning the model hyperparameters and re-optimizing the training parameters. For example, the optimal combination of hyperparameters can be found through methods such as grid search, random search, or Bayesian optimization, or the model performance can be further optimized by increasing or decreasing the number of training epochs or adjusting the learning rate.

[0081] Specifically, during model training and evaluation, cross-validation techniques are used to split and train the dataset multiple times. For example, k-fold cross-validation divides the dataset into k subsets, takes one subset as the test set each time, and uses the remaining subsets as the training set, repeating the training and evaluation process k times. Finally, the average of all evaluation results is used as the model performance metric.

[0082] Furthermore, the evaluation results on the test set can reveal potential shortcomings of the model. For example, if the recall rate for certain fault categories is low, specific data samples can be added for these categories or the weights of the loss function can be adjusted to balance the model's predictive power across different categories.

[0083] As can be seen, in this embodiment, the training process of the vehicle fault diagnosis model, through a combination of data partitioning, model training, performance evaluation, and parameter adjustment, achieves the transformation of the model from its initial state to its optimized state. By selecting evaluation metrics and fine-tuning parameters, the model's performance on different datasets can be ensured to be stable and reliable, effectively improving the accuracy, robustness, and practicality of fault diagnosis. This training method is particularly suitable for scenarios where vehicle diagnostic data is complex and unevenly distributed in practical applications, providing accurate and efficient technical support for intelligent vehicle fault diagnosis.

[0084] Optionally, step S230, which involves inputting the target vehicle fault data into the vehicle fault diagnosis model to obtain the target fault diagnosis result, may include the following steps:

[0085] A1. Obtain the vehicle information of the target vehicle;

[0086] A2. Determine the maintenance and repair information in the vehicle information;

[0087] A3. Determine the weight of each fault in the target vehicle fault data based on the maintenance information to obtain the first weight set;

[0088] A4. Adjust the target vehicle fault data according to the first weight set to obtain the first enhanced vehicle fault data;

[0089] A5. Preprocess the first enhanced vehicle fault data according to the input data format of the vehicle fault diagnosis model to obtain the second enhanced vehicle fault data.

[0090] A6. Input the second enhanced vehicle fault data into the vehicle fault diagnosis model to obtain the target fault diagnosis result.

[0091] Obtaining vehicle information specifically refers to acquiring relevant vehicle operating parameters and status information from the target vehicle's control unit via OBD or other diagnostic software interfaces. This information includes, but is not limited to, vehicle model, mileage, engine operating status, fuel consumption, driving habits, and environmental factors (such as temperature and humidity).

[0092] Maintenance and repair information can be obtained from the vehicle's maintenance records, including regular maintenance information, historical fault records, and information on repaired and replaced parts. The vehicle's maintenance and repair history can help determine which faults may be caused by a lack of maintenance over a long period or wear and tear on certain components.

[0093] Preprocessing includes steps such as data standardization, outlier removal, and missing data imputation. Standardization ensures all input data have the same units of measurement, preventing large variables from excessively influencing model training. Outlier removal and missing data imputation improve data quality and avoid the impact of data inconsistencies on fault diagnosis results.

[0094] Specifically, the weights corresponding to faults are determined based on maintenance information, resulting in a first weight set. The weights are set based on factors such as the historical frequency of vehicle faults and the timeliness of maintenance. Specifically, certain faults, such as engine faults and braking system faults, may have higher importance or frequency, and therefore should be assigned a larger weight in the weight set. Next, the fault data of the target vehicle is adjusted according to the first weight set to obtain first enhanced vehicle fault data. This adjustment process typically involves weighted averaging, weighted summation, or directly adjusting the fault data values ​​according to the set weights, so that fault information with higher weights receives more attention in diagnosis. Then, the diagnostic information typically includes the fault type judgment and its probability of occurrence or severity assessment. These results can be obtained by processing and predicting the input data using machine learning models. The target diagnostic information not only includes the possible fault types of the vehicle but also provides possible causes and solutions for each fault.

