Wear prediction model construction method, wear prediction method, wear prediction device and in-vehicle infotainment system

Abstract

The embodiment of the invention relates to a wear prediction model construction method, a wear prediction method, a wear prediction device and a vehicle-mounted infotainment system, and the method comprises the steps: fitting a multiple linear regression formula, and carrying out the training optimization of the multiple linear regression formula through a training data set, enabling an error term and a loss function value of a wear prediction model to be within a set range, and enabling the precision of the wear prediction model to be improved. And constructing a wear prediction model. Through the wear prediction model, the real-time wear conditions of vehicle tires in different driving environments can be accurately predicted, accurate driving data can be provided during driving, and the vehicle driving safety is improved; and meanwhile, the tire replacement can be reminded, so that the use safety and the use efficiency of the tire are improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of automotive electronic technology, and specifically to a wear prediction model construction method, a wear prediction method, a device and a vehicle computer system. Background Art

[0002] Tires are circular elastic rubber products that roll on the ground and are installed on various vehicles or machines. They are usually installed on metal rims, can support the vehicle body, buffer external impacts, achieve contact with the road surface and ensure the driving performance of the vehicle. Tires are often used under complex and harsh conditions. They are subjected to various deformations, loads, forces, high and low temperature effects during driving, and have high load-bearing performance, traction performance, and buffering performance.

[0003] Tires are an important part of vehicles, and their wear status is the core indicator for judging vehicle driving data and tire service life. Tire wear monitoring mainly relies on traditional detection methods, such as physical wear detection of tires through external measuring equipment, or inferring tire wear based on tire mileage; these methods are often time-consuming and cannot monitor the degree of tire wear in real time and accurately. With the increase in vehicle mileage and the influence of different driving behaviors, such as sudden acceleration and sudden braking, the degree of tire wear is difficult to accurately predict. Summary of the invention

[0004] In view of the above problems, the present invention provides a wear prediction model construction method, wear prediction method, device and vehicle system, which are used to solve the problem that tire detection in the prior art is time-consuming and cannot monitor the degree of tire wear in real time and accurately. With the increase of vehicle mileage and the influence of different driving behaviors, the degree of tire wear is difficult to accurately predict.

[0005] According to one aspect of an example of the present invention, a method for constructing a wear prediction model is provided, the method comprising: S1, obtaining a training data set and preprocessing the training data set; S2, fitting a multiple linear regression formula, and calculating a loss function according to the multiple linear regression formula and the mean square error formula; S3, calculating the regression coefficient gradient in the multivariate linear regression formula according to the loss function, and updating the regression coefficient according to the regression coefficient gradient and a preset learning rate; S4, looping through steps S1-S3 until the multivariate linear regression formula satisfies a preset stop condition, thereby generating a wear prediction model.

[0006] In some optional embodiments, in step S1, preprocessing the training data set includes: Performing data cleaning processing on the training data set; and / or, performing feature standardization processing on the training data set.

[0007] In some optional embodiments, in step S2, the multivariate linear regression formula is specifically: , in, To predict the tire wear degree in the next stage, For mileage, , , are the accelerations of the vehicle on the x-axis, y-axis, and z-axis, respectively. is the current tire wear degree, β 0. β 1. β 2. β 3. β 4. β 5 is the regression coefficient, ε is the error term.

[0008] In some optional embodiments, in step S2, according to the multivariate linear regression formula and the mean square error formula, calculating the loss function specifically includes: By inputting the training data set into the multivariate linear regression formula, the predicted tire wear degree is calculated; The expected value of the sum of squares of the deviations between the predicted tire wear degree and the actual observed value is calculated as the loss function through the mean square error formula.

[0009] In some optional embodiments, in step S3, calculating the regression coefficient gradient in the multivariate linear regression formula according to the loss function specifically includes: Calculate using the gradient calculation formula: ,in, It is i The sample j Features.

[0010] In some optional embodiments, in step S3, the regression coefficient is updated according to the regression coefficient gradient and the preset learning rate, specifically including: The updated regression coefficient is obtained by subtracting the original regression coefficient from the product of the regression coefficient gradient and the preset learning rate.

[0011] In some optional embodiments, the preset stop condition includes any one or more of reaching a preset number of iterations, a regression coefficient gradient being less than a first threshold, and a loss function change being less than a second threshold.

