Method and system for estimating road adhesion coefficient and adjusting traction force in real time

Abstract

The invention relates to a method and system for estimating a road adhesion coefficient and adjusting traction force in real time. The method comprises the steps that the initial road adhesion coefficient of a vehicle is predicted according to a road texture image; correcting the initial road adhesion coefficient according to the road contour information to obtain an initial corrected road adhesion coefficient; calculating the slip rate of the vehicle according to the wheel speed and the vehicle speed of the vehicle; correcting the initial corrected pavement adhesion coefficient according to the slip rate to obtain the initial corrected pavement adhesion coefficient; calculating the limit driving torque of the vehicle according to the final road adhesion coefficient; whether the driving torque is adjusted or not is judged according to the real-time driving torque and the limit driving torque of the vehicle, and if yes, the driving torque of the vehicle is adjusted; if not, the driving torque of the vehicle is not adjusted; according to the invention, a response can be quickly made at the moment that the vehicle is about to slip, so that the response time of the system is greatly shortened.

Description

Technical Field

[0001] This invention relates to the field of automotive safety control technology, specifically to a method and system for adjusting traction force in real time by estimating the road surface adhesion coefficient. Background Technology

[0002] This invention relates to automotive safety control technology, focusing on traction control and dynamic performance optimization of electric vehicles, encompassing intelligent sensor data fusion, road condition perception, vehicle dynamics modeling, and motor torque regulation. It is applicable to various types of electric vehicles in complex road conditions, especially improving driving safety and performance on low-adhesion surfaces such as wet, slippery, and icy surfaces, and is a key technology at the intersection of intelligent connected vehicles and electric vehicle power control systems.

[0003] Current mainstream TCS (Traction Control System) technology revolves around the core logic of "slip detection - torque correction." It acquires vehicle dynamics data through wheel speed sensors and IMU (Inductively Coupled Vehicle) sensors, and combines this with a vehicle dynamics model to estimate the vehicle's slip ratio and the road surface adhesion coefficient. When the slip ratio exceeds a set threshold, it is considered slippage and triggers a torque reduction command. This means that current technology only intervenes after slippage has already occurred, aiming to maintain driving stability by suppressing excessive slippage. However, this requires waiting for slippage to occur before intervention, which, for electric vehicles with extremely fast torque response, can easily lead to a risk of temporary loss of control due to control delays. Furthermore, relying on indirect parameters to estimate road conditions makes it difficult to directly obtain the adhesion coefficient, resulting in low torque utilization and poor power performance. Finally, it is not optimized for the characteristics of electric vehicle motors, and the torque adjustment strategy is mismatched with the motor's rapid response characteristics, easily leading to frequent tire wear and driving jerks. Summary of the Invention

[0004] To address the technical problems in existing technologies, such as the need to control the vehicle only after it has already slipped, and the risk of temporary loss of control due to control delays in electric vehicles with extremely fast torque response, this invention provides a method and system for real-time adjustment of traction force based on the estimated road surface adhesion coefficient.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: A method for adjusting traction force in real time by estimating the road surface adhesion coefficient includes the following steps: Predict the initial road adhesion coefficient of the vehicle based on the road texture image; The initial road surface adhesion coefficient is corrected based on the road surface contour information to obtain the initial corrected road surface adhesion coefficient; The slip ratio of a vehicle is calculated based on its wheel speed and overall speed. Based on the slip ratio, determine whether to correct the initial modified road surface adhesion coefficient. If yes, correct the initial modified road surface adhesion coefficient to obtain the final road surface adhesion coefficient; otherwise, set the final road surface adhesion coefficient to the initial modified road surface adhesion coefficient. The ultimate driving torque of the vehicle is calculated based on the final road surface adhesion coefficient. The system determines whether to adjust the driving torque based on the vehicle's real-time driving torque and the limit driving torque. If yes, the driving torque of the vehicle is adjusted; otherwise, the driving torque of the vehicle is not adjusted.

[0006] The beneficial effects of this invention are: by predicting the coefficient of adhesion in advance based on the road surface image, the driving torque of the vehicle can be controlled using the predicted coefficient of adhesion, and then real-time adjustments can be made during vehicle operation. It can be seen that this invention adjusts the driving torque based on the advance estimation of the road surface coefficient of adhesion, rather than intervening only after wheel slippage occurs. Combined with the fast response characteristics of electric vehicle motors, it can react quickly at the moment when the vehicle is about to slip, greatly shortening the system response time.