[0095] It is evident that by combining historical vehicle maintenance information, weighted adjustments, and data preprocessing, the diagnostic results can more accurately reflect the current fault status of the target vehicle. This effectively reduces errors caused by differences in vehicle data and changes in the external environment, improving the real-time nature and accuracy of vehicle fault diagnosis. This not only provides effective support for vehicle maintenance but also offers timely fault warnings to vehicle owners, preventing potentially serious malfunctions.

[0096] Optionally, step A3, determining the weight corresponding to each fault in the target vehicle fault data based on the maintenance information to obtain a first weight set, may include the following steps:

[0097] A31. Based on the maintenance information, count the number of times each fault occurs to obtain the first fault count set;

[0098] A32. Determine the second fault frequency corresponding to each fault in the first fault frequency set to obtain the fault frequency set;

[0099] A33. Based on the preset fault frequency and weight mapping relationship, determine the weight corresponding to each second fault frequency in the fault frequency set to obtain the first weight set.

[0100] The maintenance and repair information includes a statistical analysis of the frequency of each fault occurrence, resulting in a first fault frequency set. The frequency of fault occurrence reflects its severity and is a crucial basis for determining diagnostic priority. For example, if a fault occurs repeatedly, it suggests that it may be a common vehicle problem or closely related to factors such as the vehicle's operating environment and driving habits. Statistical analysis of these fault occurrence frequencies helps assign appropriate weights to the fault data, improving the accuracy and relevance of the fault diagnosis system.

[0101] This involves establishing a mapping rule based on historical data or empirical models to clarify the relationship between fault frequency and weight values. For example, machine learning algorithms or manually defined rules can be used to assign higher weights to frequently occurring faults and lower weights to occasionally occurring faults. This weight setting ensures that the model focuses more on frequently occurring and impactful faults during diagnosis, avoiding overemphasis on rare faults, thereby improving diagnostic efficiency and accuracy.

[0102] Specifically, the fault frequency is calculated based on the first fault occurrence set, typically normalized over a time window. The fault frequency considers not only the number of times a fault occurs but also the time interval between occurrences and vehicle usage. For example, if a fault occurs frequently within a short period, its fault frequency is high; if it occurs over a longer period, its frequency can be appropriately reduced. This calculation further refines the fault frequency, thereby improving the fault diagnosis model's ability to determine the importance of different faults. Finally, the final first weight set is used, and a reasonable weight is assigned to each fault, allowing the system to prioritize more serious or common faults based on their frequency and importance during fault diagnosis.

[0103] Step S240: Push the target fault diagnosis result to the vehicle diagnostic device.

[0104] The target fault diagnosis result is calculated based on the vehicle fault diagnosis model and is used to characterize the current fault category and potential risks of the target vehicle. The fault display module on the vehicle diagnostic equipment presents the target fault diagnosis result to users or maintenance technicians in an intuitive and easy-to-understand manner.

[0105] Specifically, after the vehicle fault diagnosis model outputs the target fault diagnosis result, the result data (such as fault category, fault location, severity, etc.) is formatted to ensure compatibility with the display standards of different vehicle diagnostic devices. For example, complex multidimensional diagnostic results can be organized into forms, charts, or alarm information. Then, the diagnostic results are pushed to the vehicle diagnostic device via wireless communication (such as 4G / 5G, V2XEvolution). During the push process, the data can be encrypted to ensure the security and integrity of the diagnostic information. The fault display interface of the vehicle diagnostic device visualizes the received target fault diagnosis result. For example, the severity of the fault can be distinguished by color (green for normal, yellow for warning, and red for severe), or an interactive interface can be provided to display detailed diagnostic information and suggested repair solutions; no specific limitations are imposed here.

[0106] In addition, to ensure fast and stable data transmission between the vehicle diagnostic equipment and the fault diagnosis server, a dual-channel redundant transmission technology was designed. Even if the main channel fails, the data can still be seamlessly switched through the backup channel to ensure that fault data can be transmitted in real time.

[0107] It should be noted that the target fault diagnosis results can not only be transmitted to the in-vehicle equipment for display, but also be simultaneously displayed via mobile terminals (such as smartphones and tablets) or remote monitoring platforms. Real-time push notifications and simultaneous display across multiple terminals improve the efficiency and convenience of information transmission.

