A method, device and apparatus for controlling a steer-by-wire system of a vehicle with road feel feedback

Abstract

The application relates to the technical field of automobile steering control, in particular to a kind of automobile steer-by-wire control method, equipment and device with road feeling feedback, the method comprises: obtaining rack force data through strain gauge sensor on steering rack, and obtaining yaw rate in vehicle running process;Calculate the first significant value and the second significant value of the rack force data in each time window, and then obtain the characteristic value of the overall frequent fluctuation and transient response impact of the rack force data in each time window;The characteristic value of the expected deviation and mutation degree of yaw rate in each time window is obtained by the deviation degree between each yaw rate and expected yaw rate and the mutation condition of all yaw rates in the time window, and then the road condition comprehensive interference coefficient is calculated, and the steer-by-wire of automobile is controlled by using sliding mode control algorithm.The application improves the control precision and anti-interference ability of steer-by-wire system.

Description

Technical Field

[0001] This application relates to the field of automotive steering control technology, specifically to a method, device, and apparatus for automotive steer-by-wire control with road feel feedback. Background Technology

[0002] As the core actuator of intelligent driving, the steer-by-wire system replaces the traditional mechanical connection with electronic signals, realizing the physical decoupling between the steering system and driving operation, which can significantly improve the vehicle's handling precision and agility.

[0003] However, because the steering wheel and steering wheels are completely decoupled, the system cannot directly transmit the actual force between the tires and the road surface, leaving the driver with little perception of road conditions and vehicle status. Therefore, road feel feedback is particularly important for steer-by-wire systems. In actual operation, steer-by-wire systems face interference from nonlinear factors such as component aging, momentary sensor failures, and severe fluctuations in external loads. This causes a difference between the feedback torque (such as return torque and road impact feel) and the feel under actual mechanical connection, especially in complex and changing road conditions, where its adaptive capability is poor, posing a serious challenge to precise steering control. There is a problem of insufficient accuracy in steer-by-wire control. Summary of the Invention

[0004] To address the aforementioned technical problems, the purpose of this application is to provide a method, device, and apparatus for steer-by-wire control of automobiles with road feel feedback. The specific technical solution adopted is as follows: In a first aspect, embodiments of this application provide a vehicle steer-by-wire control method with road feel feedback, comprising the following steps: The strain gauge sensor on the steering rack acquires rack force data and the yaw rate during vehicle movement. By utilizing the changing trend characteristics and fluctuation significance of the average level of rack force data within each time window, the first significant value of rack force data within each time window is obtained. The high-frequency energy ratio of rack force data is analyzed, and combined with the steepness of the rack force data distribution, the second significant value of rack force data within each time window is obtained. In this way, the characteristic values ​​of the overall frequent fluctuations and transient response impacts of rack force data within each time window are obtained. By analyzing the degree of deviation between each yaw angular velocity and the expected yaw angular velocity within the time window, as well as the abrupt changes in all yaw angular velocities within the time window, characteristic values ​​of the expected deviation and the degree of abrupt changes in the yaw angular velocity within each time window are obtained. By utilizing the temporal difference between two eigenvalues, the comprehensive road condition interference coefficient is obtained, and a sliding mode control algorithm is used to control the vehicle's steer-by-wire.

[0005] Preferably, the mean value of all rack force data within each time window is calculated, and a trend test is performed on the mean rack force of each time window and the multiple time windows before it to obtain a trend estimate. The absolute value of the normalized result of the trend estimate is used as the trend change coefficient.

[0006] Preferably, the first-order difference sequence of the rack force mean between each time window and the previous multiple time windows is obtained, and the standard deviation of the rack force mean between each time window and the previous multiple time windows is obtained. The number of times the signs of any two adjacent data in the first-order difference sequence are opposite is counted. The ratio of the obtained number to the total number of data in the first-order difference sequence is used as the alternation coefficient. The product of the alternation coefficient and the standard deviation is used as the significance of the change in rack force data in each time window.

[0007] Preferably, the first significant value of the rack force data within each time window is positively correlated with the trend change coefficient and the change significance, respectively.

