CN111845710B - Whole vehicle dynamic performance control method and system based on road surface adhesion coefficient identification - Google Patents

Abstract

The vehicle dynamic performance control method and system based on the identification of road adhesion coefficient, for traditional power four-wheel drive vehicles, comprehensively considers the collaborative working mechanism of each vehicle subsystem and the horizontal and longitudinal coupling influence mechanism of vehicle dynamics, and proposes an improved Global control strategy for the vehicle's horizontal and longitudinal dynamic performance. It combines modern control theory and vehicle dynamics, and uses an agile dynamics control method to improve the agility, smoothness and handling stability of the vehicle in corners on different road surfaces, and improve the driver's driving experience. Compared with the existing technology, it can improve the steering response of the vehicle under daily use conditions, improve the safety, driving stability and cornering smoothness of the vehicle. It can also be used for reference in the field of automatic driving vehicle control for automatic driving. Optimization of vehicle control technology, especially for coupling control in curves.

Description

Vehicle dynamic performance control method and system based on road adhesion coefficient identification

Technical field

The invention relates to the technical field of vehicle dynamic performance control, and specifically relates to a global control strategy for improving the driving performance of curved vehicles based on road adhesion coefficient.

Background technique

Since the mid-1980s, some well-known automobile manufacturers and researchers at home and abroad have successively carried out research on chassis integrated control and achieved many research results. Different chassis integrated control methods provide diverse vehicle control strategies, and various algorithm solutions improve vehicle performance to varying degrees. For example, the dynamics integrated management system installed in some existing commercially available models can integrate the chassis control system to the maximum extent from both the software and hardware levels, integrating drive, braking and steering control. The intelligent management improves the dynamic performance of the vehicle and enables intervention control before the vehicle becomes unstable. In the existing technology that realizes the integrated control of braking and steering partly through algorithms, the global integrated control of braking, steering and suspension can also be better formed. However, in the current existing technology, little consideration is given to the identification of the maximum road adhesion coefficient under normal driving conditions of the vehicle. Therefore, the dynamics of the global vehicle control strategy provided are obviously insufficient, making it unsuitable for the vehicle. The adaptability to actual driving conditions still has shortcomings that cannot be overcome.

Contents of the invention

Without solving the technical problems existing in the above-mentioned prior art, the present invention provides a vehicle dynamic performance control method based on the identification of the maximum road adhesion coefficient, which mainly includes the following steps:

Step 1: Determine the working conditions based on the vehicle's driving data and identify the vehicle's current situation of entering, exiting a curve, or driving in a non-curve;

Step 2: Use driving experience data to calibrate the vehicle agility control parameters corresponding to different road surfaces and different vehicle speeds, that is, the expected longitudinal acceleration;

Step 3: Combine the working conditions determined in Step 1, as well as the expected longitudinal acceleration, actual longitudinal acceleration, and brake pedal pressure, determine the driver's intention and decide to output the longitudinal acceleration command;

Step 4: Calculate the longitudinal tangent stiffness and longitudinal secant stiffness of the vehicle based on the longitudinal adhesion coefficient and longitudinal slip rate of the single wheel, and use the recursion of Gaussian distribution to estimate the maximum road adhesion coefficient;

Step 5: For the longitudinal acceleration command output by Step 3, use the estimated maximum road adhesion coefficient to provide a gain for the longitudinal acceleration command, and realize dynamic adjustment of torque according to the gained acceleration to achieve vehicle acceleration and deceleration following adjust.

Further, in the step one, the working condition is judged based on the vehicle's driving data, specifically using the following data: the vehicle's lateral acceleration G _y , lateral impact degree Steering wheel angle θ and steering wheel angular speed/>

Further, the second step specifically uses driving experience data: first-order inertial system delay time T, system gain C _xy , and through data analysis, the understeering degree k and the measured vehicle wheelbase L, steering wheel angle to front wheel angle are obtained The transmission ratio i, steering wheel angle δ and steering wheel angular speed For the vehicle speed V, the following expected longitudinal acceleration G _x is obtained, where s represents the Laplace factor:

Further, judging the driver's intention and deciding to output the longitudinal acceleration command in step three specifically includes the following process:

When it is judged that the working condition is entering a curve: If the vehicle's current _longitudinal acceleration a

(1). If the current longitudinal acceleration of the vehicle a _x ≤ G _x , it means that the driver’s braking intensity is insufficient and it is necessary to increase the braking intensity. Then the longitudinal acceleration for decision-making should be a _x ;