[0108] As can be seen, the above method can push the target fault diagnosis results to the vehicle diagnostic equipment and display them, which can quickly guide users to identify the fault type and provide clear guidance for subsequent maintenance or operation.

[0109] The above primarily describes the solutions of the embodiments of this application from the perspective of the method execution process. It is understood that, in order to achieve the above functions, the server includes the corresponding hardware structure and / or software modules for executing each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in the embodiments provided herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0110] This application embodiment can divide the server into functional units according to the above method example. For example, each function can be divided into a separate functional unit, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0111] When dividing each function into modules according to its corresponding function. Figure 3 This application provides a functional unit block diagram of a vehicle fault remote diagnosis device based on big data. The vehicle fault remote diagnosis device 300 based on big data is applied to a server, which is communicatively connected to a vehicle diagnostic device, and includes:

[0112] Determining unit 310 is used to determine the target vehicle that requires fault diagnosis;

[0113] Acquisition unit 320 is used to acquire target vehicle fault data of the target vehicle;

[0114] The calculation unit 330 is used to input the target vehicle fault data into the vehicle fault diagnosis model to obtain the target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results of multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model through the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data and one fault diagnosis result.

[0115] Control unit 340 is used to push the target fault diagnosis result to the vehicle diagnostic equipment.

[0116] In one possible embodiment, where the plurality of vehicle fault data includes n fault codes and n first fault frequencies, where n is an integer greater than 1, and the plurality of fault diagnosis results include m diagnosis result data, where m is an integer greater than 1 and m is less than or equal to n, the calculation unit 330 is specifically used for:

[0117] Determine the correlation strength value between the fault code and the first fault frequency to obtain n correlation strength values, where the correlation strength value represents the degree of correlation between the fault code and the first fault frequency;

[0118] A first data matrix is ​​constructed based on the m diagnostic results and the n correlation strength values. The first data matrix is ​​an m*n dimensional matrix.

[0119] Feature selection is performed on the first data matrix to obtain a vehicle fault feature dataset;

[0120] The initial vehicle fault diagnosis model is trained based on the vehicle fault feature dataset to obtain the vehicle fault diagnosis model.

[0121] In one possible embodiment, in performing feature selection on the first data matrix to obtain a vehicle fault feature dataset, the computing unit 330 is specifically used for:

[0122] The first similarity matrix is ​​obtained by calculating the similarity between each pair of rows in the first data matrix. The first similarity matrix is ​​an m*m dimensional matrix.

[0123] The first similarity matrix is ​​subjected to sparsity processing to obtain a second similarity matrix. Specifically, if the element value of the target element in the first similarity matrix is ​​greater than or equal to a preset first threshold, the target element is retained; otherwise, if the element value of the target element is less than the first threshold, the target element is set to 0. The target element is any element in the first similarity matrix.

[0124] A vehicle fault feature dataset is generated based on the second similarity matrix and the m diagnostic results data.

[0125] In one possible embodiment, in generating a vehicle fault feature dataset based on the second similarity matrix and the m diagnostic result data, the computing unit 330 is specifically used for:

[0126] Construct a diagnostic result matrix based on the m diagnostic result data;

[0127] The vehicle fault feature dataset is determined based on the second similarity matrix and the diagnostic result matrix.

[0128] In one possible embodiment, in training the initial vehicle fault diagnosis model based on the vehicle fault feature dataset to obtain the vehicle fault diagnosis model, the computing unit 330 is specifically used for:

[0129] The vehicle fault feature dataset is divided into a training set and a test set;

[0130] The initial vehicle fault diagnosis model is trained based on the training set to obtain the first vehicle fault diagnosis model.

[0131] The first vehicle fault diagnosis model is evaluated based on the test set to obtain the model evaluation results;

[0132] Based on the model evaluation results, the parameters of the first vehicle fault diagnosis model are adjusted to obtain the vehicle fault diagnosis model.