[0007] Based on the above technical solution, the present invention can be further improved as follows.

[0008] Furthermore, predicting the initial road adhesion coefficient of the vehicle based on the road surface texture image includes the following steps: The road surface texture image is classified using computer vision recognition methods to obtain road surface classification information; Construct a data table showing the correspondence between road surface classification and road surface adhesion coefficient; Based on the road surface classification information, the road surface adhesion coefficient corresponding to the road surface classification information is matched from the comparison relationship data table to obtain the initial road surface adhesion coefficient.

[0009] Furthermore, the initial road surface adhesion coefficient is corrected based on the road surface contour information to obtain an initial corrected road surface adhesion coefficient, including the following steps: The road surface contour information is obtained by using lidar to collect reflected waves from the road surface in the wheel area; The initial road surface adhesion coefficient is weighted and corrected based on the road surface contour information to obtain a first corrected road surface adhesion coefficient. The first modified road surface adhesion coefficient is further modified using the vehicle longitudinal dynamics equation to obtain the second modified road surface adhesion coefficient; The second modified road surface adhesion coefficient is optimized using the Kalman filter optimization algorithm to obtain the initial modified road surface adhesion coefficient.

[0010] Further, the initial road surface adhesion coefficient is weighted and corrected based on the road surface contour information to obtain a first corrected road surface adhesion coefficient, including the following steps: The weight value is determined based on the road surface contour information; The first modified road surface adhesion coefficient is obtained by multiplying the initial road surface adhesion coefficient by the weight value.

[0011] Furthermore, the first modified road surface adhesion coefficient is further modified using the vehicle longitudinal dynamics equation to obtain the second modified road surface adhesion coefficient, including the following steps: Construct the longitudinal dynamic equations of the vehicle; Substitute the wheel longitudinal force and wheel vertical load into the vehicle longitudinal dynamics equation to deduce the measured adhesion coefficient. The first modified road surface adhesion coefficient is corrected based on the measured adhesion coefficient to obtain the second modified road surface adhesion coefficient.

[0012] Furthermore, the second modified road surface adhesion coefficient is optimized using a Kalman filter optimization algorithm to obtain the initial modified road surface adhesion coefficient, including the following steps: Construct the Kalman state equation and the Kalman observation equation; The second modified road surface adhesion coefficient is dynamically smoothed using the Kalman state equation and the Kalman observation equation to filter out noise interference, thereby obtaining the second modified road surface adhesion coefficient.

[0013] Furthermore, the formula for calculating the vehicle's slip ratio based on the vehicle's wheel speed and overall speed is as follows: ; in, This represents the slip ratio. Indicates the wheel speed, This indicates the vehicle speed.

[0014] Further, determining whether to correct the initial modified road surface adhesion coefficient based on the slip ratio includes the following steps: If the slip ratio is less than or equal to a preset slip ratio threshold, then it is determined that the initial corrected road surface adhesion coefficient will not be corrected. If the slip ratio is greater than the preset slip ratio threshold, it is determined that the initial corrected road surface adhesion coefficient needs to be corrected.

[0015] Furthermore, the ultimate driving torque of the vehicle is calculated based on the final road surface adhesion coefficient, using the following formula: ; in, This indicates the ultimate driving torque. This represents the final road surface adhesion coefficient. Indicates the vertical load on the wheel. This indicates the rolling radius of the wheel.

[0016] To address the aforementioned technical problems, this invention also provides a system for real-time adjustment of traction force based on the estimated road surface adhesion coefficient, the specific technical content of which is as follows: A system for real-time adjustment of traction force based on prediction of road surface adhesion coefficient includes: The coefficient estimation module is used to predict the initial road adhesion coefficient of the vehicle based on the road texture image. The feedback optimization module is used to correct the initial road surface adhesion coefficient based on the road surface contour information to obtain an initial corrected road surface adhesion coefficient; calculate the vehicle's slip ratio based on the vehicle's wheel speed and vehicle speed; determine whether to correct the initial corrected road surface adhesion coefficient based on the slip ratio; if yes, correct the initial corrected road surface adhesion coefficient to obtain a final road surface adhesion coefficient; if no, set the final road surface adhesion coefficient to the initial corrected road surface adhesion coefficient. The torque dynamic adjustment module is used to calculate the vehicle's limit drive torque based on the final road surface adhesion coefficient; and to determine whether to adjust the drive torque based on the vehicle's real-time drive torque and the limit drive torque. If yes, the drive torque of the vehicle is adjusted; otherwise, the drive torque of the vehicle is not adjusted. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating how to adjust traction force in real time based on the estimated road surface adhesion coefficient, as described in an embodiment of the present invention. Figure 2 This is a schematic block diagram of a system for real-time adjustment of traction force based on the estimated road surface adhesion coefficient, as described in an embodiment of the present invention. Detailed Implementation