[0008] Preferably, high-frequency data are extracted from the rack force data within the time window, the arithmetic sum of squares of the high-frequency data within the time window is taken as the impact response intensity, and the arithmetic sum of squares of all rack force data within the time window is taken as the total energy intensity. The ratio of the impact response intensity to the total energy intensity is taken as the high-frequency significance of the rack force data in each time window. The kurtosis of all rack force data in each time window is calculated, and the second significance value is positively correlated with the high-frequency significance and the kurtosis, respectively.

[0009] Preferably, the weighted sum of the first and second significant values ​​is used as the characteristic value of the overall frequent fluctuations and transient response impacts of the rack force data within each time window.

[0010] Preferably, the absolute values ​​of the differences between each yaw angular velocity within the time window and the expected yaw angular velocity extracted at the same time are accumulated. The ratio of this accumulated result to the sum of the absolute values ​​of all expected yaw angular velocities within the time window is used as the deviation characteristic value of each window. The abrupt change points in the yaw angular velocities within the time window are extracted. The absolute value of the difference between the normalized result of each abrupt change point within the time window and the mean of the normalized results of all yaw angular velocities except the abrupt change points is calculated. The average of the absolute values ​​of the differences calculated for all abrupt change points within each time window is used as the abrupt change coefficient of the yaw angular velocity within each time window. The characteristic values ​​of the expected deviation and the degree of abrupt change of the yaw rate within each time window are positively correlated with the deviation characteristic value and the abrupt change coefficient, respectively.

[0011] Preferably, the two feature values ​​corresponding to each time window and the previous multiple time windows are arranged in ascending order of time, the distance between the two sequences is calculated, and the negative exponential function of the distance is used as the similarity weight. The product of the average of the two feature values ​​obtained for each time window and the similarity weight is used as the comprehensive road condition interference coefficient of each time window. W times the mean of the normalized results of the comprehensive road condition interference coefficients of the previous multiple time windows is used as the switching gain of the next time window, where W is greater than 1.

[0012] Secondly, embodiments of this application also provide a vehicle steer-by-wire control device with road feel feedback, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described vehicle steer-by-wire control methods with road feel feedback.

[0013] Thirdly, embodiments of this application also provide a vehicle steer-by-wire control device with road feel feedback, which stores a computer program. When the computer program is executed by a processor, it implements any one of the above-described vehicle steer-by-wire control methods with road feel feedback.

[0014] As can be seen from the above, the steer-by-wire control method, device, and apparatus with road feel feedback provided in this application have at least the following beneficial effects: This application addresses the challenge of achieving stable and precise steering in steer-by-wire systems under complex and changing road conditions. It extracts rack force characteristics from multiple angles, including the changing trend of road surface adhesion, frequent fluctuations, and transient impacts from road unevenness. Simultaneously, it combines the expected deviation of yaw rate with abrupt change detection to characterize abnormal vehicle dynamic response, thus accurately assessing the comprehensive interference of road conditions on steer-by-wire. Based on this interference coefficient, the main controller optimizes the switching gain of the sliding mode control. Its advantage lies in achieving mechanical decoupling while adaptively compensating for nonlinear disturbances caused by complex road conditions, thereby improving the control accuracy and anti-interference capability of the steer-by-wire system. Attached Figure Description

[0015] Figure 1 A flowchart illustrating the steps of a vehicle steer-by-wire control method with road feel feedback provided in this application. Detailed Implementation

[0016] The following description, in conjunction with the accompanying drawings, details a specific scheme for a vehicle steer-by-wire control method, device, and apparatus with road feel feedback provided in this application.

[0017] Please see Figure 1 The diagram illustrates a flowchart of a vehicle steer-by-wire control method with road feel feedback according to an embodiment of this application, including the following steps: Step 1: Obtain rack force data through strain gauge sensors on the steering rack, and obtain the yaw rate during vehicle movement.

[0018] In this embodiment, the steer-by-wire system mainly consists of three parts: a steering wheel module, a steering actuator module, and a main controller module. The steering wheel module includes a road feel feedback motor and a steering angle sensor, primarily responsible for receiving the steering angle transmitted by the steering wheel and providing road feel feedback information. The steering actuator module includes a steering actuator motor and a transmission mechanism. Its main function is to receive a steering signal from the main controller, generate corresponding torque and steering angle from the actuator motor to drive the actuator, thereby causing the wheels to rotate by the corresponding angle to complete the steering operation. The main controller processes sensor data through a high-speed network and sends control signals to the corresponding actuators according to the control strategy. To monitor road surface conditions, this embodiment installs strain gauge sensors on the steering rack to collect rack force data, setting the rack force data acquisition frequency to 5kHz. Simultaneously, a microelectromechanical system (MEMS) gyroscope is used to collect the yaw rate during vehicle movement, setting the yaw rate acquisition frequency to 200Hz.