[0133] In one possible embodiment, the calculation unit 330 is specifically used for inputting the target vehicle fault data into the vehicle fault diagnosis model to obtain the target fault diagnosis result:

[0134] Obtain the vehicle information of the target vehicle;

[0135] Determine the maintenance and repair information from the vehicle information;

[0136] Based on the maintenance information, the weight corresponding to each fault in the target vehicle fault data is determined to obtain a first weight set;

[0137] The target vehicle fault data is adjusted according to the first weight set to obtain the first enhanced vehicle fault data;

[0138] The first enhanced vehicle fault data is preprocessed according to the input data format of the vehicle fault diagnosis model to obtain the second enhanced vehicle fault data.

[0139] The second enhanced vehicle fault data is input into the vehicle fault diagnosis model to obtain the target fault diagnosis result.

[0140] In one possible embodiment, in determining the weight corresponding to each fault in the target vehicle fault data based on the maintenance information to obtain a first weight set, the calculation unit 330 is specifically used for:

[0141] The number of occurrences of each fault is counted based on the maintenance information to obtain the first fault count set;

[0142] Determine the second fault frequency corresponding to each fault in the first fault frequency set to obtain the fault frequency set;

[0143] Based on the preset fault frequency and weight mapping relationship, the weight corresponding to each second fault frequency in the fault frequency set is determined to obtain the first weight set.

[0144] It should be noted that the specific functional implementation of the big data-based remote vehicle fault diagnosis device 300 is described above. Figure 2 The description of the big data-based remote vehicle fault diagnosis method illustrates, for example, the determination of the relevant content of the execution of S210 by unit 310. The various units or modules in the big data-based remote vehicle fault diagnosis device 300 can be individually or entirely merged into one or more other units or modules, or some of the units or modules can be further divided into multiple functionally smaller units or modules. This achieves the same operation without affecting the technical effect of the embodiments of the present invention. The aforementioned units or modules are based on logical functional division. In practical applications, the function of one unit (or module) is implemented by multiple units (or modules), or the function of multiple units (or modules) is implemented by one unit (or module). The specific implementation of each operation can adopt the corresponding description of the method embodiments shown above. The big data-based remote vehicle fault diagnosis device 300 can be used to execute the above method embodiments of this application, and will not be described again here.

[0145] As can be seen, the vehicle fault remote diagnosis device based on big data described in this application first determines the target vehicle requiring fault diagnosis, then acquires the target vehicle's fault data, and then inputs the target vehicle fault data into a vehicle fault diagnosis model to obtain the target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results from multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model using the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data and one fault diagnosis result. Finally, the target fault diagnosis result is pushed to the vehicle diagnosis device. In this way, by acquiring large-scale sample vehicle fault data and fault diagnosis results through a big data platform, establishing and training a vehicle fault diagnosis model based on the large-scale vehicle data acquired through the big data platform using machine learning methods, and using the vehicle fault diagnosis model to perform remote fault diagnosis on the target vehicle, the accuracy and speed of vehicle fault diagnosis are improved.

[0146] The following is combined Figure 4 The server in the embodiments of this application will be described. Figure 4This application provides a schematic diagram of the structure of a server, as shown in the embodiment of the present application. Figure 4 As shown, the server 400 includes one or more processors 410, a memory 420, a communication interface 430, and one or more programs 421. The processors are connected to the memory and the communication interface via an internal communication bus.

[0147] The processor 410 may be, for example, a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, cells, and circuits described in conjunction with the disclosure of this application. The processor may also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0148] The memory 420 can be volatile memory such as dynamic random access memory (DRAM) or non-volatile memory such as a hard disk drive (HDD). The memory 420 stores a set of executable program code, and the processor 410 calls one or more programs 421 stored in the memory 420 to execute some or all of the steps of the method described in the above embodiment of the vehicle fault remote diagnosis method based on big data. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0149] The communication interface 430 can be a transceiver, transceiver circuit, etc., and is not limited here.

[0150] It is understood that the server may include more or fewer structural elements than those shown in the block diagram above, such as a power module, physical buttons, a Wi-Fi module, a speaker, a Bluetooth module, sensors, and a display module, etc., without limitation. It is understood that the server may be equipped with... Figure 1 The system architecture described above.

[0151] This application also provides a computer-readable storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes a server.

[0152] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments. The computer program product may be a software installation package, and the computer may include a server.

