Vehicle health monitoring and early warning method, device, equipment and storage medium

Abstract

The invention discloses a vehicle health monitoring and early warning method, device and equipment and a storage medium, and relates to the technical field of big data processing, and the method comprises the steps: obtaining system data; performing factor analysis on the system data to obtain common factor data; and training, optimizing and evaluating a prediction model according to the common factor data to obtain a target model so as to carry out vehicle health monitoring and early warning. According to the invention, the efficiency and accuracy of vehicle health monitoring and early warning can be improved.

Description

Technical Field

[0001] The present application relates to the field of big data processing technology, and in particular to vehicle health monitoring and early warning methods, devices, equipment and storage media. Background Art

[0002] With the rapid development of intelligent transportation and Internet of Things technologies, vehicle health monitoring and early warning have become an important means to ensure driving safety and efficient vehicle operation. Traditional vehicle health monitoring and early warning methods mostly rely on regular maintenance and manual inspection, which have problems such as low efficiency and insufficient accuracy. In recent years, vehicle health monitoring methods based on big data and machine learning have gradually attracted attention, among which the XGBoost algorithm has been widely used due to its high efficiency and accuracy. However, directly applying raw data to the XGBoost model may lead to problems such as high model complexity, large amount of calculation and unstable prediction effect. How to fully consider the impact of factors such as feature correlation, data dimension and model training complexity on the model monitoring and early warning effect when applying the XGBoost model is a technical problem currently faced. Summary of the invention

[0003] The main purpose of this application is to provide a vehicle health monitoring and early warning method, device, equipment and storage medium, aiming to solve the technical problem of how to improve the efficiency and accuracy of vehicle health monitoring and early warning.

[0004] To achieve the above objectives, the present application proposes a vehicle health monitoring and early warning method, the method comprising:

[0005] Get system data;

[0006] Performing factor analysis on the system data to obtain common factor data;

[0007] The prediction model is trained, optimized and evaluated according to the common factor data to obtain a target model for vehicle health monitoring and early warning.

[0008] In one embodiment, the step of performing factor analysis on the system data to obtain common factor data includes:

[0009] Representing the sample features in the system data with letters;

[0010] Factor analysis is performed on the sample characteristics represented by letters to obtain common factor data.

[0011] In one embodiment, the step of performing factor analysis on the sample characteristics represented by letters to obtain common factor data includes:

[0012] Classify the sample features represented by letters according to preset standards to obtain classification results;

[0013] Establishing a weight coefficient matrix and an eigenvector matrix, and calculating weight coefficient values ​​according to the weight coefficient matrix and the eigenvector matrix;

[0014] According to the classification results and the weight coefficient values, the linear combination values ​​of the sample features are calculated respectively, and the linear combination values ​​are used as common factor data.

[0015] In one embodiment, the step of training, optimizing and evaluating the prediction model according to the common factor data to obtain a target model for vehicle health monitoring and early warning includes:

[0016] Dividing the common factor data into training set data, validation set data and test set data according to a preset ratio;

[0017] Training the prediction model according to the training set data to obtain a training model;

[0018] Based on the validation set data and the test set data, the training model is Bayesian optimized and evaluated to obtain a target model for vehicle health monitoring and early warning.

[0019] In one embodiment, the step of performing Bayesian optimization and evaluation on the training model based on the validation set data and the test set data to obtain a target model for vehicle health monitoring and early warning includes:

[0020] Initializing the proxy function and the prior distribution based on the validation set data;

[0021] Upsampling the proxy function to obtain a first sampling point;

[0022] Calculate a posterior distribution based on the first sampling point and the prior distribution;

[0023] Based on the posterior distribution, the proxy function is maximized to obtain a second sampling point, and according to the second sampling point, an objective function is obtained;

[0024] Returning to the step of initializing the proxy function and the prior distribution based on the validation set data until the change in the objective function satisfies the preset conditions, thereby obtaining an optimized objective function;

[0025] According to the optimized objective function, a training model under target hyperparameters is obtained;

[0026] Evaluate the training model under the target hyperparameters based on the test set data to obtain an evaluation result;

[0027] When the evaluation result does not meet the preset requirements, return to the step of obtaining system data;

[0028] When the evaluation result meets the preset requirements, completing the model evaluation;

[0029] The training model under the target hyperparameters is used as the target model to perform vehicle health monitoring and early warning.

[0030] In one embodiment, the step of using the training model under the target hyperparameter as the target model to perform vehicle health monitoring and early warning includes:

[0031] According to the target model, the vehicle health is monitored and predicted to obtain monitoring results;

[0032] When the monitoring result of the target model is that the vehicle health is abnormal, a vehicle health warning is issued;

[0033] When the monitoring result of the target model is that the vehicle is in normal health, the vehicle health warning is not issued.

