Wheel steering angle calibration model training and prediction method, system, equipment and medium

Abstract

The invention relates to the technical field of automobile control systems, and discloses a wheel steering angle calibration model training and prediction method, system, device and medium, and the method comprises the steps: obtaining the sample data of vehicle turning, the sample data comprising a steering wheel angle, a vehicle speed interval and a wheel steering angle; in response to a to-be-trained model, obtaining prediction data of the sample data; model parameters of the to-be-trained model are updated to reduce loss of the prediction data and the sample data, a calibration model is obtained, and the calibration model is used for predicting the wheel steering angle in response to the steering wheel angle and the vehicle speed interval. The method has the advantages that the problems of tedious operation and measurement errors in a traditional method are avoided, the internal parameters of the model are automatically adjusted according to different application scenes, and the output precision is continuously improved along with data accumulation.

Description

Technical Field

[0001] The invention relates to the technical field of automobile control systems, and in particular to a wheel steering angle calibration model training and prediction method, system, equipment and medium. Background Art

[0002] With the continuous advancement and widespread application of autonomous driving technology, the accuracy of the planning and control system is not only related to the comfort experience of passengers, but also a core element to ensure road traffic safety. As one of the core parameters in this system, the precise control of the turning radius is directly related to the vehicle's path planning, obstacle avoidance ability, and response speed in emergency situations. Given the technical challenges of directly measuring the steering angle of the vehicle's wheels, indirect calculation using the transmission ratio parameter between the vehicle's steering wheel angle and the wheel steering angle has become a practical and efficient method.

[0003] The transmission ratio parameter not only reflects the physical characteristics of the vehicle power system, but is also the cornerstone of the automatic driving algorithm to achieve precise control. By accurately calibrating the transmission ratio, the system can accurately convert the steering wheel instructions issued by the driver or the automatic driving system into the actual turning angle of the wheel, ensuring that the vehicle follows the expected trajectory and reducing the deviation or instability caused by errors.

[0004] Traditional steering wheel angle transmission ratio calibration methods mainly include manual measurement and automatic measurement.

[0005] (1) Manual measurement method

[0006] Although manual measurement can directly reflect the driving characteristics of the vehicle under different conditions, its inherent limitations, such as large measurement errors, long time consumption, and high operator skill requirements, limit its application in the high-precision requirements of autonomous driving. In addition, manual measurement is difficult to cope with complex and changing road environments and real-time requirements.

[0007] (2) Automatic measurement method

[0008] In order to overcome the shortcomings of manual measurement, automatic measurement methods have emerged. By integrating advanced technologies such as high-precision GPS, inertial navigation system (INS), laser radar (LiDAR) or visual odometer, the automatic measurement system can record and analyze the vehicle's driving trajectory and posture in real time, thereby realizing automatic calibration of the transmission ratio parameters. This method significantly improves the efficiency and accuracy of calibration, but still faces the following challenges:

[0009] Batch measurement and angle fixation have high requirements: Existing automatic measurement solutions often need to be performed in batches according to different steering wheel angles, and the steering wheel angle needs to be kept stable, which increases the complexity and time cost of the operation.

[0010] Environmental adaptability: Different road conditions, tire wear, load changes and other factors may affect the actual value of the transmission ratio, requiring the calibration system to have a high degree of environmental adaptability and real-time adjustment capabilities.

[0011] Data processing and algorithm optimization: High-precision positioning data needs to be processed by complex algorithms to eliminate noise and calculate the transmission ratio. The accuracy and efficiency of the algorithm directly affect the calibration results.

[0012] Therefore, it is urgent to develop a method for wheel steering angle calibration training and prediction to solve the above problems. Summary of the invention

[0013] In order to solve the above technical problems, the present invention provides a wheel steering angle calibration model training and prediction method, system, equipment and medium. By acquiring a large amount of basic data of vehicle road tests as sample data, a model is constructed, and the model is trained in different speed ranges to obtain the corresponding relationship between the steering wheel angle and the wheel steering angle, thereby constructing an accurate transmission ratio model.