[0019] Step 2: Utilize the changing trend characteristics and fluctuation significance of the average level of rack force data within each time window to obtain the first significant value of rack force data within each time window. Analyze the high-frequency energy proportion of rack force data and, combined with the steepness of rack force data distribution, obtain the second significant value of rack force data within each time window. Then, obtain the characteristic values ​​of the overall frequent fluctuations and transient response impacts of rack force data within each time window.

[0020] Road feel feedback is crucial for the stability of steer-by-wire systems. In traditional steering systems, the driver directly senses the interaction force between the tires and the road surface through a mechanical connection, allowing for intuitive adjustments to steering input to maintain stability. Steer-by-wire systems, however, disconnect this mechanical connection. Without precise and timely road feel feedback, the driver may fail to perceive wheel slippage and adhesion limits, leading to potentially delayed or excessive steering input and disrupting the vehicle's dynamic balance. Furthermore, nonlinear factors and complex, variable road conditions can easily affect the reliability of road feel feedback, reducing steer-by-wire accuracy. The following analysis addresses this issue.

[0021] First, the torque generated by the rotation of the wheel is transmitted to the rack through the steering mechanism. The rack force data can directly reflect the mechanical state of the contact interface between the tire and the road surface, which includes two key types of information: road adhesion characteristics and road unevenness.

[0022] Specifically, the difference in frictional resistance to the tires from different road surfaces (such as dry asphalt, wet and slippery surfaces, gravel, or ice and snow) directly changes the rack force. On low-adhesion surfaces, the tire's lateral force easily reaches its saturation limit, resulting in a significant reduction in the rack force required to generate the same steering angle. Therefore, changes in road surface adhesion affect the stability of steer-by-wire. When the adhesion varies significantly across different road sections, the rack force data will exhibit a clear trend in amplitude. For example, when suddenly transitioning from dry asphalt to a wet and slippery surface, the road adhesion decreases significantly, and the overall rack force data becomes smaller.

[0023] Therefore, to monitor road surface conditions in real time during driving, the duration of the time window is set to 0.1 seconds in this embodiment. Then, the mean of all rack force data within each time window is calculated. A trend test algorithm is used to test the trend of the mean rack force for the i-th time window and the preceding N (30 values). If the number of historical time windows is less than 30, missing values ​​are filled using edge copying. The output of the trend test algorithm is a trend estimate. In this embodiment, Sen's Slope test algorithm is used; the specific trend test process is existing technology and will not be elaborated upon here. Furthermore, the trend estimate is normalized using the max-min normalization method in this embodiment, with a normalization range of [-1, 1]. The absolute value of the normalization result is used as the trend change coefficient of the rack force data within the i-th time window. The larger the trend change coefficient, the more obvious the trend characteristic of the mean rack force change in a short period of time.

[0024] Furthermore, the road surface adhesion state may exhibit alternating or recurring changes, such as transitioning from dry asphalt to a wet and slippery surface and then back to dry asphalt, leading to frequent fluctuations in rack force. Therefore, a first-order difference sequence of the mean rack force values ​​for the i-th time window and the previous N time windows is obtained. The number of times the signs of adjacent data points in the first-order difference sequence are opposite is counted, and the ratio of this count to the total number of data points in the first-order difference sequence is used as the alternation coefficient. Then, the standard deviation of the mean rack force values ​​for the i-th time window and the previous N time windows is calculated. The product of the alternation coefficient and the standard deviation is used as the significance of the rack force data change within the i-th time window. The greater the significance, the more pronounced the frequent alternation of the road surface adhesion state.