[0034] In one embodiment, the step of obtaining system data includes:

[0035] Acquiring raw data, wherein the raw data includes a user's driving behavior, vehicle data, and environmental data;

[0036] The raw data is cleaned and preprocessed to obtain system data.

[0037] In addition, to achieve the above purpose, the present application also proposes a vehicle health monitoring and early warning device, the device comprising:

[0038] A data acquisition module, used to acquire system data;

[0039] A data analysis module, used for performing factor analysis on the system data to obtain common factor data;

[0040] The model processing and application module is used to train, optimize and evaluate the prediction model according to the common factor data to obtain the target model for vehicle health monitoring and early warning.

[0041] In addition, to achieve the above-mentioned purpose, the present application also proposes a vehicle health monitoring and early warning device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the vehicle health monitoring and early warning method as described above.

[0042] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the vehicle health monitoring and early warning method described above are implemented.

[0043] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the vehicle health monitoring and early warning method as described above.

[0044] One or more technical solutions proposed in this application have at least the following technical effects:

[0045] By reducing data complexity through factor analysis and establishing a Bayesian optimized XGBoost model, the problems of manual feature selection and data preprocessing for vehicle health monitoring and early warning based on a single machine learning algorithm are solved, which is labor-intensive and easy to introduce subjective errors. The original data is directly applied when making predictions using the XGBoost model, resulting in high model complexity, large amount of calculation and unstable prediction results. Compared with the existing technology, the efficiency and accuracy of vehicle health monitoring and early warning are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0048] Figure 1 A flow chart of the first embodiment of the vehicle health monitoring and early warning method of the present application;

[0049] Figure 2 A system architecture diagram of the vehicle health monitoring and early warning method provided in Example 1 of the present application;

[0050] Figure 3 A flow chart of the second embodiment of the vehicle health monitoring and early warning method of the present application;

[0051] Figure 4 A flow chart of the vehicle health monitoring and early warning method embodiment 3 provided in this application;

[0052] Figure 5 A flowchart of the fourth embodiment of the vehicle health monitoring and early warning method of the present application is provided;

[0053] Figure 6 A schematic diagram of the original data composition of the vehicle health monitoring and early warning method provided in Example 4 of the present application;

[0054] Figure 7 A schematic diagram of vehicle health monitoring and early warning related data of the vehicle health monitoring and early warning method provided in Example 4 of the present application;

[0055] Figure 8 This is a schematic diagram of the module structure of the vehicle health monitoring and early warning device according to an embodiment of the present application;

[0056] Fig. 9 This is a schematic diagram of the device structure of the hardware operating environment involved in the vehicle health monitoring and early warning method in the embodiment of the present application.

[0057] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0058] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0059] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0060] The main solution of the embodiment of the present application is: obtaining system data; performing factor analysis on the system data to obtain common factor data; training, optimizing and evaluating the prediction model based on the common factor data to obtain the target model for vehicle health monitoring and early warning.

[0061] In this embodiment, for ease of description, the following description is made taking the internal actuator of the vehicle health monitoring and early warning system as the execution subject.

[0062] Since vehicle health monitoring and early warning based on a single machine learning algorithm in the existing technology often requires manual feature selection and data preprocessing, the workload is large and it is easy to introduce subjective errors. When using the XGBoost model for prediction, the original data is directly applied, resulting in high model complexity, large amount of calculation and unstable prediction effect.

[0063] This application provides a solution to reduce data complexity through factor analysis and establish a Bayesian optimized XGBoost model to improve the efficiency and accuracy of vehicle health monitoring and early warning.

[0064] It can be seen from the above embodiments that the present application reduces data complexity through factor analysis and establishes a Bayesian optimized XGBoost model, thereby overcoming the manual feature selection and data preprocessing of vehicle health monitoring and early warning based on a single machine learning algorithm, which is labor-intensive and prone to introduce subjective errors, and directly applies the original data when using the XGBoost model for prediction, resulting in technical problems such as high model complexity, large amount of calculation and unstable prediction effect, thereby achieving improved efficiency and accuracy of vehicle health monitoring and early warning.

[0065] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, etc. The following takes the internal actuator of the vehicle health monitoring and early warning system as an example to illustrate this embodiment and the following embodiments.

[0066] Based on this, the embodiment of the present application provides a vehicle health monitoring and early warning method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the vehicle health monitoring and early warning method of the present application.

[0067] In this embodiment, the vehicle health monitoring and early warning method includes steps S10 to S30:

[0068] Step S10, obtaining system data.