[0014] In a first aspect, a first object of the present invention is to provide a wheel steering angle calibration model training method, comprising:

[0015] Acquire sample data of vehicle turning, wherein the sample data includes a steering wheel angle, a vehicle speed range, and a wheel steering angle;

[0016] In response to the model to be trained, obtaining prediction data of the sample data;

[0017] The model parameters of the model to be trained are updated to reduce the loss of the predicted data and the sample data, and a calibration model is obtained, wherein the calibration model is used to predict the wheel steering angle in response to the steering wheel angle and the vehicle speed range.

[0018] In a second aspect, a second object of the present invention is to provide a method for predicting a wheel steering angle, comprising:

[0019] Acquiring operating parameters of the vehicle turning; wherein the operating parameters include a steering wheel angle and a vehicle speed;

[0020] The wheel steering angle of the vehicle is predicted according to the operating parameters and the calibration model; wherein the calibration model is trained using the wheel steering angle calibration model training method described in the first aspect.

[0021] In a third aspect, a third object of the present invention is to provide a wheel steering angle calibration model training system, comprising:

[0022] A sample data acquisition module, used to acquire sample data of vehicle turning, wherein the sample data includes a steering wheel angle, a vehicle speed range, and a wheel steering angle;

[0023] A sample data prediction module, used for obtaining predicted data of the sample data in response to the model to be trained;

[0024] A calibration model determination module is used to update the model parameters of the model to be trained to reduce the loss of the predicted data and the sample data and obtain a calibration model; wherein the calibration model is used to predict the wheel steering angle in response to the steering wheel angle and the vehicle speed range.

[0025] In a fourth aspect, a fourth object of the present invention is to provide an electronic device, comprising a processor and a memory for storing instructions executable by the processor; when the processor is configured to execute the instructions, the electronic device implements the method described in the first aspect.

[0026] In a fifth aspect, a fifth object of the present invention is to provide a computer-readable storage medium, wherein the storage medium stores a computer program; the computer program is executed by a processor to implement the method as described in the first aspect.

[0027] The embodiments of the present invention have the following technical effects:

[0028] 1. This application proposes a wheel steering angle model training method, which uses the steering wheel angle vehicle speed range and wheel steering angle as sample data for training, and obtains a model with input as vehicle speed range and steering wheel angle, and output as wheel steering angle, which is convenient for calibrating the relationship between the two in practice.

[0029] 2. As the speed increases, the side slip angle will be generated, thus affecting the change of the wheel steering angle. The dynamic characteristics are significantly different. Therefore, different calculation methods are used for low-speed and high-speed steering angles to obtain the wheel steering angle. The calculation results are more accurate, which improves the accuracy and robustness of the wheel steering angle calculation results. The trained model is also more accurate and can accurately capture the dynamic change relationship between the steering wheel and the wheel steering angle at different speeds. It ensures that the training model can provide high-precision prediction and control capabilities at different driving speeds.

[0030] 3. This application directly uses the real data obtained from actual road tests as input to model the relationship between the steering wheel and the wheel steering angle; it avoids the specific scene selection and repeated manual operations required in the traditional calibration process, improves the calibration efficiency, and enhances the practicality and reliability of the calibration results;

[0031] 4. This application uses a polynomial formula as the model to be trained, which not only ensures that the model can better approximate complex nonlinear relationships, but also controls the computational complexity through reasonable mathematical forms; mean square error and gradient descent algorithms are introduced to adjust the fitting parameters; while maintaining the calculation accuracy, the amount of calculation is reduced and the execution efficiency of the algorithm is improved;

[0032] 5. This application continuously learns different sample data, and the model can automatically adjust its internal parameters to adapt to a variety of application scenarios; it simplifies the calibration process and can ensure the stability and reliability of the autonomous driving system in different environments;