[0018] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0019] like Figure 1 As shown in the figure, this embodiment provides a method for adjusting traction force in real time by estimating the road surface adhesion coefficient, including the following steps: S1. Predict the initial road adhesion coefficient of the vehicle based on the road texture image; Predicting the initial road adhesion coefficient of a vehicle based on a road surface texture image includes the following steps: S101. Classify the road surface texture image based on computer vision recognition method to obtain road surface classification information; S102. Construct a data table showing the correspondence between road surface classification and road surface adhesion coefficient; S103. Match the road surface adhesion coefficient corresponding to the road surface classification information from the comparison relationship data table to obtain the initial road surface adhesion coefficient.

[0020] The vehicle-mounted camera captures images of the road surface in front of and around the vehicle at a set frame rate, extracting feature information such as road texture, color, and light reflection in the driving areas corresponding to the four tires. Millimeter-wave radar detects changes in road distance and reflected wave intensity in the areas corresponding to the four tires at a set frequency, and obtains data on road surface smoothness, hardness, and softness characteristics in each area; Wheel speed sensors collect the rotational speed of the four wheels in real time, while acceleration sensors collect the longitudinal and lateral acceleration of the vehicle. Combined with vehicle attitude information, the driving state of each wheel is distinguished.

[0021] An improved ResNet convolutional neural network model is used to classify the road surface images corresponding to the four tires. The obtained road surface classification information is matched with the pre-stored "road surface feature-μ database" to match the initial road surface adhesion coefficient. The road surface classification information includes 12 typical road surfaces such as dry road surface, wet road surface, or icy and snowy road surface. For example, the road surface adhesion coefficient μ of dry road surface can take the range of [0.7, 0.9], the road surface adhesion coefficient μ of wet road surface can take the range of [0.4, 0.6], and the road surface adhesion coefficient μ of icy and snowy road surface can take the range of [0.1, 0.3].

[0022] The final output is the initial road surface adhesion coefficients μ_FL_raw, μ_FR_raw, μ_RL_raw, and μ_RR_raw corresponding to the four tires; μ_FL_raw, μ_FR_raw, μ_RL_raw, and μ_RR_raw represent the initial road surface adhesion coefficients of the four wheels respectively.

[0023] S2. Correct the initial road surface adhesion coefficient based on the road surface contour information to obtain the initial corrected road surface adhesion coefficient; The initial road surface adhesion coefficient is corrected based on the road surface contour information to obtain the initial corrected road surface adhesion coefficient, including the following steps: S201. Use a lidar to collect reflected waves from the road surface in the wheel area to obtain the road surface contour information; S202. The initial road surface adhesion coefficient is weighted and corrected according to the road surface contour information to obtain the first corrected road surface adhesion coefficient. The initial road surface adhesion coefficient is weighted and corrected based on the road surface contour information to obtain a first corrected road surface adhesion coefficient, including the following steps: The weight value is determined based on the road surface contour information; The first modified road surface adhesion coefficient is obtained by multiplying the initial road surface adhesion coefficient by the weight value.

[0024] S203. The first modified road surface adhesion coefficient is further modified using the vehicle longitudinal dynamics equation to obtain the second modified road surface adhesion coefficient. The first modified road surface adhesion coefficient is further modified using the vehicle longitudinal dynamics equation to obtain the second modified road surface adhesion coefficient, including the following steps: Construct the longitudinal dynamic equations of the vehicle; Substitute the wheel longitudinal force and wheel vertical load into the vehicle longitudinal dynamics equation to deduce the measured adhesion coefficient. The first modified road surface adhesion coefficient is corrected based on the measured adhesion coefficient to obtain the second modified road surface adhesion coefficient.