[0069] It should be noted that the system data is raw data that has been cleaned and preprocessed, including user driving behavior data, vehicle data, and environmental data.

[0070] Step S20, performing factor analysis on the system data to obtain common factor data.

[0071] It should be noted that factor analysis is a statistical method that is used to study the correlation between variables and simplify multiple variables into several representative factors. These factors are unobservable variables that reflect the underlying structure of the data. The main purpose of factor analysis is to reduce the complexity of the data while retaining the most important information. Common factor data are unobservable latent variables extracted through factor analysis, which can explain most of the variance between the original variables.

[0072] Step S30, training, optimizing and evaluating the prediction model according to the common factor data to obtain a target model for vehicle health monitoring and early warning.

[0073] It should be noted that the prediction model is the XGBoost model. XGBoost (eXtreme Gradient Boosting) is an optimized distributed gradient boosting library based on the Gradient Boosting Tree (GBT). It has performed multiple optimizations and extensions on the basis of GBT to improve the performance and efficiency of the model. The XGBoost model is widely used in various machine learning tasks due to its excellent prediction performance and computational efficiency. The target model is an XGBoost model whose prediction rate meets certain requirements, such as a prediction accuracy of 80%, a prediction accuracy of 90%, etc.

[0074] The common factor data is input into the prediction model XGBoost model for training, the training results are optimized, and it is evaluated whether the optimized model can meet the preset requirements. The model that meets the requirements is used as the target model for subsequent vehicle health monitoring and early warning.

[0075] Reference Figure 2 , Figure 2 A system architecture diagram of a vehicle health monitoring and early warning method is provided, wherein the system comprises: a vehicle raw data acquisition module; a raw data cleaning and preprocessing module; a data factor analysis module; an XGBoost classification model training unit; a Bayesian optimization module of the classification model; and a vehicle health monitoring and early warning module based on the classification model.

[0076] The vehicle raw data acquisition module is used to obtain some raw data information of users from multiple data sources of the Internet of Vehicles. The raw data cleaning and preprocessing module screens and processes the raw data. The data factor analysis module is used to reduce the data dimension. The XGBoost classification model training unit is used to train the XGBoost model through the processed data. The Bayesian optimization module of the classification model is used to optimize the selection of XGBoost model parameters. The vehicle health monitoring and early warning module based on the classification model is used to perform vehicle health monitoring and early warning.

[0077] This embodiment provides a vehicle health monitoring and early warning method. By reducing data complexity through factor analysis and establishing a Bayesian optimized XGBoost model, it overcomes the technical problems of manual feature selection and data preprocessing for vehicle health monitoring and early warning based on a single machine learning algorithm, which is labor-intensive and prone to introduce subjective errors, and directly applying the original data when using the XGBoost model for prediction, resulting in high model complexity, large amount of calculation and unstable prediction effect, thereby achieving improved efficiency and accuracy of vehicle health monitoring and early warning.

[0078] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 3 , step S20 includes steps S21-S22:

[0079] Step S21, expressing the sample features in the system data with letters.

[0080] The sample features in the system data are represented by letters, for example, driving behavior preference (x1), driving mileage (x2), driving time (x3), acceleration and deceleration frequency (x4), braking habits (x5), vehicle age (x6), speed (x7), fuel consumption (x8), engine temperature (x9), tire pressure (x10), etc. 10 )、Fault code(x 11 )、Abnormal alarm(x 12 )、Maintenance data(x 13 )、Maintenance data(x 14 ), historical fault records (x 15 )、ABS status(x 16 ), ESP status (x 17 ), brake system status (x 18 )、Power steering system(x 19 )、Weather(x 20 )、Temperature(x 21 ), road condition (x 22 )、Traffic conditions(x 23 )、Light(x 24 )wait.

[0081] Step S22, performing factor analysis on the sample characteristics represented by letters to obtain common factor data.

[0082] By performing factor analysis on the data represented by letters, the above sample characteristics can be attributed to six major factors.

[0083] In a feasible implementation, step S22 may include steps S201 to S203:

[0084] Step S201, classifying the sample features represented by letters according to a preset standard to obtain a classification result.

[0085] According to the preset standards, the sample characteristics represented by letters are divided into six categories of factors. The first category of factors is represented by fac1, which mainly reflects the driver's driving behavior and driving style characteristics. The second category of factors is represented by fac2, which mainly reflects the basic information data of the vehicle, including basic data such as vehicle age, power consumption, and battery status. The third category of factors is represented by fac3, which mainly reflects the historical data information related to the vehicle, including historical repair and maintenance records and other information. The fourth category of factors is represented by fac4, which mainly reflects the real-time status of related control and braking sensors, including ABS status, ESP status, braking system status and other aspects. The fifth category of factors is represented by fac5, which mainly reflects the data of key environmental factors affecting vehicle health. The sixth category of factors is represented by fac6, which mainly reflects the factor information composed of other factors that cannot be covered by linear combination of the first five categories of factors.