[0033] In summary, this application has broad application prospects and important practical value in the research and development of autonomous driving technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1 is a flow chart of a wheel steering angle calibration model training method provided by an embodiment of the present invention;

[0036] Figure 2 is a kinematic model for low-speed steering provided by an embodiment of the present invention;

[0037] Figure 3 is a kinematic model for high-speed steering provided by an embodiment of the present invention;

[0038] Figure 4 It is a steering wheel transmission ratio model provided by the embodiment of the present invention when the vehicle speed range is 0 to 40 km / h;

[0039] Figure 5 is a steering wheel transmission ratio model provided by an embodiment of the present invention when the vehicle speed range is greater than 40 km / h;

[0040] Figure 6 It is a schematic diagram of the result of comparing the curve obtained by the steering wheel calibration model at 0-40 km / h provided by the embodiment of the present invention with the value provided by the supplier. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0042] First, let’s explain the following terms:

[0043] The steering ratio, also known as the steering ratio or steering ratio, is the ratio that describes the relationship between the steering wheel steering angle and the wheel steering angle. The setting of the steering ratio needs to take into account a variety of factors, including vehicle type, purpose, suspension system, and driver's control habits. The steering ratio can be expressed by a formula.

[0044] Vehicle road testing is an important test for motor vehicle manufacturers to ensure vehicle performance and quality. The vehicle's handling performance, braking performance, sliding performance and emissions are fully tested to ensure the vehicle's ability to drive in different environments and on different roads. Various test data and phenomena will be recorded in detail during the road test.

[0045] The present application directly uses vehicle road test data to model the relationship between the steering wheel angle and the wheel steering angle; avoids the specific scene selection and repeated manual operations required in the traditional calibration process, improves the calibration efficiency, and enhances the practicality and reliability of the calibration results.

[0046] Example 1

[0047] Figure 1 is a flow chart of a wheel steering angle calibration model training method provided by an embodiment of the present invention. Figure 1 , specifically including:

[0048] S1. Obtain sample data of vehicle turning, where the sample data includes a steering wheel angle, a vehicle speed range, and a wheel steering angle.

[0049] The steering wheel angle is obtained through the angle sensor of the electric power steering system of the car. When the steering wheel is turned from the middle position to one side, it can turn about one to one and a half circles. That is, the steering wheel angle can directly obtain the value through the angle sensor. The speed range can be set according to the actual situation. This application does not limit this. In principle, the more the number of intervals and the smaller the intervals, the more accurate the trained model will be. The wheel steering angle needs to be calculated using the dynamic formula.

[0050] The calculation of wheel steering angle is analyzed as follows:

[0051] First, when a vehicle is traveling on a curved road, the wheels will slip, that is, the centerline of the wheel deviates from its original straight driving direction. At this time, the angle between the wheel plane and the direction of movement of the wheel center is the slip angle. However, according to the running speed of the vehicle, the calculation of the wheel steering angle needs to be divided into two cases: low speed and high speed:

[0052] When the car is turning at low speed: the tire does not need to generate lateral force. In this case, there is no slip angle when the tire rolls.

[0053] When the car is turning at high speed: the dynamic model needs to consider the side slip angle of the vehicle tires, and the vehicle may understeer or oversteer. It is also because of the appearance of lateral acceleration that the steering equation is different from that when turning at low speed. Even if the steering wheel angle is kept fixed, the turning radius of the vehicle is not consistent at different speeds.

[0054] Therefore, it is necessary to perform different calculations on the wheel steering angles under different states to reduce the calculation error of the model and improve the accuracy of the transmission ratio calibration, so as to more accurately describe and predict the steering behavior of the vehicle.

[0055] From the above, we can see that the wheel steering angle can be divided into two different situations, namely high speed and low speed.

[0056] In summary, before obtaining sample data of vehicle turning, the wheel steering angle is determined according to the vehicle speed and a preset speed threshold.