(2). If the current longitudinal acceleration of the vehicle a _x > G _x , it means that the driver may avoid obstacles or make an emergency stop. At this time, in order to ensure the driver's driving safety, the longitudinal acceleration for decision-making is G _x ;

(3). If the current longitudinal acceleration of the vehicle a _x > 0, it means that the driver has insufficient driving experience, and the longitudinal acceleration for decision-making is G _x ;

When it is judged that the working condition is exiting a curve, the following situations are corresponding:

(1 ₎ . If _the current acceleration of the vehicle _a

(2). If the vehicle’s current acceleration a _x =0, the determined longitudinal acceleration is G _x ;

(3). _If the current vehicle acceleration _a

Further, the specific process of outputting the maximum road adhesion coefficient in step 4 is:

pass The relationship is estimated to obtain the longitudinal attachment coefficient μ _x of a single wheel; where, /> is the estimated value of the pure longitudinal force of a single wheel, which is estimated from the wheel edge driving moment. Its calculation formula is/> Among them, Re is the effective rolling radius of the tire; among them, F _z,i and T _t,i are the vertical force and wheel rim torque of the single wheel respectively, and the superscript ^ indicates the estimated value of the corresponding parameter. Whether it is an SUV or a car driven by a four-wheel hub motor, the wheel drive torque is the cause of vehicle acceleration. In other words, the use adhesion coefficient estimated from the wheel drive torque is always "leading" in phase with the vehicle body acceleration, so/> The vehicle's longitudinal tangent stiffness is defined as/> However, since most of the time the vehicle is driving on a small slope and the vehicle motion does not enter a strong nonlinear zone, the vertical load transfer ratio of the tires caused by the acceleration, braking and lateral movement of the vehicle itself is very small. Therefore, the vertical load transfer ratio is ignored. The effect of load change rate, and thus the tangent stiffness can be further defined as/> The linear area of the tire in the slip rate-use adhesion coefficient curve, that is, the corresponding curve slope at the starting point of the curve, is the secant stiffness k of the vehicle, where the longitudinal secant stiffness is k _x . According to the unitire model, it can be known _that the product of the _real -time normalized secant stiffness and the real-time slip rate is the use of adhesion coefficient _μ The secant stiffness of , we can get/> Among them,/> is the normalized pure longitudinal secant stiffness of a single wheel,/> is the normalized pure longitudinal tangent stiffness of a single wheel. When/> This shows that the corresponding slip rate of the working point used for Gaussian recursion cannot be too small, and the obtained working point belongs to a period of data in which the slip rate continues to increase under a certain drive;/> This shows that the corresponding longitudinal acceleration of the working point used for Gaussian recursion cannot be too small, and the obtained working point belongs to a period of data with increasing longitudinal acceleration under a certain drive; v _x > v _{x, thro} > 0, which shows that For the working point of Gaussian recursion, its corresponding longitudinal vehicle speed cannot be too small. This article selects a period of data with increasing driving force under a certain drive to estimate μ _{x and max} . This is because, in the process of increasing driving force, although there is interference from tread vibration, compared with the process of falling driving force, the tread is in a "tight" state during this process, and the corresponding data points are different under different road surfaces. The separation characteristics are more "obvious" and more suitable for μ _{x, max} estimation. After the Save_flag judgment is completed, Cnt_flag judgment needs to be performed. The main judgment logic is as follows: Save_flag==1, that is, at least the Save_flag condition is met; Related to the relationship between the normalized tangent stiffness XBS and the normalized secant stiffness k, when the tire force operating point changes from the linear area to the quasi-linear area, the relationship between XBS and k, which are originally equal, will become k>XBS, and As the degree of slippage deepens, the difference between XBS and k will become larger and larger. The significance of this condition is to use the real-time values of XBS and _k to roughly pick out the working point of the quasi-linear region; The significance of this condition is to filter out the operating points in the strongly nonlinear region. At this point, the recursive estimation input analysis for the Gaussian distribution is completed, the road adhesion coefficient is estimated based on the standard normal distribution, and the maximum road adhesion coefficient is output.