[0086] Step S202, establishing a weight coefficient matrix and a eigenvector matrix, and calculating weight coefficient values ​​according to the weight coefficient matrix and the eigenvector matrix.

[0087] For the weight coefficient in the factor It can be expressed as a matrix A:

[0088]

[0089] Let X represent all eigenvectors (x1, x2, x3, …, x 24 ), that is, X=(x1,x2,x3,…,x 24 ),pass The values ​​of all elements of the matrix can be decomposed and calculated, where Each element in the matrix A is the weight coefficient in the factor analysis The value of , that is, the weight coefficient value of each factor is calculated.

[0090] It should be noted that the VAR (Vector AutoRegression) model is a multivariate time series model used to capture the linear relationship between multiple time series. The VAR model can predict the future value of each variable while taking into account the current and past values ​​of other variables.

[0091] Step S203: Calculate the linear combination values ​​of the sample features according to the classification results and the weight coefficient values, and use the linear combination values ​​as common factor data.

[0092] According to the classification results and weight coefficient values, the linear combination values ​​of each type of factors are calculated respectively.

[0093]

[0094] That is the 6-dimensional common factor data.

[0095] By classifying sample features and calculating linear combination values, the data dimension is reduced and the efficiency of data processing is improved.

[0096] This embodiment provides a vehicle health monitoring and early warning method, which represents the sample features in the system data with letters; factor analysis is performed on the sample features represented by letters to obtain common factor data, thereby reducing data dimension and complexity and improving the efficiency of subsequent data processing and model training.

[0097] Based on the first embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 4 , step S30 includes steps S31 to S33:

[0098] Step S31, dividing the common factor data into training set data, validation set data and test set data according to a preset ratio.

[0099] It should be noted that the data in the training set is used to train the XGBoost prediction model, the data in the validation set is used for Bayesian optimization to select the optimal hyperparameters of the model, and the data in the test set is used to evaluate the prediction effect of the model.

[0100] For the common factor data after factor analysis, the data is divided into training set data, validation set data and test set data according to a preset ratio, for example, 7:2:1.

[0101] Step S32, training the prediction model according to the training set data to obtain a training model.

[0102] The training set data is brought into the XGBoost model for training. The loss function of the model is defined as:

[0103]

[0104] in, is the loss after the t-th vehicle health monitoring warning fitting, n is the number of vehicle id samples, l represents the L2 norm, y i Indicates the health monitoring warning status of the i-th vehicle, represents the first t-1 fitting values, fac i (fac1, fac2, ..., fac6) represents the six factor indicators converted from all characteristic indicators in the vehicle health monitoring and early warning indicator system, f t(fac i ) represents the fitting value of the tth iteration, γ and λ are the hyperparameters in the model, T represents the number of leaf nodes, ω is the weight of each leaf node, and a module for iterative training model is established for the loss function of the XGBoost model.

[0105] Step S33, based on the validation set data and the test set data, the training model is subjected to Bayesian optimization and evaluation to obtain a target model for vehicle health monitoring and early warning.

[0106] It should be noted that Bayesian optimization is a strategy for global optimization, which is particularly suitable for black-box function optimization problems that have high computational costs and cannot directly calculate gradients.

[0107] The validation set data is used to select the optimal hyperparameters of the vehicle health monitoring and early warning model, and the test set data is used to evaluate the prediction effect of the model to obtain the target model for vehicle health monitoring and early warning.

[0108] In a feasible implementation, step S33 may include steps A01 to A10:

[0109] Step A01, initializing the proxy function and prior distribution based on the validation set data.

[0110] It should be noted that the surrogate function is a function that approximates the target function. Since the evaluation cost of the target function (such as the loss function of the machine learning model) is very high, we cannot directly perform an exhaustive search on it. Therefore, we use a surrogate function to approximate the target function. The most common surrogate function is the Gaussian Process (GP), which can provide the mean and variance of the prediction, thereby describing the uncertainty of the prediction. The purpose of the surrogate function is to simulate the behavior of the target function on the learned data points and predict the value of the target function on the new data points. Prior distribution is a concept in Bayesian statistics. It represents the mathematical expression of the belief or knowledge of unknown parameters or assumptions before obtaining new observation data. Prior distribution is usually based on past experience, historical data or subjective judgment. Most prior functions can use Gaussian process regression or other functions, which are mainly used to estimate the behavior of the target function in the location area.