[0025] Furthermore, in this embodiment, a first significant value of the rack force data within each time window will be constructed. This first significant value is positively correlated with both the trend change coefficient and the change significance. It should be noted that in this embodiment, the positive correlation means that the dependent variable increases (decreases) as the independent variable increases (decreases), and the variables exhibit the same trend. The specific calculation relationship can be set by the implementer according to actual circumstances; this embodiment does not impose any special restrictions on this. Preferably, in this embodiment, the first significant value of the rack force data within the i-th time window is the product of the trend change coefficient and the change significance of the rack force data within the i-th time window. The larger the first significant value of the rack force data within the i-th time window, the more significant the overall change or frequent fluctuations in the road surface adhesion state.

[0026] Furthermore, when the wheels drive over potholes, seams, or speed bumps, the vertical and longitudinal impacts on the tires are transmitted through the steering system, manifesting as high-frequency, transient pulse characteristics in the rack force data. Therefore, the amplitude and frequency characteristics of the rack force can reflect information about the unevenness or obstacles on the road surface. Strong impacts may cause the steering actuator to produce unexpected transient responses, affecting the stability of steer-by-wire.

[0027] Therefore, high-frequency data is extracted from the rack force data within each time window. Taking the i-th time window as an example, high-frequency data is extracted from the rack force data within the time window. The arithmetic sum of the squares of the high-frequency data within the time window is used as the impact response intensity, and the arithmetic sum of the squares of all rack force data within the time window is used as the total energy intensity. The ratio of the impact response intensity to the total energy intensity is used as the high-frequency significance of the rack force data in each time window. Preferably, this embodiment uses a Butterworth bandpass filter to bandpass filter the rack force data within the time window to extract high-frequency data, so as to retain the high-frequency components that can reflect the road impact and obtain high-frequency data. In this embodiment, a fourth-order Butterworth bandpass filter is used, and the upper and lower cutoff frequencies of the Butterworth bandpass filter are 5 Hz and 200 Hz, respectively. In practical applications, implementers can also use high-pass filtering to filter the rack force data within each time window to extract high-frequency data. This embodiment does not impose any special restrictions on this. The greater the high-frequency significance, the higher the proportion of high-frequency impact energy caused by road potholes and joints in the rack force data within that time window.

[0028] Furthermore, under the influence of road surface unevenness impact, the stronger and more intense the transient impact in the rack force data, the more obvious the extreme peak value appears. Then, the kurtosis of all rack force data in the i-th time window is calculated. The larger the kurtosis, the steeper the distribution of rack force data in that time window, and the greater the possibility of extreme transient pulses.

[0029] Then, the second significance value of the rack force data within each time window is calculated by combining the high-frequency significance and the kurtosis, wherein the second significance value is positively correlated with both the high-frequency significance and the kurtosis. Preferably, in this embodiment, the product of the high-frequency significance and the kurtosis of the rack force data within the i-th time window is taken as the second significance value of the rack force data within the i-th time window. The larger the obtained second significance value, the more obvious the proportion of high-frequency signals and transient impact characteristics caused by road surface unevenness in the rack force data.

[0030] Furthermore, the weighted sum of the first and second significant values ​​is used as the characteristic value of the overall frequent fluctuations and transient response impacts of the rack force data within each time window. In this embodiment, the calculation formula for the characteristic value of the overall frequent fluctuations and transient response impacts of the rack force data within the i-th time window is as follows: In the formula, , These are the weighting coefficients, and When the road surface has obvious unevenness, The value should be relatively large to enhance sensitivity to high-frequency transient impacts. In this embodiment, it is set to... , Among them, the results The larger the value, the more significant the road disturbance characteristics reflected by the rack force data within the i-th time window.

[0031] In the above formula, the parameter The overall trend and frequent fluctuation characteristics of road surface adhesion state were quantified, and the parameters were... This reflects the proportion of high-frequency information and transient impact characteristics in the rack force data caused by road surface unevenness. Weighted fusion is used to accurately capture the road surface condition information in the rack force data.

[0032] Step 3: By analyzing the degree of deviation between each yaw angular velocity and the expected yaw angular velocity within the time window, as well as the abrupt changes in all yaw angular velocities within the time window, characteristic values ​​of the expected deviation and the degree of abrupt changes in the yaw angular velocity within each time window are obtained.

[0033] Furthermore, yaw rate reflects the overall rotational motion response of the vehicle, including its dynamic stability and road surface adhesion utilization. When a vehicle is in a stable steering state, the actual yaw rate should ideally match the expected yaw rate. However, complex changes in road conditions can affect vehicle stability, leading to significant deviations between the actual yaw rate and the model's expected value, and even triggering skidding or loss of trajectory control.