[0057] The specific calculation method is as follows:

[0058] When the vehicle speed is less than or equal to a preset speed threshold, determining the wheel steering angle according to the vehicle wheelbase, the vehicle linear speed, and the vehicle angular speed;

[0059] When the vehicle speed is greater than a preset speed threshold, the wheel steering angle is determined based on the vehicle wheelbase, front axle mass, rear axle mass, front wheel cornering stiffness, rear wheel cornering stiffness, and the turning radius from the steering center to the center of mass after the front and rear wheels of the vehicle corner.

[0060] Among them, the vehicle angular velocity is the rate of change of the angle of the wheel around the axis per unit time, and the vehicle linear velocity is the instantaneous velocity at a certain moment in the vehicle's driving process. Both the vehicle angular velocity and the vehicle linear velocity are obtained through the chassis inertial measurement unit. The vehicle wheelbase is the distance from the center of the front axle to the center of the rear axle, which is obtained through the design parameters of the vehicle or manual measurement. The front axle mass is the weight borne by the front axle of the vehicle, that is, the load on the front axle, and the rear axle mass is the weight borne by the rear axle of the vehicle, that is, the load on the rear axle. The front axle mass and the rear axle mass can be obtained by installing sensors on the axle to monitor the axle load, or by calculating them in other ways, which will not be elaborated here. The lateral stiffness of the front and rear wheels is a performance parameter of the tire, which can be directly obtained according to the tire manufacturer and model. The above parameters are all included in the vehicle road test data.

[0061] The preset speed threshold here can divide the vehicle speed into a low-speed interval and a high-speed interval. For example, if the preset speed threshold is 40km / h, the range of 0 to 40km / h is the low-speed turning interval, and the range greater than 40km / h is the high-speed turning interval. Here is an example of the division range, which is determined based on the actual operation of the vehicle and the needs of data analysis. The operator can also make other divisions based on experience and actual conditions, which are not specifically limited or explained here.

[0062] (1) When the vehicle is turning at low speed, if the rear wheels are not turning, the steering center must be on the extension line of the rear axle.

[0063] First of all, it should be noted that when a car turns, the body of the car moves along an arc line, and the center of this arc line is the turning center of the car; the center of mass is the point of action of the resultant force of gravity on various parts of the vehicle, and the letter O is used in this application to represent the position of the center of mass.

[0064] See also Figure 2 , the vehicle kinematic model is simplified to a wheel steering model, the upper black frame is the front wheel, and the lower black frame is the rear wheel. r R0 represents the turning radius from the steering center to the rear axle of the car, and R0 represents the turning radius from the steering center to the center of mass. According to the geometric relationship during steering, the turning radius R0 is much larger than L r , L r Refers to the distance from the rear wheel axle to the center of mass of the vehicle, ignoring R r The error from R0, the distance from the steering center to the center of mass is approximately equal to the distance from the steering center to the rear axle, that is: R r ≈R0.

[0065] And the turning radius:

[0066] R r =v / w,

[0067] Where, v is the vehicle linear velocity, w is the vehicle angular velocity;

[0068] The wheel steering angle in the low speed range is:

[0069] x1=arctan(L / R r )≈arctan(L / R0),

[0070] Where L is the wheelbase of the vehicle. According to the geometric relationship, x1 = θ, where θ is the angle between the line segment from the steering center to the rear axle and the line segment from the steering center to the center of the front axle.

[0071] (2) When the vehicle is turning at high speed, consider the side slip angle of the vehicle tires. Figure 3 According to the vehicle dynamics model, we can get:

[0072]

[0073] Among them, a f is the front wheel slip angle, a r is the rear wheel slip angle, M f is the front axle mass, M r is the rear axle mass, C f is the front wheel cornering stiffness, C r is the rear wheel cornering stiffness.

[0074] According to the geometric relationship x2=δ+a f -a r , where δ is the angle between the line segment from the steering center to the rear axle and the line segment from the steering center to the front axle when the vehicle swerves, that is, the actual steering angle.

