Control method and apparatus for autonomous vehicle, computer device, and storage medium

1. A control method for an autonomous vehicle, comprising:

obtaining a current steering wheel angle, a current vehicle speed and a current yaw rate of the autonomous vehicle;

correcting the current steering wheel angle based on a first correction deviation coefficient and a second correction deviation coefficient obtained in a previous cycle to generate a corrected steering wheel angle;

inputting the corrected steering wheel angle and the current vehicle speed into a preset vehicle dynamic model to obtain an estimated yaw rate;

obtaining a first yaw rate deviation value between the current yaw rate and the estimated yaw rate;

processing the first yaw rate deviation value by a preset closed-loop algorithm to obtain a first correction deviation coefficient and a second correction deviation coefficient of a current cycle; and

performing correction processing on a target steering wheel angle based on the first correction deviation coefficient and the second correction deviation coefficient of the current cycle to obtain a corrected target steering wheel angle, and controlling the autonomous vehicle to drive based on the corrected target steering wheel angle.

2. The control method of claim 1 , further comprising:

selecting a desired yaw rate based on a scenario; and

generating the target steering wheel angle based on the desired yaw rate and the current yaw rate.

3. The control method of claim 2 , wherein generating the target steering wheel angle based on the desired yaw rate and the current yaw rate comprises:

transforming the desired yaw rate by a preset reference model to generate a reference yaw rate;

obtaining a second yaw rate deviation value between the current yaw rate and the reference yaw rate;

performing correction processing on the desired yaw rate based on a preset model reference adaptive algorithm and the second yaw rate deviation value to obtain a target desired yaw rate; and

inputting the target desired yaw rate into a vehicle dynamic inverse model to obtain the target steering wheel angle.

4. The control method of claim 1 , wherein correcting the current steering wheel angle based on the first correction deviation coefficient and the second correction deviation coefficient obtained in the previous cycle to generate the corrected steering wheel angle comprises:

correcting the current steering wheel angle, by using a first formula, based on the first correction deviation coefficient and the second correction deviation coefficient obtained in the previous cycle to generate the corrected steering wheel angle, wherein the first formula is expressed by:

δ real2 =(δ real1 −bias n-1 )/slope n-1 ;

where δ real1 is the current steering wheel angle, δ real2 is the corrected steering wheel angle, slope n-1 is the first correction deviation coefficient of the previous cycle, bias n-1 is the second correction deviation coefficient of the previous cycle, and n−1 represents the previous cycle.

5. The control method of claim 1 , wherein performing the correction processing on the target steering wheel angle based on the first correction deviation coefficient and the second correction deviation coefficient of the current cycle comprises:

correcting the target steering wheel angle, by using a second formula, based on the first correction deviation coefficient and the second correction deviation coefficient of the current cycle to obtain the corrected target steering wheel angle, wherein second formula is expressed by:

δ cmd2 =(δ cmd1 +bias n )×slope n ;

where δ cmd1 is the target steering wheel angle, δ cmd2 is the corrected target steering wheel angle, slope n is the first correction deviation coefficient of the current cycle, bias n is the second correction deviation coefficient of the current cycle, and n represents the current cycle.

6. The control method of claim 1 , wherein the first correction deviation coefficient is expressed by:

slope= I slope ({dot over (φ)} real )*∫({dot over (φ)} est −{dot over (φ)} real ) dt ; and

the second correction deviation coefficient is expressed by:

bias= I bias ({dot over (φ)} real )*∫({dot over (φ)} est −{dot over (φ)} real ) dt,

where {dot over (φ)} est represents the estimated yaw rate, {dot over (φ)} real represents the current yaw rate, I slope ({dot over (φ)} real ) represents that {dot over (φ)} slope is a function of {dot over (φ)} real , and I bias ({dot over (φ)} real ) represents that bias is a function of {dot over (φ)} real .