[0025] S204. The second modified road surface adhesion coefficient is optimized using the Kalman filter optimization algorithm to obtain the initial modified road surface adhesion coefficient.

[0026] The second modified road surface adhesion coefficient is optimized using the Kalman filter optimization algorithm to obtain the initial modified road surface adhesion coefficient, including the following steps: Construct the Kalman state equation and the Kalman observation equation; The second modified road surface adhesion coefficient is dynamically smoothed using the Kalman state equation and the Kalman observation equation to filter out noise interference, thereby obtaining the second modified road surface adhesion coefficient.

[0027] Radar data correction: The radar data is weighted and corrected based on the intensity of the reflected waves in the corresponding areas of the four tires. For example, if the intensity of the reflected waves in the corresponding area of ​​a certain tire is high, the weighting ratio of the μ value of that tire is reduced, and the corrected μ value (μ_FL_corr1, μ_FR_corr1, μ_RL_corr1, μ_RR_corr1) is output.

[0028] Dynamic model correction: Based on the longitudinal dynamic equation (FX is the longitudinal force of the tire, m is the vehicle mass, and F is the rolling resistance), combined with the vertical load FZ of each wheel, the friction characteristics of the road surface where each wheel is located are inferred by μ=FX / FX, and the μ value after radar correction is further corrected. The corrected μ value is output, that is, the corrected μ values ​​of the four wheels are μ_FL_corr2, μ_FR_corr2, μ_RL_corr2, and μ_RR_corr2, respectively.

[0029] Kalman filter optimization: Construct the state equation and observation equation for μ estimation, perform dynamic smoothing on the fused μ value, filter out sensor noise (such as feature misidentification caused by backlighting of the camera), and finally output the high-precision μ value corresponding to the four tires. The high-precision μ value corresponding to the four tires is represented as μ_FL, μ_FR, μ_RL, and μ_RR, respectively, and is sent to the torque dynamic control module simultaneously.

[0030] S3. Calculate the vehicle's slip ratio based on the vehicle's wheel speed and vehicle speed; S4. Determine whether to correct the initial modified road surface adhesion coefficient based on the slip ratio. If yes, correct the initial modified road surface adhesion coefficient to obtain the final road surface adhesion coefficient; otherwise, set the final road surface adhesion coefficient to the initial modified road surface adhesion coefficient. The formula for calculating the slip ratio of a vehicle based on its wheel speed and overall speed is as follows: S = (V1 - V2) / V1; Wherein, S represents the slip ratio, V1 represents the wheel speed, and V2 represents the vehicle speed.

[0031] Determining whether to correct the initial modified road surface adhesion coefficient based on the slip ratio includes the following steps: S401. If the slip ratio is less than or equal to a preset slip ratio threshold, it is determined that the initial corrected road surface adhesion coefficient will not be corrected. S402. If the slip ratio is greater than the preset slip ratio threshold, it is determined that the initial corrected road surface adhesion coefficient needs to be corrected.

[0032] By continuously monitoring the vehicle's driving status, calculating the wheel slip ratio, determining the deviation between the actual driving state and the ideal state, and feeding back the deviation information to the road adhesion coefficient estimation module and the drive torque dynamic control module, the system achieves closed-loop control. This mainly involves the following steps: ① Step 3.1: Vehicle status monitoring, Based on the actual wheel speed V1 of the four wheels and the reference vehicle speed V2 of the four wheels, calculate the slip ratio S of the four wheels = (V1-V2) / V1.

[0033] ② Step 3.3: Deviation Judgment and Feedback When S exceeds a certain threshold, it indicates that the estimated μ value of the wheel is too high. The feature matching weight of the wheel needs to be re-optimized based on the S value to make the estimated μ value closer to the true μ value.

[0034] ③ Step 3.4: Dynamic correction iteration, Based on the feedback deviation, the feature matching weights are re-optimized to obtain the corrected μ value; The ultimate driving torque Tmax is recalculated based on the corrected μ value, and the driving torque command issued to the power system is updated accordingly, forming a complete closed loop of "perception-decision-correction". The ultimate driving torque refers to the limit torque at which the wheels will not slip when driving on this road surface. Once this limit torque is exceeded, the wheels will slip.

[0035] S5. Calculate the vehicle's ultimate driving torque based on the final road surface adhesion coefficient; The ultimate driving torque of the vehicle is calculated based on the final road surface adhesion coefficient, using the following formula: Tmax = μ × FZ × r; Where Tmax represents the ultimate driving torque, μ represents the final road surface adhesion coefficient, FZ represents the wheel vertical load, and r represents the wheel rolling radius.