[0012] According to another aspect of an embodiment of the present invention, a tire wear prediction method is provided, the method comprising: Get the current tire wear degree, mileage, and three-axis acceleration; The current tire wear degree, mileage, and three-axis acceleration are input into the wear prediction model produced by the wear prediction model construction method as described above to obtain the predicted tire wear degree.

[0013] According to another aspect of an embodiment of the present invention, a wear prediction device is provided, the device comprising: Data acquisition module, used to obtain the current tire wear degree, mileage, and three-axis acceleration; The model prediction module is used to input the current tire wear degree, mileage, and three-axis acceleration into the wear prediction model produced by the wear prediction model construction method as described above to obtain the predicted tire wear degree.

[0014] According to another aspect of an embodiment of the present invention, there is provided a vehicle system, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform the operation of the tire wear prediction method as described above.

[0015] The present invention provides a wear prediction model construction method, a wear prediction method, a device and a vehicle system, and the beneficial effects are: the present invention constructs a wear prediction model by fitting a multivariate linear regression formula and training and optimizing the multivariate linear regression formula through a training data set, so that the error term and loss function values ​​of the wear prediction model are within a set range. The wear prediction model can accurately predict the real-time wear of vehicle tires in different driving environments, which can not only provide accurate driving data during driving and improve vehicle driving safety; at the same time, it can also improve the safety and efficiency of tire use by reminding tire replacement.

[0016] The above description is only an overview of the technical solution of the example of the present invention. In order to more clearly understand the technical means of the embodiment of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the example of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings are only used to illustrate the embodiments and are not to be considered as limiting the present invention. In addition, the same reference symbols are used to represent the same components throughout the accompanying drawings. In the accompanying drawings: Figure 1The present invention provides a schematic flow chart of a wear prediction model construction method according to Embodiment 1; Figure 2 The present invention provides a schematic flow chart of step S2 of embodiment 1; Figure 3 The present invention provides a schematic flow chart of a wear prediction method according to Embodiment 2; Figure 4 The present invention provides a schematic structural diagram of a wear prediction device according to Embodiment 3; Figure 5 A schematic structural diagram of a vehicle system according to Embodiment 4 of the present invention is shown. DETAILED DESCRIPTION

[0018] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0019] Embodiment 1: Figure 1 An embodiment of a wear prediction model construction method of the present invention is shown. The method is applied to vehicle tire wear degree detection, and the wear prediction model construction method can be stored in the memory of the vehicle system through an application program. The method specifically includes: S1, obtain a training data set and preprocess the training data set; in step S1, the training data set can be obtained by testing the tire life cycle data of the test vehicle on different test sections to form a training data set; in addition, the training data set can be obtained based on the historical data collected from the vehicle. The training data set is The preprocessing of the training data set includes: performing data cleaning processing on the training data set; and / or performing feature standardization processing on the training data set.

[0020] S2, fitting the multiple linear regression formula, and calculating the loss function according to the multiple linear regression formula and the mean square error formula; in step S2, the multiple linear regression formula is specifically: ,in, To predict the tire wear degree in the next stage, For mileage, , , are the accelerations of the vehicle on the x-axis, y-axis, and z-axis, respectively. is the current tire wear degree, β 0. β 1. β 2. β 3. β 4. β5 is the regression coefficient, ε is the error term. This application comprehensively considers the mileage, three-axis acceleration and current tire wear by fitting a multiple linear regression formula; compared with simple mileage estimation, the tire wear predicted by this application is more accurate. Among them, multiple linear regression is a statistical method used to describe the linear relationship between a dependent variable predicting the tire wear and multiple independent variables mileage, three-axis acceleration, and current tire wear.

[0021] S3, according to the loss function, calculate the regression coefficient gradient in the multivariate linear regression formula, and update the regression coefficient according to the regression coefficient gradient and the preset learning rate; in step S3, according to the loss function, calculate the regression coefficient gradient in the multivariate linear regression formula, specifically including: calculating by the gradient calculation formula: ,in, It is i The sample j According to the regression coefficient gradient and the preset learning rate, the regression coefficient is updated, specifically including: obtaining the updated regression coefficient by subtracting the original regression coefficient from the product of the regression coefficient gradient and the preset learning rate.

[0022] S4, looping through steps S1-S3 until the multivariate linear regression formula satisfies the preset stop condition, and a wear prediction model is generated. In step S4, the present invention loops through steps S1-S3 according to the training data set, so that the regression coefficient is applicable to the calculation of each training data set and the loss function and the error term are minimized, and then stops executing steps S1-S3, and generates a wear prediction model.