[0034] Therefore, this embodiment utilizes a two-degree-of-freedom linear vehicle model to obtain the desired yaw rate. This method is a known technique, and implementers can also use other existing methods to extract the desired yaw rate during vehicle operation. The sampling frequency of the desired yaw rate is the same as the sampling frequency of the actual yaw rate. For the i-th time window, the absolute value of the difference between each yaw rate within the time window and the desired yaw rate at the same time is accumulated. The ratio of this accumulated result to the sum of the absolute values ​​of all desired yaw rates within the time window is used as the deviation characteristic value of each window. To avoid the denominator being zero during the ratio calculation, a very small positive number is added to the denominator, which is set to [value missing]. The larger the obtained deviation characteristic value, the greater the deviation between the actual value and the expected value of the yaw rate within that time window.

[0035] In addition, uneven road surface adhesion can also affect steering stability. Especially when there is a large difference in the adhesion conditions between the left and right wheels (for example, one side is on an asphalt road while the other side is on an icy or snowy road), the vehicle will generate unexpected yaw moment, resulting in a significant change in the actual yaw rate.

[0036] Therefore, for all yaw angular velocities within the i-th time window, this embodiment uses the median absolute deviation (MAD) anomaly detection method to detect abrupt changes, outputting an anomaly probability value for each yaw angular velocity. An anomaly probability threshold is set to three times the MAD. The specific process of the median absolute deviation (MAD) anomaly detection method is existing technology and will not be elaborated in this embodiment. Further, yaw angular velocities with anomaly probability values ​​greater than or equal to this threshold are considered abrupt changes. The minimax method is used to normalize all yaw angular velocities within the time window, with a normalization range of [0,1]. The absolute value of the difference between the normalized result for each abrupt change point and the mean of the normalized results for all yaw angular velocities except the abrupt change point is calculated. The average of the absolute values ​​of these differences calculated for all abrupt changes is used as the abrupt change coefficient of the yaw angular velocity within the i-th time window. The obtained abrupt change coefficient reflects the degree of abrupt change in the actual yaw angular velocity. Specifically, if the abnormal probability values ​​of all yaw angular velocities within the i-th time window are not greater than the abnormal probability threshold, it indicates that no mutation point has been detected, and the mutation coefficient is set to 0.

[0037] Therefore, based on the deviation characteristic value and the mutation coefficient, the characteristic values ​​of the expected deviation and mutation degree of the yaw rate within each time window are obtained. Specifically, the characteristic values ​​of the expected deviation and mutation degree of the yaw rate within each time window are positively correlated with the deviation characteristic value and the mutation coefficient, respectively.

[0038] Preferably, in this embodiment, the characteristic value of the expected deviation and the degree of change of the yaw rate in the i-th time window is calculated by the sum of the deviation characteristic value and the change coefficient. The larger the characteristic value of the expected deviation and the degree of change of the yaw rate, the more significant the deviation characteristic and the degree of change of the yaw rate from the expected value in the time window.

[0039] Step 4: Utilize the temporal difference between the two feature values ​​to obtain the comprehensive road condition interference coefficient, and use the sliding mode control algorithm to control the vehicle's steer-by-wire.

[0040] Furthermore, changes in road surface adhesion conditions first affect the tires, causing the torque generated by wheel rotation to be transmitted to the rack via the steering mechanism, resulting in an immediate response in the rack force data. In contrast, yaw rate is the result of changes in tire force caused by changes in road surface conditions, processed through the dynamic response of the vehicle's suspension, mass, and other inertial systems; its change lags behind the rack force. While the road condition information reflected by rack force and yaw rate has its own emphasis, they are also coupled. For example, severe road unevenness impacts can exacerbate yaw rate anomalies. The more consistent the change characteristics of rack force and yaw rate under complex road conditions, the more severe the impact of road conditions on steer-by-wire.

[0041] Therefore, the two characteristic values ​​corresponding to each time window and the previous multiple time windows are arranged in ascending order of time, and then the distance between the two sequences is calculated. The ratio of the average of the two characteristic values ​​obtained for each time window to the distance is used as the comprehensive road condition interference coefficient for each time window. In the process of calculating the ratio, a very small positive number is added to the denominator to avoid the situation where the denominator is zero and cannot be calculated.