[0036] S6. Determine whether to adjust the drive torque based on the vehicle's real-time drive torque and the limit drive torque. If yes, adjust the vehicle's drive torque; otherwise, do not adjust the vehicle's drive torque. If the real-time drive torque is greater than or equal to the limit drive torque, it is determined that the drive torque needs to be adjusted; if the real-time drive torque is less than the limit drive torque, it is determined that the drive torque does not need to be adjusted. Therefore, the method for adjusting the vehicle's drive torque is to adjust the real-time drive torque so that the limit drive torque is less than the limit drive torque.

[0037] The dynamic drive torque control module primarily determines the maximum drive torque corresponding to the adhesion limit based on the μ value obtained from the road surface adhesion coefficient estimation module. Based on this, it issues drive torque commands to the power system to ensure that the drive torque remains within a safe and efficient range. The module's process mainly consists of the following steps: Step 2.1: Determine the maximum driving torque corresponding to the adhesion limit of each wheel. The real-time vertical loads Fz_FL, Fz_FR, Fz_RL, and Fz_RR of each tire on the front and rear axles were calculated using a dynamic model. Calculate the maximum permissible driving torques Tmax_FL, Tmax_FR, Tmax_RL, and Tmax_RR for each wheel under the current road surface adhesion limit. These torques are the critical values ​​at which slippage does not occur.

[0038] Step 2.2: Monitor the current drive torque in real time. It receives the current actual driving torque Tact of each drive source and the current required torque Treq of the driver from the power system feedback, and compares them in real time with the maximum driving torque Tmax corresponding to the adhesion limit of each wheel.

[0039] Step 2.3: Issue a drive torque command based on the adhesion limit; The drive torque command is issued based on the drive type, including: Drive type of single motor controlling single drive shaft: The motor drive torque is the controlled object. When the actual motor torque Tact is close to the shaft Tmax (the sum of Tmax of the left and right wheels) and Treq is greater than Tmax, it indicates that the actual drive torque is about to reach the adhesion limit and there is still a need for increased torque. At this time, the TCS pre-control function is activated and a torque reduction request is sent to the power system to ensure that the actual torque does not exceed the adhesion limit, until Treq < Tmax and then it exits.

[0040] Drive type of single motor controlling single wheel: The motor drive torque is the controlled object. When the actual motor torque Tact is close to the wheel Tmax and the wheel Treq is greater than Tmax, it indicates that the actual drive torque is about to reach the adhesion limit and there is still a need for increased torque. At this time, the TCS pre-control function is activated and a torque reduction request is sent to the power system to ensure that the actual torque does not exceed the adhesion limit, until Treq < Tmax and then it exits.

[0041] This invention, through its embodiment, predicts the coefficient of adhesion in advance based on road surface images, and uses the predicted coefficient of adhesion to control the driving torque of the vehicle. It then makes real-time adjustments during vehicle operation. As can be seen, this invention adjusts the driving torque based on the advance estimation of the road surface coefficient of adhesion, rather than intervening only after wheel slippage occurs. Combined with the rapid response characteristics of electric vehicle motors, it can react quickly the moment the vehicle is about to slip, greatly shortening the system's response time.

[0042] By precisely controlling the drive torque to always be at the slippage boundary, the road surface adhesion potential can be maximized, avoiding the problem of traditional traction control systems conservatively limiting torque for fear of slippage. This algorithm can release the maximum allowable torque based on the real-time μ value, greatly improving acceleration performance on various road surfaces.

[0043] In terms of driving stability, lateral stability is significantly improved on low-traction surfaces (such as icy or slippery roads), reducing the probability of dangerous situations such as vehicle skidding and fishtailing. Simultaneously, precise torque control helps improve the energy efficiency of electric vehicles, extending their driving range to some extent.

[0044] By reducing tire wear and avoiding frequent wheel slippage and the impact wear caused by the drastic torque adjustment after slippage, this invention can effectively extend the service life of electric vehicle tires.

[0045] The driving experience is greatly enhanced, with significantly improved stability and comfort. Passengers experience significantly less bumping and shaking during acceleration, deceleration, and cornering, resulting in a smoother and more comfortable driving experience. Simultaneously, it reduces power interruptions caused by slippage and the jerking sensation during recovery, making the power delivery of electric vehicles smoother.