[0153] It should be noted that, for the sake of simplicity, the above embodiments are all described as a series of actions. Those skilled in the art should understand that this application is not limited to the described order of actions, as some steps in the embodiments of this application can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions, steps, modules, or units involved are not necessarily essential to the embodiments of this application.

[0154] In the above embodiments, the descriptions of each embodiment in this application have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0155] 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. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

[0156] The steps of the methods or algorithms described in the embodiments of this application can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in RAM, flash memory, ROM, EPROM, electrically erasable programmable read-only memory (EEPROM), registers, hard disk, portable hard disk, read-only optical disk (CD-ROM), or any other form of storage medium well known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and storage medium can reside in an ASIC. Furthermore, the ASIC can reside in a terminal device or management device. Alternatively, the processor and storage medium can exist as discrete components in the terminal device or management device.

[0157] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in the embodiments of this application can be implemented, in whole or in part, by software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, in the form of a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., digital video discs (DVDs)), or semiconductor media (e.g., solid-state disks (SSDs)).

[0158] The modules / units included in the various devices and products described in the above embodiments can be software modules / units, hardware modules / units, or a combination of both. For example, for devices and products applied to or integrated into a chip, all modules / units can be implemented using hardware methods such as circuits, or at least some modules / units can be implemented using software programs that run on a processor integrated within the chip, while the remaining (if any) modules / units can be implemented using hardware methods such as circuits. For devices and products applied to or integrated into a chip module, all modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components of the chip module, or at least some modules / units can be implemented using hardware methods such as circuits. The implementation is achieved through a software program that runs on a processor integrated within the chip module. The remaining modules / units (if any) can be implemented using hardware methods such as circuits. For various devices and products applied to or integrated into terminal equipment, each of their modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components within the terminal equipment. Alternatively, at least some modules / units can be implemented using a software program that runs on a processor integrated within the terminal equipment, while the remaining modules / units (if any) can be implemented using hardware methods such as circuits.

[0159] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the embodiments of this application. It should be understood that the above descriptions are merely specific embodiments of the embodiments of this application and are not intended to limit the protection scope of the embodiments of this application. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solutions of the embodiments of this application should be included within the protection scope of the embodiments of this application.

Claims

1. A remote vehicle fault diagnosis method based on big data, characterized in that, Applied to a server, the server being communicatively connected to vehicle diagnostic equipment, the method includes: Identify the target vehicle that requires fault diagnosis; Obtain the target vehicle fault data; The target vehicle fault data is input into the vehicle fault diagnosis model to obtain the target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results of multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model through the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data and one fault diagnosis result. The target fault diagnosis result is pushed to the vehicle diagnostic equipment; The multiple vehicle fault data include n fault codes and n first fault frequencies, where n is an integer greater than 1; the multiple fault diagnosis results include m diagnosis result data, where m is an integer greater than 1 and m is less than or equal to n; the method further includes: Determine the correlation strength value between the fault code and the first fault frequency to obtain n correlation strength values, where the correlation strength value represents the degree of correlation between the fault code and the first fault frequency; A first data matrix is ​​constructed based on the m diagnostic results and the n correlation strength values. The first data matrix is ​​an m*n dimensional matrix. Feature selection is performed on the first data matrix to obtain a vehicle fault feature dataset. This process includes: calculating the similarity between pairs of rows in the first data matrix to obtain a first similarity matrix, where the first similarity matrix is ​​an m*m dimensional matrix; performing sparsity processing on the first similarity matrix to obtain a second similarity matrix, specifically: retaining the target element if its value is greater than or equal to a preset first threshold, otherwise setting the target element to 0 if its value is less than the first threshold; the target element is any element in the first similarity matrix; and generating the vehicle fault feature dataset based on the second similarity matrix and the m diagnostic result data. The initial vehicle fault diagnosis model is trained based on the vehicle fault feature dataset to obtain the vehicle fault diagnosis model.

2. The method as described in claim 1, characterized in that, The process of generating a vehicle fault feature dataset based on the second similarity matrix and the m diagnostic result data includes: Construct a diagnostic result matrix based on the m diagnostic result data; The vehicle fault feature dataset is determined based on the second similarity matrix and the diagnostic result matrix.