7. A control apparatus for an autonomous vehicle, comprising:

a processor; and

a memory, configured to store executable program codes;

wherein the processor is configured to run a program corresponding to the executable program codes by reading the executable program codes stored in the memory, to:

obtain a current steering wheel angle, a current vehicle speed and a current yaw rate of a vehicle;

correct the current steering wheel angle based on a first correction deviation coefficient and a second correction deviation coefficient obtained in a previous cycle to generate a corrected steering wheel angle;

input the corrected steering wheel angle and the current vehicle speed into a preset vehicle dynamic model to obtain an estimated yaw rate;

obtain a first yaw rate deviation value between the current yaw rate and the estimated yaw rate;

process the first yaw rate deviation value by a preset closed-loop algorithm to obtain a first correction deviation coefficient and a second correction deviation coefficient of a current cycle; and

perform correction processing on a target steering wheel angle based on the first correction deviation coefficient and the second correction deviation coefficient of the current cycle to obtain a corrected target steering wheel angle, and control the autonomous vehicle to drive based on the corrected target steering wheel angle.

8. The control apparatus of claim 7 , wherein the processor is further configured to run a program corresponding to the executable program codes by reading the executable program codes stored in the memory, to:

select a desired yaw rate based on a scenario; and

generate the target steering wheel angle based on the desired yaw rate and the current yaw rate.

9. The control apparatus of claim 8 , wherein the processor is configured to run a program corresponding to the executable program codes by reading the executable program codes stored in the memory, to:

transform the desired yaw rate by a preset reference model to generate a reference yaw rate;

obtain a second yaw rate deviation value between the current yaw rate and the reference yaw rate;

perform correction processing on the desired yaw rate based on a preset model reference adaptive algorithm and the second yaw rate deviation value to obtain a target desired yaw rate; and

input the target desired yaw rate into a vehicle dynamic inverse model to obtain the target steering wheel angle.

10. The control apparatus of claim 7 , wherein the processor is configured to run a program corresponding to the executable program codes by reading the executable program codes stored in the memory, to:

correct the current steering wheel angle, by using a first formula, based on the first correction deviation coefficient and the second correction deviation coefficient obtained in the previous cycle to generate the corrected steering wheel angle, wherein the first formula is expressed by:

δ real2 =(δ real1 −bias n-1 )/slope n-1 ;

where δ real1 is the current steering wheel angle, δ real2 is the corrected steering wheel angle, slope n-1 is the first correction deviation coefficient of the previous cycle, bias n-1 is the second correction deviation coefficient of the previous cycle, and n−1 represents the previous cycle.

11. The control apparatus of claim 7 , wherein the processor is configured to run a program corresponding to the executable program codes by reading the executable program codes stored in the memory, to:

correct the target steering wheel angle, by using a second formula, based on the first correction deviation coefficient and the second correction deviation coefficient of the current cycle to obtain the corrected target steering wheel angle, wherein second formula is expressed by:

δ cmd2 (δ cmd1 +bias)×slope;

where δ cmd1 is the target steering wheel angle, δ cmd2 is the target steering wheel angle obtained after the correction processing, slope is the first correction deviation coefficient of the current cycle, bias is the second correction deviation coefficient of the current cycle, and n represents the current cycle.

12. The control apparatus of claim 7 , wherein the first correction deviation coefficient is expressed by:

slope= I slope ({dot over (φ)} real )*∫({dot over (φ)} est −{dot over (φ)} real ) dt ; and

the second correction deviation coefficient is expressed by:

bias= I bias *({dot over (φ)} real )*∫({dot over (φ)} est −{dot over (φ)} real ) dt,

where {dot over (φ)} est represents the estimated yaw rate, {dot over (φ)} real represents the current yaw rate, I slope ({dot over (φ)} real ) represents that I slope is a function of {dot over (φ)} real and I bias ({dot over (φ)} real ) represents that I bias is a function of {dot over (φ)} real .

