# US 11,999,380 B1 · Autonomous vehicle trajectory generation and optimization

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 11,999,380 B1
- Document ID: US11999380B1
- Status: Granted
- Kind: B1
- Application number: 17/554,693
- Filed: 2021-12-17
- Granted: 2024-06-04
- Earliest claimed: 2021-12-17
- Assignees: Zoox, Inc.
- Applicants: Zoox, Inc.
- Inventors: Joseph Funke; Steven Cheng Qian; Kazuhide Okamoto; Jacob Patrick Thalman; Sriram Narayanan
- Web page: https://patentgrep.com/patent/US11999380B1

## CPC

- B60W 60/0011: {involving control alternatives for a single driving scenario, e.g. planning several paths to avoid obstacles}
- B60W 2520/06: Direction of travel
- B60W 2520/10: Longitudinal speed
- B60W 2520/12: Lateral speed
- B60W: CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT

## Abstract

Techniques are discussed for generating and optimizing trajectories for controlling autonomous vehicles in performing on-route and off-route actions within a driving environment. A planning component of an autonomous vehicle can receive or generate time-discretized (or temporal) trajectories for the autonomous vehicle to traverse an environment. Trajectories can be optimized, for example, based on the lateral and longitudinal dynamics of the vehicle, using loss functions and/or costs. In some examples, the temporal optimization of a trajectory may include resampling a previous trajectory based on the differences in the time sequences of the temporal trajectories, to ensure temporal consistency of trajectories across planning cycles. Constraints also may be applied during temporal optimization in some examples, to control or restrict driving maneuvers that are not supported by the autonomous vehicle.

## Claims

### Claim 1

A system comprising:
one or more processors; and
one or more non-transitory computer-readable media storing computer-executable instructions that, when executed, cause the system to perform operations comprising:
determining a first trajectory for a vehicle, wherein the first trajectory includes a first set of time-discretized vehicle states associated with a first set of times;
controlling the vehicle for a first duration of time based at least in part on the first trajectory;
receiving an updated trajectory defined by a second set of time-discretized vehicle states associated with a second set of times, the second set of times differing from the first set of times;
determining, based at least in part on the first trajectory and the second set of times, a resampled trajectory including a third set of time-discretized vehicle states associated with the second set of times, wherein determining the resampled trajectory comprises applying an offset value to the first set of time-discretized vehicle states, wherein each of the third set of vehicle states is associated with a fixed time and is independent of a fixed physical location, and wherein the resampled trajectory comprises a trajectory for moving the vehicle from an off-route current vehicle state to a target vehicle state located on a driving lane of a route-based driving environment;
determining a consistency metric between the updated trajectory and the resampled trajectory, based at least in part on a vehicle state difference determined based on the second and third sets of time-discretized vehicle states;
determining a control trajectory for the vehicle, based at least in part on the consistency metric; and
controlling the vehicle based at least in part on the control trajectory.

### Claim 2 (depends on 1)

The system as recited in claim 1, wherein the first trajectory includes a first vehicle control command associated with a first time, and wherein determining the resampled trajectory comprises:
instantiating a vehicle model based on the vehicle;
applying a first vehicle state corresponding to the first time to the vehicle model; and
executing the first vehicle control command on the vehicle model for a duration of time based on a time difference between the first set of time-discretized vehicle states and the second set of time-discretized vehicle states.

### Claim 3 (depends on 1)

The system as recited in claim 1, wherein the first set of time-discretized vehicle states is associated with a first time step duration, and wherein determining the resampled trajectory comprises:
determining a velocity of the vehicle; and
determining, based at least in part on the velocity, a second time step duration associated with the second set of time-discretized vehicle states.

### Claim 4

A method comprising:
determining a first trajectory for a vehicle, wherein the first trajectory includes a first set of vehicle states spaced temporally from one another;
resampling the first trajectory into a second set of vehicle states spaced temporally from one another and differing in time from the first set of vehicle states, wherein resampling the first trajectory comprises determining the second set of vehicle states by applying an offset value to the first set of vehicle states, wherein each of the second set of vehicle states is associated with a fixed time and is independent of a fixed physical location;
determining an optimal trajectory for moving the vehicle from an off-route current vehicle state to a target vehicle state located on a driving lane of a route-based driving environment, the optimal trajectory comprising a third set of vehicle states spaced temporally from one another, wherein the third set of vehicle states correspond in time to the second set of vehicle states;
determining a consistency metric between the first trajectory and the optimal trajectory, based at least in part on a vehicle state difference determined based on the second and thirds set of vehicle states;
determining a control trajectory for the vehicle, based at least in part on the consistency metric; and
controlling the vehicle based at least in part on the control trajectory.

### Claim 5 (depends on 4)

The method of claim 4, wherein the first trajectory includes a first vehicle state comprising a first vehicle control command associated with a first time, and wherein resampling the first trajectory comprises:
instantiating a vehicle model based on the vehicle;
applying the first vehicle state to the vehicle model; and
executing the first vehicle control command on the vehicle model for a duration of time based on a time difference between the first set of vehicle states and the second set of vehicle states.

