# US 11,305,765 B2 · Method for predicting movement of moving objects relative to an autonomous driving vehicle

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

- Number: US 11,305,765 B2
- Document ID: US11305765B2
- Status: Granted
- Kind: B2
- Application number: 16/392,354
- Filed: 2019-04-23
- Published: 2020-10-29
- Granted: 2022-04-19
- Earliest claimed: 2019-04-23
- Assignees: Baidu USA LLC
- Applicants: Baidu USA LLC
- Inventors: Kecheng Xu; Yajia Zhang; Hongyi Sun; Jiacheng Pan; Jinghao Miao
- Web page: https://patentgrep.com/patent/US11305765B2

## CPC

- B60W 30/0956: {the prediction being responsive to traffic or environmental parameters}
- B60W 30/18145: {Cornering}
- B60W 30/09: Taking automatic action to avoid collision, e.g. braking and steering
- B60W 30/18159: {Traversing an intersection}
- B60W 60/0011: {involving control alternatives for a single driving scenario, e.g. planning several paths to avoid obstacles}
- B60W 60/00274: {considering possible movement changes}
- G06N 20/00: Machine learning
- B60W 2554/4045: Intention, e.g. lane change or imminent movement
- B60W 2720/125: Lateral acceleration
- 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
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

In response to perceiving a moving object, one or more possible object paths of the moving object are determined based on the prior movement predictions of the moving object, for example, using a machine-learning model, which may be created based on a large amount of driving statistics of different vehicles. For each of the possible object paths, a set of trajectory candidates is generated based on a set of predetermined accelerations. Each of the trajectory candidates corresponds to one of the predetermined accelerations. A trajectory cost is calculated for each of the trajectory candidates using a predetermined cost function. One of the trajectory candidates having the lowest trajectory cost amongst the trajectory candidates is selected. An ADV path is planned to navigate the ADV to avoid collision with the moving object based on the lowest costs of the possible object paths of the moving object.

## Claims

### Claim 1

A computer-implemented method for operating an autonomous driving vehicle (ADV), the method comprising:
in response to perceiving a moving object, determining one or more possible object paths based on prior movement predictions of the moving object;
for each of the possible object paths,
generating a set of trajectory candidates based on a set of predetermined accelerations,
calculating a trajectory cost for each of the trajectory candidates using a predetermined cost function, and
selecting one of the trajectory candidates with a lowest trajectory cost amongst the trajectory candidates to represent the possible object path; and
planning an ADV path to navigate the ADV to avoid collision with the moving object based on the lowest trajectory costs associated with the possible object paths of the moving object.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
for each of the possible object paths, calculating a moving probability based on the trajectory cost of the selected trajectory candidate; and
selecting one of the possible object paths having a highest moving probability, wherein the ADV path is planned based on the selected possible object path of the moving object.

### Claim 3 (depends on 2)

The method of claim 2, wherein calculating a moving probability of a possible object path comprises:
determining a likelihood probability based on the lowest trajectory cost associated with the possible object path, wherein the likelihood probability represents a likelihood that the moving object will move according to the possible object path; and
calculating the moving probability of the possible object path based on the likelihood probability and a prior probability of the possible object path.

### Claim 4 (depends on 1)

The method of claim 1, wherein each of the trajectory candidates is associated with one of the predetermined accelerations with which the moving object moves along the trajectory candidate.

### Claim 5 (depends on 1)

The method of claim 1, wherein calculating a cost for each of the trajectory candidates comprises:
calculating a centripetal acceleration cost for the trajectory candidate using a first predetermined cost function;
calculating a collision cost for the trajectory candidate using a second predetermined cost function; and
calculating the trajectory cost of the trajectory candidate based on the centripetal acceleration cost and the collision cost.

### Claim 6 (depends on 5)

The method of claim 5, wherein calculating a centripetal acceleration cost for the trajectory candidate comprises:
determining a plurality of trajectory points along the trajectory candidate;
for each of the trajectory points, calculating a centripetal acceleration for the trajectory point; and
calculating the centripetal acceleration cost based on the centripetal accelerations of the trajectory points.

### Claim 7 (depends on 6)

The method of claim 6, wherein the centripetal acceleration of each trajectory point is determined based on a speed of the moving object and a curvature at the trajectory point of the trajectory candidate.

### Claim 8 (depends on 6)

The method of claim 6, wherein the trajectory points are selected evenly in time along the trajectory candidate.