[0075] The wheel steering angle in the high-speed section is:

[0076]

[0077] Among them, R is the turning radius from the steering center to the center of mass after the front and rear wheels of the car deviate sideways.

[0078] exist Figure 3 In the figure, R1 is the turning radius from the steering center to the center of mass when the front and rear wheels of the vehicle do not deviate sideways, and δ1 is the turning radius from the steering center to the center of mass when the vehicle does not deviate sideways. Obviously, δ and δ1 are not the same.

[0079] It should be noted that, in principle, the sample size of sample data should be large enough to ensure the stability and reliability of the results.

[0080] S2. In response to the model to be trained, obtain prediction data of the sample data.

[0081] Considering the impact of the vehicle on the turning radius in different speed ranges, the vehicle's steering equation is a nonlinear relationship. Different speed ranges can be trained using different formulas. For example, a linear formula can be used for low-speed ranges. However, in this application, the same polynomial linear model is used for different speed ranges to reduce the complexity of interval allocation while ensuring accuracy.

[0082] The model to be trained includes a polynomial model. The relationship between the steering wheel angle and the wheel steering angle is as follows:

[0083] y=w0+w1x+w2x 2 +w3x 3 +…+w n x n ,

[0084] Where y is the steering wheel angle, x is the wheel steering angle, and w i (i=0,1,2,...,n) is a model parameter, and n is an integer greater than 1.

[0085] For example, in the present application, a cubic polynomial equation may be used to represent the relationship between the vehicle steering wheel angle and the wheel steering angle in each vehicle speed range. Compared with a low-order polynomial, a cubic polynomial can fit nonlinear trends more accurately and capture complex data patterns, including bends and inflection points. Compared with a high-order polynomial, it can reduce the required computing and storage resources while maintaining a certain fitting accuracy, ensuring computing efficiency and feasibility, and avoiding the problem of overfitting.

[0086] That is:

[0087] y=w0+w1x+w2x 2 +w3x 3 .

[0089] Substituting the wheel steering angle in the sample data into the above formula and assigning arbitrary values ​​to the model parameters first, the predicted data of the steering wheel angle can be obtained.

[0090] S3. Update the model parameters of the model to be trained to reduce the loss of prediction data and sample data, and obtain a calibration model; wherein the calibration model is used to respond to the steering wheel angle and the vehicle speed range to predict the wheel steering angle.

[0091] Specifically, for the multiple divided vehicle speed intervals, the model parameters of the model to be trained are updated according to the predicted data and sample data in each vehicle speed interval to obtain the calibration model of each vehicle speed interval.

[0092] Specifically in actual calculations, the number of vehicle speed intervals can be appropriately increased to further improve the accuracy of model training. In theory, the more intervals there are and the smaller the range of each interval, the more accurate the parameter results from later training will be.

[0093] In a preferred embodiment, the following divisions are made: 0 to 40 km / h is a vehicle speed range, 40 km / h to 60 km / h and greater than 60 km / h, three models are set respectively, and corresponding data are substituted for calculation.

[0094] It should be noted that the vehicle speed interval here is different from the two intervals divided by the preset speed threshold. The former is to train the model in different vehicle speed intervals during model training, while the latter is to divide the model by considering the sideslip angle when calculating the front wheel steering angle.

[0095] A loss function is introduced in the process of updating the model parameters of the model to be trained to evaluate the matching degree between the mathematical model prediction data and the sample data. In a specific embodiment, the loss function is the mean square error, which is expressed by the following formula:

[0096]

[0097] Among them, y i (i=1,2,...,n) is the sample data of the i-th sample, is the predicted data of the ith sample, and n is the number of samples. The mean square error can directly calculate the average of the square of the difference between the predicted data and the sample data, reflecting the degree of deviation between the predicted value and the true value. The derivative of the mean square error is a linear function of the difference between the predicted value and the true value, which simplifies the optimization process and makes the calculation more efficient.

[0098] In linear regression, the gradient descent algorithm is often used to minimize the error function to find the optimal model parameters.