3. The method as described in claim 1, characterized in that, The step of training the initial vehicle fault diagnosis model based on the vehicle fault feature dataset to obtain the vehicle fault diagnosis model includes: The vehicle fault feature dataset is divided into a training set and a test set; The initial vehicle fault diagnosis model is trained based on the training set to obtain the first vehicle fault diagnosis model. The first vehicle fault diagnosis model is evaluated based on the test set to obtain the model evaluation results; Based on the model evaluation results, the parameters of the first vehicle fault diagnosis model are adjusted to obtain the vehicle fault diagnosis model.

4. The method according to any one of claims 1-3, characterized in that, The step of inputting the target vehicle fault data into the vehicle fault diagnosis model to obtain the target fault diagnosis result includes: Obtain the vehicle information of the target vehicle; Determine the maintenance and repair information from the vehicle information; Based on the maintenance information, the weight corresponding to each fault in the target vehicle fault data is determined to obtain a first weight set; The target vehicle fault data is adjusted according to the first weight set to obtain the first enhanced vehicle fault data; The first enhanced vehicle fault data is preprocessed according to the input data format of the vehicle fault diagnosis model to obtain the second enhanced vehicle fault data. The second enhanced vehicle fault data is input into the vehicle fault diagnosis model to obtain the target fault diagnosis result.

5. The method as described in claim 4, characterized in that, The step of determining the weight corresponding to each fault in the target vehicle fault data based on the maintenance information to obtain a first weight set includes: The number of occurrences of each fault is counted based on the maintenance information to obtain the first fault count set; Determine the second fault frequency corresponding to each fault in the first fault frequency set to obtain the fault frequency set; Based on the preset fault frequency and weight mapping relationship, the weight corresponding to each second fault frequency in the fault frequency set is determined to obtain the first weight set.

6. A remote vehicle fault diagnosis device based on big data, characterized in that, Applied to a server, the server is communicatively connected to vehicle diagnostic equipment. The device includes a determining unit, an acquiring unit, a calculating unit, and a controlling unit, wherein: The determining unit is used to determine the target vehicle that needs fault diagnosis; The acquisition unit is used to acquire the target vehicle fault data of the target vehicle; The computing unit is used to input the target vehicle fault data into the vehicle fault diagnosis model to obtain the target fault diagnosis result. The vehicle fault diagnosis model is obtained by acquiring multiple vehicle fault data and multiple fault diagnosis results of multiple sample vehicles through a big data platform, and training an initial vehicle fault diagnosis model through the multiple vehicle fault data and the multiple fault diagnosis results. Each sample vehicle corresponds to one vehicle fault data and one fault diagnosis result. The control unit is used to push the target fault diagnosis result to the vehicle diagnostic equipment; The multiple vehicle fault data includes n fault codes and n first fault frequencies, where n is an integer greater than 1. The multiple fault diagnosis results include m diagnosis result data, where m is an integer greater than 1 and m is less than or equal to n. Determine the correlation strength value between the fault code and the first fault frequency to obtain n correlation strength values, where the correlation strength value represents the degree of correlation between the fault code and the first fault frequency; A first data matrix is ​​constructed based on the m diagnostic results and the n correlation strength values. The first data matrix is ​​an m*n dimensional matrix. Feature selection is performed on the first data matrix to obtain a vehicle fault feature dataset. This process includes: calculating the similarity between pairs of rows in the first data matrix to obtain a first similarity matrix, where the first similarity matrix is ​​an m*m dimensional matrix; performing sparsity processing on the first similarity matrix to obtain a second similarity matrix, specifically: retaining the target element if its value is greater than or equal to a preset first threshold, otherwise setting the target element to 0 if its value is less than the first threshold; the target element is any element in the first similarity matrix; and generating the vehicle fault feature dataset based on the second similarity matrix and the m diagnostic result data. The initial vehicle fault diagnosis model is trained based on the vehicle fault feature dataset to obtain the vehicle fault diagnosis model.

7. A server, characterized in that, include: Processor, memory, communication interface, and one or more programs; The one or more programs are stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps of the method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1-5.

Images