[0111] The validation set data is used as the basic data of the proxy function, and the proxy function and prior distribution are initialized by sampling the validation set data.

[0112] Step A02: up-sample the proxy function to obtain a first sampling point.

[0113] According to the acquisition function of the XGBoost model, the next sampling point is selected for function evaluation, and the obtained sampling point is the first sampling point.

[0114] It should be noted that upsampling is the selection of new sampling points in the parameter space. These points are determined based on the prediction of the proxy function and the evaluation of the acquisition function. The purpose of this process is to find the parameter configuration that is most likely to improve the objective function value. The acquisition function defines how to select the next point for function evaluation based on the prediction of the proxy model, which is mainly used for selecting the next sampling point in the current state.

[0115] Step A03: Calculate the posterior distribution according to the first sampling points and the prior distribution.

[0116] According to the first sampling point and the prior distribution, when the evaluation value of the objective function is obtained at the first sampling point, the posterior distribution is calculated according to the Bayesian theorem.

[0117] It should be noted that Bayes' theorem is an important theory in probability theory. It describes how the probability of a hypothesis changes given relevant evidence or data. Bayes' theorem provides a method for calculating conditional probability, that is, calculating the probability of an event occurring when other conditional probabilities are known. The mathematical expression of Bayes' theorem is as follows:

[0118] P(H|E)=P(E|H)·P(H) / P(E)

[0119] Among them, P(H|E) represents the posterior probability, that is, the probability that hypothesis H is true when evidence E occurs, that is, the posterior distribution; P(E|H) represents the likelihood, that is, the probability that evidence E occurs when hypothesis H is true; P(H) represents the prior probability, that is, the probability that hypothesis H is true before considering any evidence, that is, the prior distribution; P(E) represents the marginal probability, that is, the probability that evidence E occurs regardless of whether hypothesis H is true.

[0120] Step A04: maximizing the proxy function based on the posterior distribution to obtain a second sampling point, and obtaining an objective function based on the second sampling point.

[0121] It should be noted that in Bayesian optimization, maximizing the proxy function is the process of selecting the next sampling point by optimizing the acquisition function, and the objective function is the loss function of the XGBoost model.

[0122] According to the acquisition function of the XGBoost model, the proxy function is maximized based on the posterior distribution, and the obtained sampling point is the second sampling point. The function is evaluated according to the second sampling point to calculate the value of the objective function.

[0123] Step A05, returning to the step of initializing the proxy function and the prior distribution based on the validation set data, until the change in the objective function meets the preset conditions, thereby obtaining the optimized objective function.

[0124] Repeat the steps of initializing the proxy function and the prior distribution based on the validation set data, and perform Bayesian optimization on the training model again until the change in the objective function meets the preset conditions in a preset number of iterations. The objective function at this time is the optimized objective function.

[0125] It should be noted that the preset number of rounds depends on the characteristics of the specific problem and the objective function, and there is no fixed value, for example, 10,000, 20,000, etc. The preset condition is that the change in the objective function is lower than a certain threshold, such as 0.1, 0.01, etc.

[0126] Step A06, obtaining a training model under target hyperparameters according to the optimized objective function.

[0127] It should be noted that the target hyperparameters are the optimal hyperparameters. The optimal hyperparameters are the parameter combinations that can optimize the objective function (such as the performance indicators of the model) under given problems and models.

[0128] Step A07, evaluating the training model under the target hyperparameters based on the test set data to obtain an evaluation result.

[0129] The test set data is input into the training model under the target hyperparameters for evaluation, and the evaluation results are obtained according to the evaluation indicators. Commonly used evaluation indicators include accuracy, precision, recall, F1 score and area under the ROC curve (AUC).

[0130] It should be noted that accuracy indicates the proportion of samples correctly predicted by the model to the total number of samples, precision indicates the proportion of samples predicted by the model to be positive that are actually positive, and recall is also called true positive rate (TPR), which indicates the proportion of samples correctly predicted by the model to be positive among all samples that are actually positive. The F1 score is the harmonic mean of precision and recall, which strikes a balance between the two. The ROC (Receiver Operating Characteristic) curve is a performance metric that shows the changes in the true positive rate and false positive rate of the model at different thresholds.

[0131] Step A08, when the evaluation result does not meet the preset requirements, return to the step of obtaining system data.

[0132] When the evaluation result does not meet the preset requirements, that is, the data of the evaluation index does not meet the preset requirements, for example, the accuracy is less than 80%, the F1 score is less than 0.9, etc., then return to the step of obtaining system data and recalculate the target model.

