Autonomous vehicle driving systems and methods for critical conditions

1. A method for autonomously operating a vehicle, the method comprising:

generating, by a processor, a search graph based on a reference trajectory, wherein the search graph includes a plurality of nodes that are offset in a lateral direction from centerline waypoint stations, and line segments that connect each node of the plurality of nodes to neighboring nodes;

generating, by the processor, a first path for the vehicle based on a cost function and the search graph, wherein the cost function is associated with possible collisions with static and dynamic objects on a road, a degree of offset from the reference trajectory, a required steering action, a degree of deviation from a previously generated trajectory;

generating, by the processor, a first trajectory based on the first path and a corridor, the corridor defines points associated with the first path and that represent safe distances from obstacles;

identifying, by the processor, at least one constraint;

modifying, by the processor, at least a first portion of the first trajectory using model predictive control that takes into account the at least one constraint to result in a second trajectory, wherein the portion is defined based on a pre-defined duration;

generating a resulting trajectory based on the first trajectory and the second trajectory; and

controlling the vehicle based on the resulting trajectory.

2. The method of claim 1 wherein the generating the resulting trajectory includes combining the first trajectory and the second trajectory, by smoothing an ending of the second trajectory with a beginning of the first trajectory to form the resulting trajectory.

3. The method of claim 1 , wherein the at least one constraint is a safe zone constraint defined around the first portion of the first trajectory as an acceptable deviation.

4. The method of claim 1 , wherein the at least one constraint is a stability constraint representing lateral motion of the vehicle in terms of side slope angle (β) and yaw rate (r={dot over (φ)}).

5. The method of claim 1 , wherein the at least one constraint is an actuator constraint representing a limitation associated with at least one of a steering system, a propulsion system, or a braking system of the vehicle.

6. The method of claim 1 , wherein the at least one constraint is based on a vehicle dynamics model.

7. The method of claim 1 ,

and wherein the search graph uses a coordinate system that indicates lateral offsets of objects relative to the reference trajectory.

8. The method of claim 7 , wherein the objects include other vehicles in the vehicle environment.

9. The method of claim 8 , wherein the search graph is one of a plurality of search graphs generated using the coordinate system to determine trajectory paths in different road scenarios.

10. An autonomous vehicle, comprising:

at least one sensor that provides sensor data about objects within a vehicle environment as vehicle object environment data and about a vehicle state as vehicle state data; and

a controller that, by a processor and based on the sensor data, is configured to:

generate, a search graph based on a reference trajectory, wherein the search graph includes a plurality of nodes that are offset in a lateral direction from centerline waypoint stations, and line segments that connect each node of the plurality of nodes to neighboring nodes;

generate a first path for the vehicle based on a cost function and the search graph, wherein the cost function is associated with possible collisions with static and dynamic objects on a road, a degree of offset from the reference trajectory, a required steering action, a degree of deviation from a previously generated trajectory;

generate a first trajectory based on the first path and a corridor, the corridor defines points associated with the first path and that represent safe distances from obstacles;

identify at least one critical condition constraint based on at least one of the vehicle or the vehicle environment;

modify at least a first portion of the first trajectory using model predictive control that takes into account the at least one critical condition constraint to result in a second trajectory, wherein the first portion is defined based on a pre-defined duration;

generate a resulting trajectory based on the first trajectory and the second trajectory; and

control the vehicle based on the resulting trajectory.

11. The autonomous vehicle of claim 10 , wherein the controller is configured to generate the first trajectory without consideration of the at least one critical condition constraint.

12. The autonomous vehicle of claim 11 , wherein the controller is configured to smooth a first end of the second trajectory to provide continuity with a second end of the first trajectory.

13. The autonomous vehicle of claim 12 ,

wherein the at least one critical condition constraint includes

a safe zone constraint defined around the first portion of the first trajectory as an acceptable deviation,

a stability constraint representing lateral motion of the vehicle in terms of side slope angle (β) and yaw rate (r={dot over (φ)}), or an actuator constraint representing a limitation associated with at least one of a steering system, a propulsion system, or a braking system of the vehicle.

14. The method of claim 1 , wherein the model predictive control includes a linear parameter varying system.