[0099] The gradient descent algorithm specifically includes the following steps:

[0100] S31. Initialize model parameters. Before the algorithm starts, the model parameters need to be randomly initialized. These initial values ​​will serve as the starting point of the iterative process. The learning rate, i.e., the step size, also needs to be set to control the step size of each iteration.

[0101] S32. Calculate the gradient of the loss function with respect to each model parameter. Under the current parameter value, calculate the gradient, and the direction of the gradient represents the direction in which the function value increases fastest.

[0102] S33, updating the model parameter values ​​according to the reverse direction of the gradient; in the gradient descent algorithm, updating the parameters along the reverse direction of the gradient (i.e., the steepest descent direction). Specifically, updating the model parameters according to a preset learning rate so that the value of the loss function decreases.

[0103] S34, repeat the above steps until the preset maximum number of iterations is reached or the loss function converges stably. The parameters will be continuously optimized during the training process to make the prediction results of the model more accurate.

[0104] It should be noted that the present application does not specifically limit the type of gradient descent algorithm. Both the loss function and the gradient descent algorithm are prior arts and will not be described in detail here.

[0105] After the above training and the model parameter values ​​obtained by the gradient descent algorithm, the expressions of the models for each speed range are obtained, that is, the calibration formula. The steering wheel rotation range is fixed. Substituting the steering wheel angle or wheel steering angle into the calibration model, the accurate wheel steering angle or steering wheel angle can be calculated. According to the formulas for different speed ranges and combined with the steering wheel angle range, the relationship diagram of the steering wheel angle and wheel steering angle can be obtained.

[0106] In a specific embodiment, a certain vehicle model is calibrated. It is divided into intervals of 0-40km / h and greater than 40km / h; the collected basic data is substituted into the model, and the wheel steering angle is first calculated according to the preset speed threshold and vehicle speed, and a polynomial model is constructed. Different model trainings are performed according to different vehicle speed intervals. During the training process, the parameters in the model are continuously updated using the loss function and the gradient descent algorithm until the preset maximum number of iterations is reached or the loss function converges stably. At this time, the corresponding model parameter values ​​are what are required, and they are substituted into the model formula to obtain the steering wheel transmission ratio formulas for each speed interval.

[0107] According to the above process, we can get Figure 4 and Figure 5 .in, Figure 4 It is a polynomial curve with a vehicle speed range of 0 to 40 km / h. The horizontal axis in the figure is the wheel steering angle (taking the front wheel as an example), and the vertical axis is the steering wheel angle. The range of the steering wheel angle is between one circle and one and a half circles. The specific formula for the relationship between the two is:

[0108] y=3.54+15.65x-0.005x 2 +0.001x 3 .

[0109] Figure 5 It is a polynomial curve when the vehicle speed range is greater than 40km / h. The horizontal axis in the figure is the wheel steering angle, and the vertical axis is the steering wheel angle. The range of the steering wheel angle is between one circle and one and a half circles. The specific formula for the relationship between the two is:

[0110] y=3.54+17.98x-0.045x 2 +0.007x 3 .

[0111] Comparing the above formula with the values ​​provided by the supplier, they are basically consistent. For example, Figure 6 It is a schematic diagram of the comparison results at 0-40km / h. The solid line represents the model of this application, and the dotted line is the formula obtained based on the numerical values ​​provided by the supplier. Obviously, the difference between the two is very small, which fully proves the accuracy of the model of this application.

[0112] This application can also be verified by:

[0113] A portion of the basic data is used as a validation set, and the wheel steering angle is calculated through the data in the validation set. The calculated wheel steering angle is substituted into the steering wheel transmission ratio formula to calculate the steering wheel angle, and the calculated steering wheel angle is compared with the steering wheel angle in the validation set. It has been verified that the steering wheel angle calculated according to the transmission ratio formula of this application and the corresponding angle value in the validation set are highly consistent, so this application has high-precision prediction capabilities.