[0046] Overall, based on more accurate real-time μ estimation and predictive control, tire wear is expected to be reduced by 25% in complex road conditions such as ice and water accumulation, range is expected to increase by 8%-12%, and acceleration is expected to improve by 10%.

[0047] like Figure 2 As shown, in some other embodiments, a system for real-time adjustment of traction force based on estimated road surface adhesion coefficient is also provided, including: The coefficient estimation module is used to predict the initial road adhesion coefficient of the vehicle based on the road texture image. The feedback optimization module is used to correct the initial road surface adhesion coefficient based on the road surface contour information to obtain an initial corrected road surface adhesion coefficient; calculate the vehicle's slip ratio based on the vehicle's wheel speed and vehicle speed; determine whether to correct the initial corrected road surface adhesion coefficient based on the slip ratio; if yes, correct the initial corrected road surface adhesion coefficient to obtain a final road surface adhesion coefficient; if no, set the final road surface adhesion coefficient to the initial corrected road surface adhesion coefficient. The torque dynamic adjustment module is used to calculate the vehicle's limit drive torque based on the final road surface adhesion coefficient; and to determine whether to adjust the drive torque based on the vehicle's real-time drive torque and the limit drive torque. If yes, the drive torque of the vehicle is adjusted; otherwise, the drive torque of the vehicle is not adjusted.

[0048] Traditional traction control (TCS) relies heavily on real-time operating condition feedback (such as intervening only after the wheel slip rate exceeds a threshold), resulting in a "slippage first, correction later" lag problem. This is especially problematic in scenarios with sudden changes in the road surface adhesion coefficient μ, such as driving from a dry road to a flooded or icy road, which can easily lead to wasted power or loss of vehicle control. This invention, however, upgrades the control logic from "passive correction" to "active prediction" through real-time μ estimation and predictive control. This allows for early detection of changes in road surface μ, dynamically adjusting engine output torque or braking intervention intensity to prevent slippage, while simultaneously balancing power and safety. Current algorithms for estimating the adhesion coefficient primarily rely on dynamic models based on wheel speed and acceleration sensors. These models require "dynamic deviation signals" (such as slippage exceeding a threshold) generated by wheel slippage to correct the μ value in reverse, essentially "accurate calculation after slippage." In the initial stage of a sudden change in road surface μ, the lack of slippage signals leads to large μ estimation errors, causing a lag in TCS intervention.

[0049] This invention addresses this pain point at its root through "multi-source sensing fusion and pre-slip stage estimation," with the following specific breakthroughs: ① Preliminary μ calculation during the pre-slip phase: using vision and radar for early detection, without relying on slip signals; ② The dynamic model shifted from "dominant estimation" to "correction optimization," reducing its dependence on slippage signals; ③ Sensor redundancy and complementarity: From a single estimation method, an innovative collaborative robust design of "camera, millimeter-wave radar and dynamic sensor" is adopted, which greatly improves the estimation accuracy.

[0050] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the concept and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for estimating road adhesion coefficient for real-time adjustment of traction force, characterized in that, The method comprises the following steps: predicting an initial road adhesion coefficient of a vehicle according to a road texture image; correcting the initial road adhesion coefficient according to road profile information to obtain an initial corrected road adhesion coefficient; calculating a slip rate of the vehicle according to wheel speed and vehicle speed of the vehicle; judging whether to correct the initial corrected road adhesion coefficient according to the slip rate, and if yes, correcting the initial corrected road adhesion coefficient to obtain a final road adhesion coefficient; if no, setting the final road adhesion coefficient as the initial corrected road adhesion coefficient; calculating a limit driving torque of the vehicle according to the final road adhesion coefficient; judging whether to adjust the driving torque according to real-time driving torque of the vehicle and the limit driving torque, and if yes, adjusting the driving torque of the vehicle; if no, not adjusting the driving torque of the vehicle.

2. The method of estimating road adhesion coefficient real-time adjusting traction force according to claim 1, characterized in that, The method for predicting an initial road adhesion coefficient of a vehicle according to a road texture image comprises the following steps: classifying the road texture image based on a computer vision recognition method to obtain road classification information; constructing a comparison relationship data table of road classification and road adhesion coefficient; matching road adhesion coefficient corresponding to the road classification information from the comparison relationship data table according to the road classification information to obtain the initial road adhesion coefficient.