### Claim 9 (depends on 5)

The method of claim 5, wherein calculating a collision cost for the trajectory candidate comprises:
determining a plurality of trajectory points along the trajectory candidate;
for each of the trajectory points, determining a relative distance between the trajectory point and the ADV at a point in time corresponding to the trajectory point; and
calculating the collision cost based on the relative distances between the trajectory points and the ADV.

### Claim 10

A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor to perform operations, the operations comprising:
in response to perceiving a moving object, determining one or more possible object paths based on prior movement predictions of the moving object;
for each of the possible object paths,
generating a set of trajectory candidates based on a set of predetermined accelerations,
calculating a trajectory cost for each of the trajectory candidates using a predetermined cost function, and
selecting one of the trajectory candidates with a lowest trajectory cost amongst the trajectory candidates to represent the possible object path; and
planning an ADV path to navigate the ADV to avoid collision with the moving object based on the lowest trajectory costs associated with the possible object paths of the moving object.

### Claim 11 (depends on 10)

The machine-readable medium of claim 10, wherein the operations further comprise:
for each of the possible object paths, calculating a moving probability based on the trajectory cost of the selected trajectory candidate; and
selecting one of the possible object paths having a highest moving probability, wherein the ADV path is planned based on the selected possible object path of the moving object.

### Claim 12 (depends on 11)

The machine-readable medium of claim 11, wherein calculating a moving probability of a possible object path comprises:
determining a likelihood probability based on the lowest trajectory cost associated with the possible object path, wherein the likelihood probability represents a likelihood that the moving object will move according to the possible object path; and
calculating the moving probability of the possible object path based on the likelihood probability and a prior probability of the possible object path.

### Claim 13 (depends on 10)

The machine-readable medium of claim 10, wherein each of the trajectory candidates is associated with one of the predetermined accelerations with which the moving object moves along the trajectory candidate.

### Claim 14 (depends on 10)

The machine-readable medium of claim 10, wherein calculating a cost for each of the trajectory candidates comprises:
calculating a centripetal acceleration cost for the trajectory candidate using a first predetermined cost function;
calculating a collision cost for the trajectory candidate using a second predetermined cost function; and
calculating the trajectory cost of the trajectory candidate based on the centripetal acceleration cost and the collision cost.

### Claim 15 (depends on 14)

The machine-readable medium of claim 14, wherein calculating a centripetal acceleration cost for the trajectory candidate comprises:
determining a plurality of trajectory points along the trajectory candidate;
for each of the trajectory points, calculating a centripetal acceleration for the trajectory point; and
calculating the centripetal acceleration cost based on the centripetal accelerations of the trajectory points.

### Claim 16 (depends on 15)

The machine-readable medium of claim 15, wherein the centripetal acceleration of each trajectory point is determined based on a speed of the moving object and a curvature at the trajectory point of the trajectory candidate.

### Claim 17 (depends on 15)

The machine-readable medium of claim 15, wherein the trajectory points are selected evenly in time along the trajectory candidate.

### Claim 18 (depends on 14)

The machine-readable medium of claim 14, wherein calculating a collision cost for the trajectory candidate comprises:
determining a plurality of trajectory points along the trajectory candidate;
for each of the trajectory points, determining a relative distance between the trajectory point and the ADV at a point in time corresponding to the trajectory point; and
calculating the collision cost based on the relative distances between the trajectory points and the ADV.

### Claim 19

A data processing system, comprising:
a processor; and
a memory coupled to the processor to store instructions, which when executed by the processor, cause the processor to perform operations, the operations including
in response to perceiving a moving object, determining one or more possible object paths based on prior movement predictions of the moving object,
for each of the possible object paths,
generating a set of trajectory candidates based on a set of predetermined accelerations,
calculating a trajectory cost for each of the trajectory candidates using a predetermined cost function, and
selecting one of the trajectory candidates with a lowest trajectory cost amongst the trajectory candidates to represent the possible object path, and
planning an ADV path to navigate the ADV to avoid collision with the moving object based on the lowest trajectory costs associated with the possible object paths of the moving object.

### Claim 20 (depends on 19)

The system of claim 19, wherein the operations further comprise:
for each of the possible object paths, calculating a moving probability based on the trajectory cost of the selected trajectory candidate; and
selecting one of the possible object paths having a highest moving probability, wherein the ADV path is planned based on the selected possible object path of the moving object.