[0133] Step A09, when the evaluation result meets the preset requirement, the model evaluation is completed.

[0134] When the evaluation result meets the preset requirements, that is, the data of the evaluation indicator meets the preset requirements, for example, the accuracy is greater than or equal to 90%, the F1 score is greater than or equal to 0.9, etc., the training model evaluation under this target hyperparameter is completed.

[0135] Step A10, using the training model under the target hyperparameters as the target model to perform vehicle health monitoring and early warning.

[0136] After completing the evaluation of the trained model under the target hyperparameters, the model is used as the target model for vehicle health monitoring and early warning.

[0137] By performing Bayesian optimization and evaluation on the training model, the training model with the optimal hyperparameters is obtained to improve the accuracy of model prediction.

[0138] In a feasible implementation manner, step A10 includes steps B10 to B30:

[0139] Step B10: monitor and predict the vehicle health according to the target model to obtain monitoring results.

[0140] According to the target model, various sensor data and other relevant data of the vehicle are monitored to evaluate the current status of the vehicle and predict possible failures or performance degradation to obtain monitoring results, such as vehicle engine overheating, vehicle tire pressure being too high, etc. The prediction time range can be 3 months, 6 months, etc.

[0141] Step B20, when the monitoring result of the target model is that the vehicle health is abnormal, a vehicle health warning is issued.

[0142] When the monitoring result shows that the vehicle health is abnormal, that is, the vehicle has or may have a fault, a vehicle health warning is issued through the display screen inside the vehicle. For example, the display screen shows the following text: "The vehicle engine is overheating, please deal with it in time."

[0143] Step B30: When the monitoring result of the target model is that the vehicle is in normal health, the vehicle health warning is not issued.

[0144] When the monitoring result shows that the vehicle is healthy and normal, that is, the vehicle has no faults within the current and predicted time range, no vehicle health warning will be issued.

[0145] The target model is used to monitor the health status of the vehicle and issue early warnings in a timely manner, thus improving the safety of vehicle operation.

[0146] This embodiment provides a vehicle health monitoring and early warning method, which divides the common factor data into training set data, validation set data and test set data according to a preset ratio; trains the prediction model according to the training set data to obtain a training model; and performs Bayesian optimization and evaluation on the training model based on the validation set data and the test set data to obtain a target model for vehicle health monitoring and early warning, thereby improving the prediction efficiency and accuracy of the XGBoost model.

[0147] Based on the first embodiment of the present application, in the fourth embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 5 , step S10 includes steps S11 to S12:

[0148] Step S11, obtaining raw data, wherein the raw data includes the user's driving behavior, vehicle data, and environmental data.

[0149] It should be noted that the original data is the original data information of the user in multiple data sources of the Internet of Vehicles, including the user's driving behavior, vehicle data and environmental data, such as Figure 6 As shown, driving behavior includes driving behavior preferences, driving mileage, driving time, acceleration and deceleration frequency, braking habits, etc. Vehicle data includes vehicle age, speed, power consumption, battery temperature, battery voltage, battery internal resistance, tire pressure, fault codes, abnormal alarms, maintenance data, repair data, historical fault records, ABS status, ESP status, braking system status, power steering system, etc. Environmental data includes weather, temperature, road conditions, traffic conditions, lighting, etc.

[0150] Obtain the user's original data information from multiple data sources in the Internet of Vehicles, store the acquired data, and record missing data with null values.

[0151] Step S12, cleaning and preprocessing the original data to obtain system data.

[0152] Perform a statistical analysis on the proportion of missing values ​​in the original data obtained, and eliminate sample features with missing values ​​exceeding a certain proportion, such as 70% or 80%. Then, fill in the data according to basic data preprocessing rules to obtain cleaned data, i.e., system data.

[0153] For example, Figure 7 The figure shows the data related to a vehicle health monitoring and early warning. A total of 1,278 data items were obtained, which contained a large amount of "##" data and character data, i.e., missing data. The relevant missing data need to be removed and filled. A statistical analysis was performed on the proportion of missing values ​​in the original data obtained. Sample features with a missing proportion of more than 70% were removed. After the removal, 1,237 relevant vehicle health data items remained. The missing values ​​in the data were filled with the median, and the discrete labels in the data were encoded to obtain the cleaned data, i.e., the system data.

[0154] This embodiment provides a vehicle health monitoring and early warning method, which obtains raw data, wherein the raw data includes user driving behavior data, vehicle data, and environmental data; cleans and preprocesses the raw data to obtain system data, thereby providing a data basis for subsequent model training, optimization, and evaluation.