13. A non-transitory computer readable storage medium having stored thereon a computer program that, when executed by a processor, causes the processor to implement a control method for an autonomous vehicle, the control method comprising:

obtaining a current steering wheel angle, a current vehicle speed and a current yaw rate of the autonomous vehicle;

correcting the current steering wheel angle based on a first correction deviation coefficient and a second correction deviation coefficient obtained in a previous cycle to generate a corrected steering wheel angle;

inputting the corrected steering wheel angle and the current vehicle speed into a preset vehicle dynamic model to obtain an estimated yaw rate;

obtaining a first yaw rate deviation value between the current yaw rate and the estimated yaw rate;

processing the first yaw rate deviation value by a preset closed-loop algorithm to obtain a first correction deviation coefficient and a second correction deviation coefficient of a current cycle; and

performing correction processing on a target steering wheel angle based on the first correction deviation coefficient and the second correction deviation coefficient of the current cycle to obtain a corrected target steering wheel angle, and controlling the autonomous vehicle to drive based on the corrected target steering wheel angle.

14. The non-transitory computer readable storage medium of claim 13 , wherein the control method further comprising:

selecting a desired yaw rate based on a scenario; and

generating the target steering wheel angle based on the desired yaw rate and the current yaw rate.

15. The non-transitory computer readable storage medium of claim 14 , wherein generating the target steering wheel angle based on the desired yaw rate and the current yaw rate comprises:

transforming the desired yaw rate by a preset reference model to generate a reference yaw rate;

obtaining a second yaw rate deviation value between the current yaw rate and the reference yaw rate;

performing correction processing on the desired yaw rate based on a preset model reference adaptive algorithm and the second yaw rate deviation value to obtain a target desired yaw rate; and

inputting the target desired yaw rate into a vehicle dynamic inverse model to obtain the target steering wheel angle.

16. The non-transitory computer readable storage medium of claim 13 , wherein correcting the current steering wheel angle based on the first correction deviation coefficient and the second correction deviation coefficient obtained in the previous cycle to generate the corrected steering wheel angle comprises:

correcting the current steering wheel angle, by using a first formula, based on the first correction deviation coefficient and the second correction deviation coefficient obtained in the previous cycle to generate the corrected steering wheel angle, wherein the first formula is expressed by:

δ real2 =(δ real1 −bias n-1 )/slope n-1 ;

where δ real1 is the current steering wheel angle, δ real2 is the corrected steering wheel angle, slope n-1 is the first correction deviation coefficient of the previous cycle, bias n-1 is the second correction deviation coefficient of the previous cycle, and n−1 represents the previous cycle.

17. The non-transitory computer readable storage medium of claim 13 , wherein performing the correction processing on the target steering wheel angle based on the first correction deviation coefficient and the second correction deviation coefficient of the current cycle comprises:

correcting the target steering wheel angle, by using a second formula, based on the first correction deviation coefficient and the second correction deviation coefficient of the current cycle to obtain the corrected target steering wheel angle, wherein second formula is expressed by:

δ cmd2 (δ cmd1 +bias n )×slope n ;

where δ cmd1 is the target steering wheel angle, δ cmd2 is the corrected target steering wheel angle, slope n is the first correction deviation coefficient of the current cycle, bias n is the second correction deviation coefficient of the current cycle, and n represents the current cycle.

18. The non-transitory computer readable storage medium of claim 13 , wherein the first correction deviation coefficient is expressed by:

I slope =I slope ({dot over (φ)} real )*∫({dot over (φ)} est −{dot over (φ)} real ) dt ; and

the second correction deviation coefficient is expressed by:

bias= I bias ({dot over (φ)} real )*∫({dot over (φ)} est −{dot over (φ)} real ) dt,

where {dot over (φ)} est represents the estimated yaw rate, {dot over (φ)} real represents the current yaw rate, I slope ({dot over (φ)} real ) represents that I slope is a function of {dot over (φ)} real , and I bias ({dot over (φ)} real ) represents that I bias is a function of {dot over (φ)} real .