[0023] The present invention constructs a wear prediction model by fitting a multiple linear regression formula and training and optimizing the multiple linear regression formula through a training data set, so that the error term and loss function values ​​of the wear prediction model are within a set range. The wear prediction model can accurately predict the real-time wear of vehicle tires in different driving environments, which can not only provide accurate driving data during driving and improve vehicle driving safety; at the same time, it can also improve the safety and efficiency of tire use by reminding tire replacement.

[0024] In step S1, the training data set is preprocessed, including: Data cleaning of the training data set includes: Data import: import the raw data into the data analysis environment, such as Python, R or SQL database. Table organization: modify the column name, delete blank lines and convert the data format of the data table. Evaluate data: check the structure of the training data set, check the data type, identify missing values, identify outliers and check duplicate values. Process outliers: process missing values ​​through deletion and filling. Process outliers: define thresholds, define reasonable thresholds to identify outliers according to the actual data situation and business needs; statistical methods, use Z-score, IQR interquartile range and other statistical methods to identify and process outliers; model-based methods, such as isolation forest and other machine learning algorithms can also be used for outlier detection and processing. Process duplicate values: process duplicate values ​​through unique identifier deduplication and similarity merging. Data conversion and normalization, including data type conversion, unit conversion, data format adjustment and data normalization of the training data set. Data consistency processing and verification and preservation.

[0025] The training data set is subjected to feature standardization. Specifically, the training data set may be subjected to feature standardization by using a Min-Max standardization algorithm or a Z-score standardization algorithm.

[0026] In step S2, the multiple linear regression formula is specifically: ,in, To predict the tire wear degree in the next stage, For mileage, , , are the accelerations of the vehicle on the x-axis, y-axis, and z-axis, respectively. is the current tire wear degree, β 0. β 1. β 2. β 3. β 4. β 5 is the regression coefficient, ε is the error term. In the present invention, the predicted tire wear degree of the next stage is calculated by a multiple linear regression formula. The multiple linear regression formula comprehensively considers the mileage, three-axis acceleration and current tire wear degree. At the same time, the mileage, three-axis acceleration, current tire wear degree and predicted tire wear degree are associated by the regression coefficient. The result of predicting the tire wear degree can be made more accurate by adjusting the value of the regression coefficient. The mileage is the mileage of the vehicle at the time of detection, the three-axis acceleration is the current detection moment, and the vehicle acceleration is decomposed on the x-axis, y-axis and z-axis to form , , Current tire wear level The tire wear degree value obtained in the previous detection cycle. When the mileage is 0, the current tire wear degree Assign a value of 0. Error term ε It is crucial to the accuracy and reliability of the model. Specifically, the error term can be limited by zero mean assumption, homoscedasticity, no autocorrelation, normal distribution and linear relationship to adjust the relationship between the predicted tire wear degree and mileage, three-axis acceleration, and current tire wear degree, so that the result of predicting tire wear degree is more accurate.

[0027] In step S2, see Figure 2 , according to the multivariate linear regression formula and the mean square error formula, the loss function is calculated specifically including: S21, by inputting the training data set into a multiple linear regression formula, the predicted tire wear degree is calculated; in step S21, the predicted tire wear degree in the training data set is calculated by the multiple linear regression formula .

[0028] S22, using the mean square error formula, calculate the expected value of the sum of squares of the deviations between the predicted tire wear degree and the actual observed value as a loss function. In step S22, the following mean square error formula is used as the loss function: in is the actual observed value. The mean square error (MSE) reflects the prediction error of the model by calculating the sum of squares of the difference between the predicted value and the true value and averaging them. The smaller the MSE value, the more accurate the prediction ability of the model.

[0029] In step S3, the regression coefficient gradient in the multivariate linear regression formula is calculated according to the loss function, specifically including: calculating by the gradient calculation formula: ,in, It is i The sample j In this embodiment, the mean square error MSE is considered to predict the degree of tire wear. The derivative of , use the chain rule to calculate the mean square error MSE for each regression coefficient βj The gradient of is derived as follows: .