[0155] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the vehicle health monitoring and early warning method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0156] This application also provides a vehicle health monitoring and early warning device, please refer to Figure 8 , the vehicle health monitoring and early warning device comprises:

[0157] The data acquisition module 10 is used to acquire system data.

[0158] The data analysis module 20 is used to perform factor analysis on the system data to obtain common factor data.

[0159] The model processing and application module 30 is used to train, optimize and evaluate the prediction model according to the common factor data to obtain a target model for vehicle health monitoring and early warning.

[0160] In one embodiment, the data analysis module 20 is further used to represent the sample features in the system data with letters; and perform factor analysis on the sample features represented with letters to obtain common factor data.

[0161] In one embodiment, the data analysis module 20 is further used to classify the sample features represented by letters according to preset standards to obtain classification results; establish a weight coefficient matrix and a feature vector matrix, and calculate the weight coefficient values ​​according to the weight coefficient matrix and the feature vector matrix; calculate the linear combination values ​​of the sample features according to the classification results and the weight coefficient values, and use the linear combination values ​​as common factor data.

[0162] In one embodiment, the model processing and application module 30 is also used to divide the common factor data into training set data, validation set data and test set data according to a preset ratio; train the prediction model according to the training set data to obtain a training model; and perform Bayesian optimization and evaluation on the training model based on the validation set data and the test set data to obtain a target model for vehicle health monitoring and early warning.

[0163] In one embodiment, the model processing and application module 30 is also used to initialize the proxy function and the prior distribution based on the validation set data; upsample the proxy function to obtain a first sampling point; calculate the posterior distribution based on the first sampling point and the prior distribution; maximize the proxy function based on the posterior distribution to obtain a second sampling point, and obtain the objective function based on the second sampling point; return to the step of initializing the proxy function and the prior distribution based on the validation set data until the change in the objective function meets the preset conditions to obtain the optimized objective function; obtain the training model under the target hyperparameters based on the optimized objective function; evaluate the training model under the target hyperparameters based on the test set data to obtain an evaluation result; when the evaluation result does not meet the preset requirements, return to the step of obtaining system data; when the evaluation result meets the preset requirements, complete the model evaluation; use the training model under the target hyperparameters as the target model for vehicle health monitoring and early warning.

[0164] In one embodiment, the model processing and application module 30 is also used to monitor and predict the vehicle health according to the target model to obtain a monitoring result; when the monitoring result of the target model is that the vehicle health is abnormal, a vehicle health warning is issued; when the monitoring result of the target model is that the vehicle health is normal, the vehicle health warning is not issued.

[0165] In one embodiment, the data acquisition module 10 is further used to acquire raw data, which includes user's driving behavior data, vehicle data and environmental data; and clean and pre-process the raw data to obtain system data.

[0166] The vehicle health monitoring and early warning device provided by the present application adopts the vehicle health monitoring and early warning method in the above-mentioned embodiment, which can solve the technical problem of how to improve the efficiency and accuracy of vehicle health monitoring and early warning. Compared with the prior art, the beneficial effects of the vehicle health monitoring and early warning device provided by the present application are the same as the beneficial effects of the vehicle health monitoring and early warning method provided by the above-mentioned embodiment, and the other technical features in the vehicle health monitoring and early warning device are the same as the features disclosed in the above-mentioned embodiment method, which will not be repeated here.

[0167] The present application provides a vehicle health monitoring and early warning device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the vehicle health monitoring and early warning method in the above-mentioned embodiment one.

[0168] Reference below Fig. 9 , which shows a schematic diagram of the structure of a vehicle health monitoring and early warning device suitable for implementing the embodiment of the present application. The vehicle health monitoring and early warning device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Fig. 9 The vehicle health monitoring and early warning device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0169] like Fig. 9As shown, the vehicle health monitoring and early warning device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the vehicle health monitoring and early warning device are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the vehicle health monitoring and early warning device to communicate wirelessly or wired with other devices to exchange data. Although the figure shows a vehicle health monitoring and early warning device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or provided instead.

[0170] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0171] The vehicle health monitoring and early warning device provided by the present application adopts the vehicle health monitoring and early warning method in the above-mentioned embodiment, which can solve the technical problem of how to improve the efficiency and accuracy of vehicle health monitoring and early warning. Compared with the prior art, the beneficial effects of the vehicle health monitoring and early warning device provided by the present application are the same as the beneficial effects of the vehicle health monitoring and early warning method provided by the above-mentioned embodiment, and the other technical features in the vehicle health monitoring and early warning device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0172] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0173] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0174] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the vehicle health monitoring and early warning method in the above-mentioned embodiment.