[0042] Preferably, in this embodiment, the characteristic values ​​of the overall frequent fluctuations and transient response impacts of the rack force data corresponding to the i-th time window and the seven time windows preceding it are arranged in ascending order of time, denoted as the first sequence in this embodiment. The characteristic values ​​of the expected deviation and abrupt change degree of the yaw rate corresponding to the i-th time window and the seven time windows preceding it are arranged in ascending order of time, denoted as the second sequence. Then, the SBD (Shape-Based Distance) distance between the first and second sequences is calculated. The smaller the distance, the more consistent the influence of road conditions on the rack force and yaw rate. Then, the average of the two characteristic values ​​obtained for the i-th time window is calculated, and the negative exponential function of the SBD distance (in this embodiment, ...) is applied. Where e is a natural constant) is used as the similarity weight, and the average value of this similarity weight is used as the comprehensive road condition interference coefficient for the i-th time window, where the minimum positive number is taken as... The higher the overall road condition interference coefficient, the greater the degree to which the steer-by-wire control is affected by various factors.

[0043] In this embodiment, by deeply analyzing the measured rack force data and yaw rate under the influence of complex road conditions during vehicle driving, and further considering the consistency of the two under the influence of road conditions within a local time window, the overall interference of road conditions on steer-by-wire is accurately evaluated.

[0044] Furthermore, this embodiment employs a sliding mode control (SMC) algorithm for steering control. The gain of the sliding mode controller is adjusted based on road condition feedback information to suppress the impact of external disturbances on the steering system. Specifically, the larger the obtained comprehensive road condition disturbance coefficient, the greater the influence of external road surface disturbances on the steering system. Therefore, a larger switching gain is set in the SMC. The SMC uses an exponential reaching law to ensure that the system state can quickly approach and maintain on the sliding surface, thereby effectively overcoming external disturbances. Conversely, if the comprehensive road condition disturbance coefficient is smaller, it indicates a less severe road surface disturbance, and a smaller switching gain can be set to reduce control output chatter and improve the smoothness of steering operations.

[0045] Preferably, in this embodiment, the switching gain is set to a value range of [0,2], and optimization is performed in units of time windows. The maximum and minimum values ​​of the comprehensive road condition interference coefficient for each time window within a recent period (set to 3 minutes in this embodiment) are obtained and used to perform maximum-minimum normalization processing on the comprehensive road condition interference coefficients obtained from each time window. The normalization interval is (0,1). The mean of the normalized results of the comprehensive road condition interference coefficients of multiple time windows preceding the current time window is denoted as T. In this embodiment, the mean of the normalized results of the comprehensive road condition interference coefficients of multiple time windows preceding the current time window is denoted as T. The switching gain is used as the switching gain for the next time window. To avoid cold start issues, the initial value of the switching gain is set to 1. When the cumulative number of time windows is insufficient, the gain parameter is reset to the initial value. The main controller calculates the steering angle based on the optimized parameters and sends control signals to the corresponding actuators, thereby realizing the steer-by-wire control of the vehicle. This helps improve the accuracy of steer-by-wire control.

[0046] Based on the same inventive concept as the above method, this application embodiment also provides a vehicle steer-by-wire control device with road feel feedback, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described vehicle steer-by-wire control methods with road feel feedback.

[0047] Meanwhile, this application also provides a vehicle steer-by-wire control device with road feel feedback, which stores a computer program. When the computer program is executed by a processor, it implements any of the above-described vehicle steer-by-wire control methods with road feel feedback.

[0048] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the protection scope of this application.

Claims

1. A method for steer-by-wire control of a vehicle with road feel feedback, characterized in that, Includes the following steps: The strain gauge sensor on the steering rack acquires rack force data and the yaw rate during vehicle movement. By utilizing the changing trend characteristics and fluctuation significance of the average level of rack force data within each time window, the first significant value of rack force data within each time window is obtained. The high-frequency energy ratio of rack force data is analyzed, and combined with the steepness of the rack force data distribution, the second significant value of rack force data within each time window is obtained. In this way, the characteristic values ​​of the overall frequent fluctuations and transient response impacts of rack force data within each time window are obtained. By analyzing the degree of deviation between each yaw angular velocity and the expected yaw angular velocity within the time window, as well as the abrupt changes in all yaw angular velocities within the time window, characteristic values ​​of the expected deviation and the degree of abrupt changes in the yaw angular velocity within each time window are obtained. By utilizing the temporal difference between two eigenvalues, the comprehensive road condition disturbance coefficient is obtained, and a sliding mode control algorithm is used to control the vehicle's steer-by-wire.