[0030] In step S3, the regression coefficient is updated according to the regression coefficient gradient and the preset learning rate, specifically including: obtaining the updated regression coefficient by subtracting the original regression coefficient from the product of the regression coefficient gradient and the preset learning rate. The formula is: , is the original regression coefficient, is the updated regression coefficient, is the preset learning rate, is the regression coefficient gradient. The preset learning rate can be set manually by adjusting the model parameters or by mechanically optimizing the error between the predicted value and the true value.

[0031] In step S4, the preset stop condition includes any one or more of reaching the preset number of iterations, the regression coefficient gradient is less than the first threshold, and the loss function change is less than the second threshold. In this embodiment, the first threshold and the second threshold can be set when the model is established to adjust the accuracy of the results of the wear prediction model. For example, the number of iterations can be set to reach 10,000 times, the first threshold is set to between 0 and 1, and the loss function change is less than 50%. By adjusting the preset stop conditions of the wear prediction model, the model calculation can be made more accurate.

[0032] Embodiment 2: Figure 3 An embodiment of a tire wear prediction method is shown. Based on the wear prediction model constructed in Embodiment 1, the method includes: S4, obtain the current tire wear degree, mileage, and three-axis acceleration; in step S4, the current vehicle mileage, three-axis acceleration, and current tire wear degree are obtained in real time by connecting to the vehicle's CAN bus, wherein the mileage is the mileage of the vehicle at the time of detection, and the three-axis acceleration is the decomposition of the vehicle acceleration on the x-axis, y-axis, and z-axis at the current detection moment, forming , , Current tire wear level The tire wear degree value obtained in the previous detection cycle. When the mileage is 0, the current tire wear degree Assign a value of 0.

[0033] S5, input the current tire wear degree, mileage, and three-axis acceleration into the wear prediction model produced by the wear prediction model construction method described in Example 1 to obtain the predicted tire wear degree. In step S5, by inputting the current tire wear degree, mileage, and three-axis acceleration into the wear prediction model, the multivariate linear regression formula of the wear prediction model is: , get the predicted tire wear level for the next stage .

[0034] The present invention constructs a wear prediction model by fitting a multiple linear regression formula and training and optimizing the multiple linear regression formula through a training data set, so that the error term and loss function values ​​of the wear prediction model are within a set range. The wear prediction model can accurately predict the real-time wear of vehicle tires in different driving environments, which can not only provide accurate driving data during driving and improve vehicle driving safety; at the same time, it can also improve the safety and efficiency of tire use by reminding tire replacement.

[0035] Embodiment 3: Figure 4 An embodiment of a wear prediction device 400 is shown, and the device comprises: The data acquisition module 410 is used to execute step S4 in embodiment 2, specifically including: obtaining the current vehicle mileage, three-axis acceleration, and current tire wear degree in real time by connecting to the vehicle's CAN bus, wherein the mileage is the mileage of the vehicle during the detection, the three-axis acceleration is the current detection moment, and the vehicle acceleration is decomposed in the x-axis, y-axis, and z-axis to form , , Current tire wear level The tire wear degree value obtained in the previous detection cycle. When the mileage is 0, the current tire wear degree Assign a value of 0.

[0036] The model prediction module 420 is used to execute step S5 in embodiment 2, specifically including: inputting the current tire wear degree, mileage, and three-axis acceleration into the wear prediction model, and using the multivariate linear regression formula of the wear prediction model: , get the predicted tire wear level for the next stage .

[0037] The present invention constructs a wear prediction model by fitting a multiple linear regression formula and training and optimizing the multiple linear regression formula through a training data set, so that the error term and loss function values ​​of the wear prediction model are within a set range. The wear prediction model can accurately predict the real-time wear of vehicle tires in different driving environments, which can not only provide accurate driving data during driving and improve vehicle driving safety; at the same time, it can also improve the safety and efficiency of tire use by reminding tire replacement.

[0038] Embodiment 4: Figure 5 The schematic diagram of the structure of the embodiment of the vehicle system of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the vehicle system.

[0039] like Figure 5As shown, the vehicle system may include: a processor, a communication interface, a memory, and a communication bus.

[0040] The processor 510, the communication interface 540, and the memory 520 communicate with each other via the communication bus 530. The communication interface is used to communicate with other devices such as a client or other server network elements. The processor is used to execute the program 550, which can specifically execute the relevant steps in the above-mentioned embodiment of the vehicle-mounted key storage method.

[0041] Specifically, the program may include program code including computer executable instructions.

