System and method for semi-autonomous driving of vehicles

1. A method for a semi-autonomous driving of a vehicle with a steer-by-wire system having a steering wheel mechanically decoupled from vehicle wheels, such that a steering angle of the steering wheel is controlled by a column motor independently from a wheels angle of the vehicle wheels controlled by a rack motor, comprising:

receiving from a semi-autonomous driving planning (SADP) system a target value for the wheels angle;

determining values for the steering angle and the wheels angle tracking the target value of the wheels angle subject to constraints including a constraint on a motion of the vehicle wheels, a constraint on a motion of a steering wheel, a constraint on an actuation of the steering wheel and the vehicle wheels, and a constraint on a relative motion of the steering wheel with respect to the motion of the vehicle wheels; and

generating control commands to the column motor and the rack motor according to the values for the steering angle and the wheels angle, wherein steps of the method are performed by a processor.

2. The method of claim 1 , wherein the determining comprises:

applying iteratively a nonlinear filter to the target value of the wheels angle to determine the values for the steering angle and the wheels angle, such that the constraints are satisfied at each iteration.

3. The method of claim 2 , wherein the nonlinear filter includes a first filter and a second filter, further comprising:

applying the first filter to the target value of the wheels angle to produce the value of the wheels angle;

scaling the value of the wheels angle to produce a scaled value of the wheels angle; and

applying the second filter to the scaled value of the wheels angle to produce the value of the steering angle.

4. The method of claim 2 , wherein the nonlinear filter includes a multivariable nonlinear filter, further comprising:

determining concurrently at each time step of the control the values for the steering angle and the wheels angle.

5. The method of claim 1 , wherein the determining comprises:

determining the values of the steering angle and the wheels angle from a invariant set that are closest to the target value.

6. The method of claim 1 , wherein the target value for the wheels angle is determined based on a state of a movement of the vehicle and an objective of the movement of the vehicle, such that the target value is misaligned with a state of the steering wheel.

7. The method of claim 1 , wherein the constraint on the motion of the vehicle wheels include one or combination of a constraint on the wheels angle, a constraint on a rate of change of the wheels angle and a constraint on an angular acceleration of the wheels angle,

wherein the constraint on the motion of the steering wheel includes one or a combination of a constraint on the steering angle, a constraint on a rate of change of the steering angle and a constraint on an angular acceleration of the steering wheel, and

wherein the constraint on the relative motion of the steering wheel with respect to the motion of the vehicle wheels includes one or combination of a constraint on a difference between a scaled wheels angle and the steering angle, a constraint on a difference between a scaled value of the wheels angle and the value of the steering angle, a constraint on a difference between a scaled rate of change of the wheels angle and the rate of change of the steering angle, and a constraint on a difference between a scaled angular acceleration of wheels angle and the angular acceleration of the steering angle.

8. The method of claim 1 , wherein the determining comprises:

optimizing a cost function specifying a first relationship between the value of the wheels angle and the target value of the wheels angle, and a second relationship between the value of the steering angle and the value of the wheels angle, wherein the optimizing is subject to the constraints.

9. The method of claim 8 , wherein the first relationship includes a squared difference between the value of the wheels angle and the target value of the wheels angle, multiplied by a nonnegative constant, and wherein the second relationship includes a squared difference between a scaled value of the wheels angle and the value of the steering angle.

10. The method of claim 9 , further comprising:

determining a value of the nonnegative constant to balance a rapidness of tracking the target value of the wheels angle with a rapidness of alignment of the steering angle with the wheels angle, wherein increasing the value of the nonnegative constant increases the rapidness of tracking the target value by the wheels angle and decreasing the value of the nonnegative constant increases the rapidness of alignment of the steering angle with the wheels angle.

11. The method of claim 1 , further comprising:

determining a feasible region of values of the steering angle, wheels angle, and states of the steering wheel and the vehicle wheels satisfying the constraint;

determining a invariant set as a part of the feasible region, such that after a one step evolution of dynamics of the steer-by-wire system the values of the steering angle, wheels angle, and the states of the steering wheel and the vehicle wheels are inside the invariant set; and

selecting values for the steering angle and the wheels angle from the invariant set.

12. The method of claim 11 , wherein the selecting is based on a solution of a constrained quadratic program.

13. A steer-by-wire system for controlling an operation of a vehicle, comprising:

a steering rack subsystem for controlling a wheels angle of vehicle wheels via a motion of a rack motor;

a steering column subsystem for controlling a steering angle of a steering wheel via a motion of a column motor, wherein the steering wheel is mechanically decoupled from the vehicle wheels, such that the steering angle of the steering wheel is controlled by the column motor independently from the wheels angle of the vehicle wheels controlled by the rack motor; and

a control system, comprising:

a semi-autonomous driving planning (SADP) system for determining a target value for the wheels angle; and

an electronic control unit (ECU) for determining values for the steering angle and the wheels angle tracking the target value of the wheels angle subject to constraints including a constraint on a motion of the vehicle wheels, a constraint on a motion of a steering wheel, a constraint on an actuation of the steering wheel and the vehicle wheels, and a constraint on a relative motion of the steering wheel with respect to the motion of the vehicle wheels, and for generating control commands to the column motor and the rack motor according to the values for the steering angle and the wheels angle.

14. The steer-by-wire system of claim 13 , wherein the ECU comprises:

a governor for nonlinearly filtering the target value of the wheels angle based on a state of the steering rack subsystem and of a state of the steering column subsystem to produce the values for the steering angle and the wheels angle;

a column controller for determining the control command to the column motor based on the value of the steering angle; and

a rack controller for determining the control command to the rack motor based on the value of the steering angle.

15. The steer-by-wire system of claim 14 , wherein the governor includes a first filter and a second filter, and wherein the governor applies the first filter to the target value of the wheels angle to produce the value of the wheel angle, scales the value of the wheel angle to produce a scaled value of the wheel angle; and applies the second filter to the scaled value of the wheel angle to produce the value of the steering angle.

16. The steer-by-wire system of claim 14 , wherein the governor includes a multivariable nonlinear filter, and wherein the governor determines concurrently at each time step of the control the values for the steering angle and the wheels angle.

17. The steer-by-wire system of claim 13 , wherein the ECU determines the values for the steering angle and the wheels angle from a invariant set by solving a constrained quadratic program selecting the closest values to the target value from the invariant set.