Further, the acceleration and deceleration follow-up adjustment process of the vehicle in step five specifically includes:

According to the determined longitudinal acceleration command, ESC braking is used to achieve deceleration following, and the dynamic torque adjustment of the EMS is used to achieve acceleration following; before executing the command, a logical judgment needs to be made on the current state. Among them, it is first judged whether the ESC is in fault. status, if it is in a fault state, it will exit the acceleration and deceleration request, and determine whether the ESC sub-function module is executed, including but not limited to ABS, HDC, HHC, VDC, DYC and other sub-functions. If it is executed, it will exit the acceleration and deceleration request. , whether the RDU is in the fault bit. If there is a fault bit, it will exit the acceleration and deceleration request. The same is true for EMS and TCU. Finally, when the TCU determines N, it uses the RDU to distribute the front and rear axle torque according to the acceleration information, executes the expected acceleration and deceleration following instructions, and performs appropriate gains based on the current estimate of the maximum adhesion coefficient, that is, a calculation is performed for the current expected acceleration and deceleration. Coefficient weighting. When the expected acceleration is greater than or equal to zero, acceleration tracking is performed through EMS torque dynamic adjustment. When the expected acceleration is less than zero, the deceleration request is executed through ESC.

Correspondingly, the present invention also provides a vehicle dynamic performance control system that executes the method provided by the present invention. The system specifically includes:

The curve recognition module determines the working conditions based on the vehicle's driving data and identifies the vehicle's current situation of entering, exiting a curve, or driving on a non-curve;

The acceleration and deceleration calculation module is used to use driving experience data to calibrate vehicle agility control parameters corresponding to different road surfaces and different vehicle speeds, that is, the expected longitudinal acceleration;

The acceleration and deceleration recognition module is used to determine the driver's intention and decide to output the longitudinal acceleration command based on the working conditions determined by the curve recognition module, as well as the expected longitudinal acceleration, actual longitudinal acceleration, and brake pedal pressure;

The road adhesion coefficient estimation module is used to calculate the longitudinal tangent stiffness and longitudinal secant stiffness of the vehicle based on the longitudinal adhesion coefficient and longitudinal slip rate of the single wheel, and estimates the maximum road adhesion coefficient using the recursion of Gaussian distribution;

An acceleration and deceleration execution module, configured to use the estimated maximum road adhesion coefficient to provide a gain for the longitudinal acceleration command output by the acceleration and deceleration identification module, and implement dynamic torque adjustment based on the gained acceleration. , to achieve vehicle acceleration and deceleration follow-up adjustment.

In order to solve the technical problems existing in the prior art, the above-mentioned method and system provided by the present invention, for traditional power four-wheel drive vehicles, comprehensively consider the cooperative working mechanism of each vehicle subsystem and the horizontal and longitudinal coupling effects of vehicle dynamics. mechanism, and proposes a global control strategy that can improve the horizontal and vertical performance of the vehicle. It combines modern control theory and vehicle dynamics, and uses an agile dynamics control method to improve the agility, smoothness and handling stability of the vehicle in corners on different road surfaces, and improve the driver's driving experience. Compared with the prior art, the present invention at least has the following beneficial effects:

1. The present invention can improve the steering response of the vehicle under daily use conditions, and improve the safety, driving stability and cornering smoothness of the vehicle;

2. The present invention can be used for reference in the field of automatic driving vehicle control, and can be used to optimize the automatic driving vehicle control technology, and can especially be used for coupling control in curves.

Description of the drawings

Figure 1 is a schematic diagram of the curve recognition process in the present invention;

Figure 2 is a schematic diagram of the acceleration and deceleration recognition process in the present invention;

Figure 3 is a schematic diagram of the identification process of the maximum longitudinal road adhesion coefficient in the present invention;

Figure 4 is a schematic diagram of the acceleration and deceleration following process in the present invention;

Figure 5 is a logic diagram of the global control strategy of vehicle dynamic performance of the present invention;

Figure 6 is a hardware architecture diagram used in an example of the present invention.

Detailed ways

The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

The vehicle dynamic performance control method based on road adhesion coefficient identification provided by the present invention mainly includes the following steps:

Step 1: Determine the working conditions based on the vehicle's driving data and identify the vehicle's current situation of entering, exiting a curve, or driving in a non-curve;

Step 2: Use driving experience data to calibrate the vehicle agility control parameters corresponding to different road surfaces and different vehicle speeds, that is, the expected longitudinal acceleration;

Step 3: Combine the working conditions determined in Step 1, as well as the expected longitudinal acceleration, actual longitudinal acceleration, and brake pedal pressure, determine the driver's intention and decide to output the longitudinal acceleration command;

Step 4: Calculate the longitudinal tangent stiffness and longitudinal secant stiffness of the vehicle based on the longitudinal adhesion coefficient and longitudinal slip rate of the single wheel, and use the recursion of Gaussian distribution to estimate the maximum road adhesion coefficient;

Step 5: For the longitudinal acceleration command output by Step 3, use the estimated maximum road adhesion coefficient to provide a gain for the longitudinal acceleration command, and realize dynamic adjustment of torque according to the gained acceleration to achieve vehicle acceleration and deceleration following adjust.