### Claim 6 (depends on 5)

The method of claim 5, wherein determining the consistency metric comprises:
determining a second vehicle state based on the resampling of the first trajectory;
determining a third vehicle state based on the optimal trajectory; and
comparing a difference between the second vehicle state and the third vehicle state to a threshold.

### Claim 7 (depends on 4)

The method of claim 4, wherein:
the first set of vehicle states is associated with a first time sequence; and
the second set of vehicle states is associated with a second time sequence, wherein the second time sequence is offset from the first time sequence.

### Claim 8 (depends on 7)

The method of claim 7, wherein the first time sequence includes a first time step duration, and wherein resampling the first trajectory comprises:
determining a velocity of the vehicle; and
determining a second time step duration for the second time sequence, based at least in part on the velocity of the vehicle, wherein the second time step duration is different from the first time step duration.

### Claim 9 (depends on 4)

The method of claim 4, wherein resampling the first trajectory comprises:
determining a vehicle state within the second set of vehicle states, wherein determining the vehicle state includes determining a steering angle associated with the vehicle state based at least in part on a velocity-based steering rate activation function.

### Claim 10 (depends on 4)

The method of claim 4, wherein resampling the first trajectory comprises:
determining a vehicle state within the second set of vehicle states, the vehicle state including a velocity and a steering angle; and
determining, based at least in part on the velocity and the steering angle, that the second set of vehicle states is compliant with a velocity-based steering angle restriction.

### Claim 11 (depends on 10)

The method of claim 10, wherein the velocity is zero, and wherein resampling the first trajectory comprises:
determining a spatial projection from the vehicle state, based at least in part on the steering angle; and
evaluating the spatial projection from the vehicle state.

### Claim 12

One or more non-transitory computer-readable media storing processor-executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
determining a first trajectory for a vehicle, wherein the first trajectory includes a first set of vehicle states spaced temporally from one another;
resampling the first trajectory into a second set of vehicle states spaced temporally from one another and differing in time from the first set of vehicle states, wherein resampling the first trajectory comprises determining the second set of vehicle states by applying an offset value to the first set of vehicle states, wherein each of the second set of vehicle states is associated with a fixed time and is independent of a fixed physical location;
determining an optimal trajectory for moving the vehicle from an off-route current vehicle state to a target vehicle state located on a driving lane of a route-based driving environment, the optimal trajectory comprising a third set of vehicle states spaced temporally from one another, wherein the third set of vehicle states correspond in time to the second set of vehicle states;
determining a consistency metric between the first trajectory and the optimal trajectory, based at least in part on a vehicle state difference determined based on the second and thirds set of vehicle states;
determining a control trajectory for the vehicle, based at least in part on the consistency metric; and
controlling the vehicle based at least in part on the control trajectory.

### Claim 13 (depends on 12)

The one or more non-transitory computer-readable media of claim 12, wherein the first trajectory includes a first vehicle state comprising a first vehicle control command associated with a first time, and wherein resampling the first trajectory comprises:
instantiating a vehicle model based on the vehicle;
applying the first vehicle state to the vehicle model; and
executing the first vehicle control command on the vehicle model for a duration of time based on a time difference between the first set of vehicle states and the second set of vehicle states.

### Claim 14 (depends on 13)

The one or more non-transitory computer-readable media of claim 13, wherein determining the consistency metric comprises:
determining a second vehicle state based on the resampling of the first trajectory;
determining a third vehicle state based on the optimal trajectory; and
comparing a difference between the second vehicle state and the third vehicle state to a threshold.

### Claim 15 (depends on 12)

The one or more non-transitory computer-readable media of claim 12, wherein:
the first set of vehicle states is associated with a first time sequence; and
the second set of vehicle states is associated with a second time sequence, wherein the second time sequence is offset from the first time sequence.

### Claim 16 (depends on 15)

The one or more non-transitory computer-readable media of claim 15, wherein the first time sequence includes a first time step duration, and wherein resampling the first trajectory comprises:
determining a velocity of the vehicle; and
determining a second time step duration for the second time sequence, based at least in part on the velocity of the vehicle, wherein the second time step duration is different from the first time step duration.

### Claim 17 (depends on 12)

The one or more non-transitory computer-readable media of claim 12, wherein resampling the first trajectory comprises:
determining a vehicle state within the second set of vehicle states, wherein determining the vehicle state includes determining a steering angle associated with the vehicle state based at least in part on a velocity-based steering rate activation function.

### Claim 18 (depends on 12)

The one or more non-transitory computer-readable media of claim 12, wherein resampling the first trajectory comprises:
determining a vehicle state within the second set of vehicle states, the vehicle state including a velocity and a steering angle; and
determining, based at least in part on the velocity and the steering angle, that the second set of vehicle states is compliant with a velocity-based steering angle restriction.