[0114] In summary, this application uses the steering wheel angle, vehicle speed range and wheel steering angle as sample data for training, and obtains a model with input as vehicle speed range and steering wheel angle or wheel steering angle, and output as wheel steering angle or steering wheel angle, which is convenient for calibrating the relationship between the two in practice.

[0115] For the wheel steering angle, as the speed increases, the sideslip angle will be generated, thus affecting the change of the wheel steering angle. There are significant differences in the dynamic characteristics. Therefore, different calculation methods are used for low-speed turning angles and high-speed turning angles to obtain the wheel steering angle; the calculation results are more accurate, which improves the accuracy and robustness of the wheel steering angle calculation results; the trained model is also more accurate, and can accurately capture the dynamic change relationship between the steering wheel and the wheel steering angle at different speeds; ensuring that the trained model can provide high-precision prediction and control capabilities at different driving speeds.

[0116] This application directly uses real data obtained from actual road tests as input to model the relationship between the steering wheel and the wheel steering angle; avoids the specific scene selection and repetitive manual operations required in the traditional calibration process; uses a polynomial formula as the model to be trained, which not only ensures that the model can better approximate complex nonlinear relationships, but also controls the calculation complexity through a reasonable mathematical form; at the same time, the mean square error and gradient descent algorithm are introduced to adjust the fitting parameters; while maintaining the calculation accuracy, the amount of calculation is reduced and the execution efficiency of the algorithm is improved.

[0117] Example 2

[0118] The present application provides a method for predicting a wheel steering angle, comprising:

[0119] Obtaining operating parameters of the vehicle turning; wherein the operating parameters include a steering wheel angle and a vehicle speed;

[0120] The wheel steering angle of the vehicle is predicted according to the operating parameters and the calibration model; wherein the calibration model is trained using the wheel steering angle calibration model training method in Example 1.

[0121] The training method of the wheel steering angle calibration model can be referred to Example 1, which will not be described in detail here.

[0122] Example 3

[0123] Based on the content of Example 1, the present application provides a vehicle steering wheel transmission ratio calibration system, including:

[0124] A sample data acquisition module is used to acquire sample data of vehicle turning, the sample data including steering wheel angle, vehicle speed range and wheel steering angle;

[0125] A sample data prediction module, connected to the sample data acquisition module, for obtaining prediction data of the sample data in response to the model to be trained;

[0126] The calibration model determination module is connected to the sample data prediction module and is used to update the model parameters of the model to be trained to reduce the loss of prediction data and sample data and obtain a calibration model; wherein the calibration model is used to respond to the steering wheel angle and vehicle speed range and predict the wheel steering angle.

[0127] For other contents, please refer to Example 1 and will not be described in detail here.

[0128] It should be noted that the division of the above-mentioned functional modules is only used as an example for illustration. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0129] Example 4

[0130] The present application provides an electronic device, which includes a processor and a memory for storing instructions executable by the processor; when the processor is configured to execute the instructions, the electronic device implements a wheel steering angle calibration model training method in Example 1 and a wheel steering angle prediction method in Example 2. The electronic device can be a tablet computer, a handheld computer, a personal computer PC and other electronic devices. The electronic device can directly display the calibration results, which is convenient for operators to directly view.

[0131] Example 5

[0132] The present application provides a computer-readable storage medium, which stores a computer program; the computer program is executed by a processor to implement a wheel steering angle calibration model training method as in Example 1 and a wheel steering angle prediction method as in Example 2.

[0133] In some embodiments, the computer readable memory is a high-speed cache memory, which can store instructions or data used or recycled by the electronic device. If the electronic device needs to use the instruction or data again, it can be directly called from the memory. Repeated access is avoided, the waiting time of the electronic device is reduced, and the efficiency of the system is improved.

[0134] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0135] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When the use is implemented in whole or in part in the form of a computer program product, the computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk SolidStateDisk (SSD)), etc.