Among them, as shown in Figure 1, in step one, the working condition is judged based on the driving data of the vehicle. Specifically, the following data are used: the lateral acceleration G _y of the vehicle, the lateral impact degree Steering wheel angle θ and steering wheel angular speed/>

As shown in Figure 2, the second step specifically uses driving experience data: first-order inertial system delay time T, system gain C _xy , and through data analysis, the understeer degree k and the measured vehicle wheelbase L are obtained. The steering wheel angle is The transmission ratio i of the front wheel angle, the steering wheel angle δ and the steering wheel angular speed For the vehicle speed V, the following expected longitudinal acceleration G _x is obtained, where s represents the Laplace factor:

As shown in Figure 3, judging the driver's intention and deciding to output the longitudinal acceleration command in step three specifically includes the following processes:

When it is judged that the working condition is entering a curve: If the vehicle's current longitudinal acceleration _a

(1). If the vehicle’s current longitudinal acceleration a _x ≤ G _x , then the decision-making longitudinal acceleration should be a _x ;

(2). If the vehicle’s current longitudinal acceleration a _x > G _x , then the decision-making longitudinal acceleration is G _x ;

(3). If the vehicle’s current longitudinal acceleration a _x >0, the decision-making longitudinal acceleration is G _x ;

When it is judged that the working condition is exiting a curve, the following situations are corresponding:

(1). If the current acceleration of the vehicle a _x > 0, the determined longitudinal acceleration is the minimum value between a _x and G _x ;

(2). If the vehicle’s current acceleration a _x =0, the determined longitudinal acceleration is G _x ;

(3). If the current acceleration of the vehicle a _x <0, the determined longitudinal acceleration is a _x .

As shown in Figure 4, the specific process of outputting the maximum road adhesion coefficient in step 4 is:

pass The relationship is estimated to obtain the longitudinal attachment coefficient μ _x of a single wheel; where, /> is the estimated value of the pure longitudinal force of a single wheel, which is estimated from the wheel edge driving moment. Its calculation formula is/> Among them, Re is the effective rolling radius of the tire; among them, F _z,i and T _t,i are the vertical force and wheel rim torque of the single wheel respectively, and the superscript ^ indicates the estimated value of the corresponding parameter; based on the tire longitudinal speed and wheel rim torque, The longitudinal slip rate S _x of the single wheel is calculated quickly; the longitudinal tangent stiffness of the vehicle is defined as/> The product of the longitudinal secant stiffness and the longitudinal slip rate of the single wheel is the adhesion coefficient μ _x =K _x S _x ; after normalization, it is obtained Among them,/> is the normalized pure longitudinal secant stiffness of a single wheel,/> For the normalized pure longitudinal tangent stiffness of a single wheel, a recursive estimation of the Gaussian distribution is performed, and the maximum road adhesion coefficient is output.

As shown in Figure 5, the acceleration and deceleration following adjustment process of the vehicle in Step 5 specifically includes:

According to the determined longitudinal acceleration command, ESC braking is used to achieve deceleration following, and the dynamic torque adjustment of the EMS is used to achieve acceleration following; before executing the command, a logical judgment needs to be made on the current state. Among them, it is first judged whether the ESC is in fault. status, if it is in a fault state, it will exit the acceleration and deceleration request, and determine whether the ESC sub-function module is executed, including but not limited to ABS, HDC, HHC, VDC, DYC and other sub-functions. If it is executed, it will exit the acceleration and deceleration request. , whether the RDU is in the fault bit. If there is a fault bit, it will exit the acceleration and deceleration request. The same is true for EMS and TCU. Finally, when the TCU determines N, it uses the RDU to distribute the front and rear axle torque according to the acceleration information, executes the expected acceleration and deceleration following instructions, and performs appropriate gains based on the current estimate of the maximum adhesion coefficient, that is, a calculation is performed for the current expected acceleration and deceleration. Coefficient weighting. When the expected acceleration is greater than or equal to zero, acceleration tracking is performed through EMS torque dynamic adjustment. When the expected acceleration is less than zero, the deceleration request is executed through ESC.