[0175] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0176] The above-mentioned computer-readable storage medium may be included in the vehicle health monitoring and early warning device; or it may exist independently without being assembled into the vehicle health monitoring and early warning device.

[0177] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the vehicle health monitoring and early warning device, the vehicle health monitoring and early warning device: obtains system data; performs factor analysis on the system data to obtain common factor data; trains, optimizes and evaluates the prediction model based on the common factor data to obtain the target model for vehicle health monitoring and early warning.

[0178] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0179] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0180] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0181] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned vehicle health monitoring and early warning method, and can solve the technical problem of how to improve the efficiency and accuracy of vehicle health monitoring and early warning. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the vehicle health monitoring and early warning method provided in the above-mentioned embodiment, and will not be repeated here.

[0182] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned vehicle health monitoring and early warning method when executed by a processor.

[0183] The computer program product provided by this application can solve the technical problem of how to improve the efficiency and accuracy of vehicle health monitoring and early warning. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the vehicle health monitoring and early warning method provided by the above embodiment, which will not be repeated here.

[0184] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A vehicle health monitoring and early warning method, characterized in that: The method comprises: Get system data; Performing factor analysis on the system data to obtain common factor data; The prediction model is trained, optimized and evaluated according to the common factor data to obtain a target model for vehicle health monitoring and early warning.

2. The method according to claim 1, characterized in that The step of performing factor analysis on the system data to obtain common factor data comprises: Representing the sample features in the system data with letters; Factor analysis is performed on the sample characteristics represented by letters to obtain common factor data.

3. The method according to claim 2, characterized in that The step of performing factor analysis on the sample characteristics represented by letters to obtain common factor data comprises: Classify the sample features represented by letters according to preset standards to obtain classification results; Establishing a weight coefficient matrix and an eigenvector matrix, and calculating weight coefficient values ​​according to the weight coefficient matrix and the eigenvector matrix; According to the classification results and the weight coefficient values, the linear combination values ​​of the sample features are calculated respectively, and the linear combination values ​​are used as common factor data.

4. The method according to claim 1, characterized in that The step of training, optimizing and evaluating the prediction model according to the common factor data to obtain the target model for vehicle health monitoring and early warning includes: Dividing the common factor data into training set data, validation set data and test set data according to a preset ratio; Training the prediction model according to the training set data to obtain a training model; Based on the validation set data and the test set data, the training model is Bayesian optimized and evaluated to obtain a target model for vehicle health monitoring and early warning.

5. The method according to claim 4, characterized in that The step of performing Bayesian optimization and evaluation on the training model based on the validation set data and the test set data to obtain a target model for vehicle health monitoring and early warning includes: Initializing the proxy function and the prior distribution based on the validation set data; Upsampling the proxy function to obtain a first sampling point; Calculate a posterior distribution based on the first sampling point and the prior distribution; Based on the posterior distribution, the proxy function is maximized to obtain a second sampling point, and according to the second sampling point, an objective function is obtained; Returning to the step of initializing the proxy function and the prior distribution based on the validation set data until the change in the objective function satisfies the preset conditions, thereby obtaining an optimized objective function; According to the optimized objective function, a training model under target hyperparameters is obtained; Evaluate the training model under the target hyperparameters based on the test set data to obtain an evaluation result; When the evaluation result does not meet the preset requirements, return to the step of obtaining system data; When the evaluation result meets the preset requirements, completing the model evaluation; The training model under the target hyperparameters is used as the target model to perform vehicle health monitoring and early warning.

6. The method according to claim 5, characterized in that The step of using the training model under the target hyperparameter as the target model to perform vehicle health monitoring and early warning includes: According to the target model, the vehicle health is monitored and predicted to obtain monitoring results; When the monitoring result of the target model is that the vehicle health is abnormal, a vehicle health warning is issued; When the monitoring result of the target model is that the vehicle is in normal health, the vehicle health warning is not issued.

7. The method according to claim 1, characterized in that The step of obtaining system data comprises: Acquiring raw data, wherein the raw data includes a user's driving behavior, vehicle data, and environmental data; The raw data is cleaned and preprocessed to obtain system data.

8. A vehicle health monitoring and early warning device, characterized in that: The device comprises: A data acquisition module, used to acquire system data; A data analysis module, used for performing factor analysis on the system data to obtain common factor data; The model processing and application module is used to train, optimize and evaluate the prediction model according to the common factor data to obtain the target model for vehicle health monitoring and early warning.

9. A vehicle health monitoring and early warning device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the vehicle health monitoring and early warning method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the vehicle health monitoring and early warning method as described in any one of claims 1 to 7 are implemented.

Images