3. The method of estimating road adhesion coefficient real-time adjusting traction force according to claim 1, characterized in that, The method for correcting the initial road adhesion coefficient according to road profile information to obtain an initial corrected road adhesion coefficient comprises the following steps: collecting reflection waves of the road surface in the wheel area by using a laser radar to obtain the road profile information; weighting correcting the initial road adhesion coefficient according to the road profile information to obtain a first corrected road adhesion coefficient; further correcting the first corrected road adhesion coefficient by using a vehicle longitudinal dynamics equation to obtain a second corrected road adhesion coefficient; optimizing the second corrected road adhesion coefficient by using a Kalman filter optimization algorithm to obtain the initial corrected road adhesion coefficient.

4. The method of estimating road adhesion coefficient real-time adjusting traction force according to claim 3, characterized in that, The method for weighting correcting the initial road adhesion coefficient according to the road profile information to obtain a first corrected road adhesion coefficient comprises the following steps: determining a weight value according to the road profile information; calculating a product of the initial road adhesion coefficient and the weight value to obtain the first corrected road adhesion coefficient.

5. The method of estimating road adhesion coefficient real-time adjusting traction force according to claim 3, characterized in that, The method for further correcting the first corrected road adhesion coefficient by using a vehicle longitudinal dynamics equation to obtain a second corrected road adhesion coefficient comprises the following steps: constructing the vehicle longitudinal dynamics equation; substituting wheel longitudinal force and wheel vertical load into the vehicle longitudinal dynamics equation to back-calculate a measured adhesion coefficient; correcting the first corrected road adhesion coefficient according to the measured adhesion coefficient to obtain the second corrected road adhesion coefficient.

6. The method of estimating road adhesion coefficient real-time adjusting traction force according to claim 3, characterized in that, The method for optimizing the second corrected road adhesion coefficient by using a Kalman filter optimization algorithm to obtain the initial corrected road adhesion coefficient comprises the following steps: constructing a Kalman state equation and a Kalman observation equation; performing dynamic smoothing processing on the second corrected road adhesion coefficient by using the Kalman state equation and the Kalman observation equation to filter out noise interference to obtain the second corrected road adhesion coefficient.

7. The method of estimating road adhesion coefficient real-time adjusting traction force according to claim 1, characterized in that, The calculation formula of the slip rate of the vehicle according to the wheel speed and the vehicle speed is as follows: ； wherein denotes the slip ratio, denotes the wheel speed, denotes the vehicle speed.

8. The method of estimating road adhesion coefficient real-time adjusting traction force according to claim 1, characterized in that, According to the slip rate, whether the initial corrected road adhesion coefficient is corrected is determined, including the following steps: If the slip rate is less than or equal to a preset slip rate threshold, it is determined that the initial corrected road adhesion coefficient is not corrected; If the slip rate is greater than the preset slip rate threshold, it is determined that the initial corrected road adhesion coefficient is to be corrected.

9. The method of estimating road adhesion coefficient for real-time adjustment of tractive effort according to claim 1, wherein, According to the final road adhesion coefficient, the limit driving torque of the vehicle is calculated, and the formula is as follows: ； wherein, represents the limit driving torque, represents the final road adhesion coefficient, represents the wheel vertical load, represents the wheel rolling radius.

10. A system for estimating a road adhesion coefficient for real-time adjustment of traction, characterized in that It includes: The coefficient estimation module is configured to predict the initial road adhesion coefficient of the vehicle according to the road texture image. The feedback optimization module is configured to correct the initial road adhesion coefficient according to the road profile information to obtain an initial corrected road adhesion coefficient, calculate the slip rate of the vehicle according to the wheel speed and the vehicle speed, determine whether the initial corrected road adhesion coefficient is corrected according to the slip rate, correct the initial corrected road adhesion coefficient if yes to obtain a final road adhesion coefficient, and let the final road adhesion coefficient be the initial corrected road adhesion coefficient if no. The torque dynamic adjustment module is configured to calculate the limit driving torque of the vehicle according to the final road adhesion coefficient. According to the real-time driving torque of the vehicle and the limit driving torque, whether the driving torque is adjusted is determined, and the driving torque of the vehicle is adjusted if yes, and the driving torque of the vehicle is not adjusted if no.

Images