[0042] The processor may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. The one or more processors included in the vehicle system may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0043] Memory is used to store programs. The memory may include high-speed RAM memory and may also include non-volatile memory, such as at least one disk storage.

[0044] The program can be called by the processor to enable the vehicle system to execute Figure 1 The following steps of embodiment 2 are as follows: S4, obtaining the current tire wear degree, mileage, and three-axis acceleration; S5, inputting the current tire wear degree, mileage, and three-axis acceleration into the wear prediction model produced by the wear prediction model construction method described in Example 1 to obtain a predicted tire wear degree.

[0045] The present invention constructs a wear prediction model by fitting a multiple linear regression formula and training and optimizing the multiple linear regression formula through a training data set, so that the error term and loss function values ​​of the wear prediction model are within a set range. The wear prediction model can accurately predict the real-time wear of vehicle tires in different driving environments, which can not only provide accurate driving data during driving and improve vehicle driving safety; at the same time, it can also improve the safety and efficiency of tire use by reminding tire replacement.

[0046] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system or other device. In addition, the embodiments of the present invention are not directed to any particular programming language.

[0047] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. Similarly, in order to simplify the present invention and help understand one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, the various features of the embodiments of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. Wherein, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, wherein each claim itself is a separate embodiment of the present invention.

[0048] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and further may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive.

[0049] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc. does not indicate any order. These words may be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be understood as limitations on the order of execution.

Claims

1. A wear prediction model construction method, characterized in that: The method comprises: S1, obtaining a training data set and preprocessing the training data set; S2, fitting a multiple linear regression formula, and calculating a loss function according to the multiple linear regression formula and the mean square error formula; S3, calculating the regression coefficient gradient in the multivariate linear regression formula according to the loss function, and updating the regression coefficient according to the regression coefficient gradient and a preset learning rate; S4, looping through steps S1-S3 until the multivariate linear regression formula satisfies a preset stop condition, thereby generating a wear prediction model.

2. The wear prediction model construction method according to claim 1, characterized in that: In step S1, the training data set is preprocessed, including: Performing data cleaning processing on the training data set; and / or, performing feature standardization processing on the training data set.

3. The wear prediction model construction method according to claim 1, characterized in that: In step S2, the multivariate linear regression formula is specifically: ， in, To predict the tire wear degree in the next stage, For mileage, , , are the accelerations of the vehicle on the x-axis, y-axis, and z-axis, respectively. is the current tire wear degree, β 0. β 1. β 2. β 3. β 4. β 5 is the regression coefficient, ε is the error term.

4. The wear prediction model construction method according to claim 3, characterized in that: In step S2, according to the multivariate linear regression formula and the mean square error formula, calculating the loss function specifically includes: Calculate and predict the degree of tire wear by inputting the training data set into the multivariate linear regression formula; The expected value of the sum of squares of the deviations between the predicted tire wear degree and the actual observed value is calculated as the loss function through the mean square error formula.

5. The wear prediction model construction method according to claim 4, characterized in that: In step S3, the regression coefficient gradient in the multivariate linear regression formula is calculated according to the loss function, specifically including: Calculate using the gradient calculation formula: ,in, It is i The sample j Features.

6. The wear prediction model construction method according to claim 5, characterized in that: In step S3, the regression coefficient is updated according to the regression coefficient gradient and the preset learning rate, specifically including: The updated regression coefficient is obtained by subtracting the original regression coefficient from the product of the regression coefficient gradient and the preset learning rate.

7. The wear prediction model construction method according to claim 5, characterized in that: The preset stop condition includes any one or more of reaching a preset number of iterations, a regression coefficient gradient being less than a first threshold, and a loss function change being less than a second threshold.

8. A tire wear prediction method, characterized in that: The method comprises: Get the current tire wear degree, mileage, and three-axis acceleration; The current tire wear degree, mileage, and three-axis acceleration are input into the wear prediction model produced by the wear prediction model construction method according to any one of claims 1 to 7 to obtain the predicted tire wear degree.

9. A wear prediction device, characterized in that: The device comprises: Data acquisition module, used to obtain the current tire wear degree, mileage, and three-axis acceleration; The model prediction module is used to input the current tire wear degree, mileage, and three-axis acceleration into the wear prediction model produced by the wear prediction model construction method according to any one of claims 1 to 7 to obtain the predicted tire wear degree.

10. A vehicle computer system, characterized in that: include: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform the operation of the tire wear prediction method as described in any one of claims 1-7.

Images