2. The steer-by-wire control method for a vehicle with road feel feedback as described in claim 1, characterized in that, Calculate the mean of all rack force data within each time window, perform a trend test on the mean rack force of each time window and the multiple time windows before it, obtain a trend estimate, and use the absolute value of the normalized result of the trend estimate as the trend change coefficient.

3. The steer-by-wire control method for a vehicle with road feel feedback as described in claim 2, characterized in that, Obtain the first-order difference sequence of the rack force mean between each time window and the previous multiple time windows, and obtain the standard deviation of the rack force mean between each time window and the previous multiple time windows. Count the number of times the signs of any two adjacent data in the first-order difference sequence are opposite. The ratio of the obtained number to the total number of data in the first-order difference sequence is used as the alternation coefficient. The product of the alternation coefficient and the standard deviation is used as the significance of the change in rack force data in each time window.

4. The steer-by-wire control method for a vehicle with road feel feedback as described in claim 3, characterized in that, The first significant value of the rack force data within each time window is positively correlated with the trend change coefficient and the change significance, respectively.

5. The steer-by-wire control method for a vehicle with road feel feedback as described in claim 1, characterized in that, High-frequency data are extracted from the rack force data within a time window. The arithmetic sum of squares of the high-frequency data within the time window is taken as the impact response intensity, and the arithmetic sum of squares of all rack force data within the time window is taken as the total energy intensity. The ratio of the impact response intensity to the total energy intensity is taken as the high-frequency significance of the rack force data in each time window. The kurtosis of all rack force data in each time window is calculated. The second significance value is positively correlated with the high-frequency significance and the kurtosis, respectively.

6. The steer-by-wire control method for a vehicle with road feel feedback as described in claim 1, characterized in that, The weighted sum of the first and second significant values ​​is used as the characteristic value of the overall frequent fluctuations and transient response impacts of the rack force data within each time window.

7. The steer-by-wire control method for a vehicle with road feel feedback as described in claim 1, characterized in that, The absolute values ​​of the differences between each yaw angular velocity within a time window and the expected yaw angular velocity extracted at the same time are accumulated. The ratio of this accumulated result to the sum of the absolute values ​​of all expected yaw angular angular velocities within the time window is used as the deviation characteristic value of each window. The abrupt change points in the yaw angular angular velocities within the time window are extracted. The absolute value of the difference between the normalized result of each abrupt change point within the time window and the mean of the normalized results of all yaw angular angular velocities except for the abrupt change points is calculated. The average of the absolute values ​​of the differences calculated for all abrupt change points within each time window is used as the abrupt change coefficient of the yaw angular angular velocities within each time window. The characteristic values ​​of the expected deviation and the degree of abrupt change of the yaw rate within each time window are positively correlated with the deviation characteristic value and the abrupt change coefficient, respectively.

8. A steer-by-wire control method for a vehicle with road feel feedback as described in claim 1, characterized in that, Arrange the two feature values ​​corresponding to each time window and the previous multiple time windows in ascending order of time, calculate the distance between the two sequences, and use the negative exponential function of the distance as the similarity weight. Multiply the average of the two feature values ​​obtained for each time window by the similarity weight as the comprehensive road condition interference coefficient for each time window. Use W times the mean of the normalized results of the comprehensive road condition interference coefficients of the previous multiple time windows as the switching gain for the next time window, where W is greater than 1.

9. A vehicle steer-by-wire control device with road feel feedback, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the vehicle steer-by-wire control method with road feel feedback as described in any one of claims 1-8.

10. A steer-by-wire control device for automobiles with road feel feedback, comprising a stored computer program, characterized in that, When the computer program is executed by the processor, it implements a vehicle steer-by-wire control method with road feel feedback as described in any one of claims 1-8.

Images