The invention provides a vehicle dynamic performance control system that executes the method provided by the invention. The system specifically includes:

The curve recognition module determines the working conditions based on the vehicle's driving data and identifies the vehicle's current situation of entering, exiting a curve, or driving on a non-curve;

The acceleration and deceleration calculation module is used to use driving experience data to calibrate vehicle agility control parameters corresponding to different road surfaces and different vehicle speeds, that is, the expected longitudinal acceleration;

The acceleration and deceleration recognition module is used to determine the driver's intention and decide to output the longitudinal acceleration command based on the working conditions determined by the curve recognition module, as well as the expected longitudinal acceleration, actual longitudinal acceleration, and brake pedal pressure;

The road adhesion coefficient estimation module is used to calculate the longitudinal tangent stiffness and longitudinal secant stiffness of the vehicle based on the longitudinal adhesion coefficient and longitudinal slip rate of the single wheel, and estimates the maximum road adhesion coefficient using the recursion of Gaussian distribution;

An acceleration and deceleration execution module, configured to use the estimated maximum road adhesion coefficient to provide a gain for the longitudinal acceleration command output by the acceleration and deceleration identification module, and implement dynamic torque adjustment based on the gained acceleration. , to achieve vehicle acceleration and deceleration follow-up adjustment.

The workflow of each of the above modules is shown in Figure 1-5. Figure 6 shows a hardware architecture diagram that can be used in an example of the present invention, which can be built using existing common modules without the need to design additional functional modules, thereby providing better practicality.

It should be understood that the sequence number of each step in the embodiment of the present invention does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention. .

Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art will understand that various changes, modifications, and substitutions can be made to these embodiments without departing from the principles and spirit of the invention. and modifications, the scope of the invention is defined by the appended claims and their equivalents.

Claims (6)