[0136] It should be noted that the terms used in the present invention are only for describing specific embodiments, rather than limiting the scope of the present application. As shown in the present specification, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular, but may also include the plural. The terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of more restrictions, the elements defined by the sentence "include one..." do not exclude the presence of other identical elements in the process, method or device including the elements.

[0137] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A wheel steering angle calibration model training method, characterized in that: include: Acquire sample data of vehicle turning, wherein the sample data includes a steering wheel angle, a vehicle speed range, and a wheel steering angle; In response to the model to be trained, obtaining prediction data of the sample data; The model parameters of the model to be trained are updated to reduce the loss of the predicted data and the sample data, and a calibration model is obtained, wherein the calibration model is used to predict the wheel steering angle in response to the steering wheel angle and the vehicle speed range.

2. The wheel steering angle calibration model training method according to claim 1, characterized in that: Before obtaining the sample data of the vehicle turning, the method further includes: The wheel steering angle is determined based on the vehicle speed and a preset speed threshold.

3. The wheel steering angle calibration model training method according to claim 2, characterized in that: Determining the wheel steering angle according to the vehicle speed and a preset speed threshold comprises: When the vehicle speed is less than or equal to the preset speed threshold, determining the wheel steering angle according to the vehicle wheelbase, the vehicle linear speed, and the vehicle angular speed; When the vehicle speed is greater than the preset speed threshold, the wheel steering angle is determined according to the vehicle wheelbase, front axle mass, rear axle mass, front wheel cornering stiffness, rear wheel cornering stiffness, and the turning radius from the steering center to the center of mass after the front and rear wheels of the vehicle corner.

4. The wheel steering angle calibration model training method according to claim 1, characterized in that: The model to be trained includes a polynomial model, and the relationship between the steering wheel angle and the wheel steering angle is as follows: y=w0+w1x+w2x 2 +w3x 3 +…+w n x n , Where y is the steering wheel angle, x is the wheel steering angle, and w i (i=0,1,2,...,n) is a model parameter, and n is an integer greater than 1.

5. The wheel steering angle calibration model training method according to claim 1, characterized in that: The updating of the model parameters of the model to be trained to reduce the loss of the prediction data and the sample data to obtain a calibration model includes: Divide the vehicle into multiple speed intervals, and update the model parameters of the model to be trained according to the predicted data and sample data in each speed interval to obtain a calibration model for each speed interval.

6. The wheel steering angle calibration model training method according to claim 1, characterized in that: The loss function in the process of updating the model parameters of the model to be trained is: Among them, MSE is the loss function, yi(i=1,2,...,m) is the true value of the i-th steering wheel angle sample, (i=1,2,...,m) is the predicted value of the i-th steering wheel angle sample, and m is the number of steering wheel angle samples.

7. A method for predicting a wheel steering angle, characterized in that: include: Acquiring operating parameters of the vehicle turning; wherein the operating parameters include a steering wheel angle and a vehicle speed; The wheel steering angle of the vehicle is predicted according to the operating parameters and the calibration model; wherein the calibration model is trained using the wheel steering angle calibration model training method described in any one of claims 1 to 6.

8. The wheel steering angle calibration model training system is characterized by: include: A sample data acquisition module, used to acquire sample data of vehicle turning, wherein the sample data includes a steering wheel angle, a vehicle speed range, and a wheel steering angle; A sample data prediction module, used for obtaining predicted data of the sample data in response to the model to be trained; A calibration model determination module is used to update the model parameters of the model to be trained to reduce the loss of the predicted data and the sample data and obtain a calibration model; wherein the calibration model is used to predict the wheel steering angle in response to the steering wheel angle and the vehicle speed range.

9. An electronic device, characterized in that: The electronic device comprises a processor and a memory for storing instructions executable by the processor; when the processor is configured to execute the instructions, the electronic device implements the method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The storage medium stores a computer program; the computer program is executed by a processor to implement the method according to any one of claims 1 to 7.

Images