1. The vehicle dynamic performance control method based on road adhesion coefficient identification is characterized by: mainly including the following steps: Step 1: Determine the working conditions based on the vehicle's driving data and identify the vehicle's current situation of entering, exiting a curve, or driving in a non-curve; Step 2: Use driving experience data to calibrate the vehicle agility control parameters corresponding to different road surfaces and different vehicle speeds, that is, the expected longitudinal acceleration; Step 3: Combine the working conditions determined in Step 1, as well as the expected longitudinal acceleration, actual longitudinal acceleration, and brake pedal pressure, determine the driver's intention and decide to output the longitudinal acceleration command; Step 4. Pass The relationship is estimated to obtain the longitudinal attachment coefficient μ _x of a single wheel; where, /> is the estimated value of the pure longitudinal force of a single wheel, which is estimated from the wheel edge driving moment. Its calculation formula is/> Among them, Re is the effective rolling radius of the tire; among them, F _z,i and T _t,i are the vertical force and wheel rim torque of the single wheel respectively, and the superscript ^ indicates the estimated value of the corresponding parameter; based on the tire longitudinal speed and wheel rim torque, The single wheel _{longitudinal slip} rate _S > Among them,/> is the normalized pure longitudinal secant stiffness of a single wheel,/> is the normalized pure longitudinal tangent stiffness of a single wheel. The superscript ^ indicates the estimated value of the corresponding parameter. Using the recursion of the Gaussian distribution, a section of data with increasing driving force under a certain drive is selected to estimate the maximum road adhesion coefficient μ _{x , max} ; specifically, it is determined that the conditions are met/> When, related to the relationship between the normalized tangent stiffness XBS and the normalized secant stiffness k, the tire force working point changes from the linear area to the quasi-linear area, and the originally equal relationship between XBS and k will become k>XBS, and As the degree of slip deepens, the difference between XBS and k will become larger and larger; using the real-time values of XBS and k, the operating point XBS _k > 0.5 in the quasi-linear region can be roughly obtained, which indicates that the tire force working for Gaussian recursion point, its corresponding normalized tangent stiffness cannot be too small, so that the operating points in the strongly nonlinear area can be filtered out; after the recursive estimation input analysis of the Gaussian distribution is completed, the road adhesion coefficient is estimated based on the standard normal distribution. Obtain the maximum road adhesion coefficient μ _{x, max} ; Step 5: For the longitudinal acceleration command output by Step 3, use the estimated maximum road adhesion coefficient to provide a gain for the longitudinal acceleration command, and realize dynamic adjustment of torque according to the gained acceleration to achieve vehicle acceleration and deceleration following adjust. 2. The method according to claim 1, characterized in that: in the step one, the working conditions are judged based on the driving data of the vehicle, specifically using the following data: lateral acceleration G _y of the vehicle, lateral impact degree Steering wheel angle θ and steering wheel angular speed/> 3. The method according to claim 2, characterized in that: the second step specifically utilizes driving experience data: first-order inertial system delay time T, system gain C _xy , and obtains understeer degree k through data analysis and measurement. The vehicle wheelbase L, the transmission ratio i from the steering wheel angle to the front wheel angle, the steering wheel angle δ and the steering wheel angular speed For the vehicle speed V, the following expected longitudinal acceleration G _x is obtained, where s represents the Laplace factor: 4. The method according to claim 2, characterized in that: judging the driver's intention and deciding to output the longitudinal acceleration command in step three specifically includes the following process: When it is judged that the working condition is entering a curve: If the vehicle's current longitudinal acceleration _a (1). If the vehicle’s current longitudinal acceleration a _x ≤ G _x , then the decision-making longitudinal acceleration should be a _x ; (2). If the vehicle’s current longitudinal acceleration a _x > G _x , then the decision-making longitudinal acceleration is G _x ; If the vehicle's current longitudinal acceleration a _x >0, the decision-making longitudinal acceleration is G _x ; When it is judged that the working condition is exiting a curve, the following situations are corresponding: (1). If the current acceleration of the vehicle a _x > 0, the determined longitudinal acceleration is the minimum value between a _x and G _x ; (2). If the vehicle’s current acceleration a _x =0, the determined longitudinal acceleration is G _x ; (3). If the current acceleration of the vehicle a _x <0, the determined longitudinal acceleration is a _x . 5. The method according to claim 1, characterized in that: the acceleration and deceleration follow-up adjustment process of the vehicle in step five specifically includes: According to the determined longitudinal acceleration command, ESC braking is used to achieve deceleration following, and the dynamic torque adjustment of the EMS is used to achieve acceleration following; before executing the command, a logical judgment needs to be made on the current state. Among them, it is first judged whether the ESC is in fault. status, if it is in the fault state, it will exit the execution of the acceleration and deceleration request, and determine whether the ESC sub-function module is executed, including the sub-functions of ABS, HDC, HHC, VDC, and DYC. If it is executed, it will exit the execution of the acceleration and deceleration request and determine whether the RDU Whether it is in the fault position, if there is a fault position, it will exit the execution of the acceleration and deceleration request, and judge the EMS and TCU in the same way; finally, when the TCU determines that there is no fault, it will use the RDU to distribute the front and rear axle torque according to the acceleration information, and execute the expected acceleration and deceleration following Instruction, based on the current estimate of the maximum adhesion coefficient, perform gain weighting on the current expected acceleration and deceleration. When the expected acceleration is greater than or equal to zero, acceleration is followed through EMS torque dynamic adjustment. When the expected acceleration is less than zero, deceleration is performed through ESC. ask. 6. A vehicle dynamic performance control system, characterized in that: executing the method according to any one of claims 1-5, the system specifically includes: The curve recognition module determines the working conditions based on the vehicle's driving data and identifies the vehicle's current situation of entering, exiting a curve, or driving on a non-curve; The acceleration and deceleration calculation module is used to use driving experience data to calibrate vehicle agility control parameters corresponding to different road surfaces and different vehicle speeds, that is, the expected longitudinal acceleration; The acceleration and deceleration recognition module is used to determine the driver's intention and decide to output the longitudinal acceleration command based on the working conditions determined by the curve recognition module, as well as the expected longitudinal acceleration, actual longitudinal acceleration, and brake pedal pressure; The road adhesion coefficient estimation module is used to calculate the longitudinal tangent stiffness and longitudinal secant stiffness of the vehicle based on the longitudinal adhesion coefficient and longitudinal slip rate of the single wheel, and estimates the maximum road adhesion coefficient using the recursion of Gaussian distribution; An acceleration and deceleration execution module, configured to use the estimated maximum road adhesion coefficient to provide a gain for the longitudinal acceleration command output by the acceleration and deceleration identification module, and implement dynamic torque adjustment based on the gained acceleration. , to achieve vehicle acceleration and deceleration follow-up adjustment.