Autonomous vehicle interactive decision making

Patent metadata

Field	Value
Publication number	US-11242054-B2
Application number	US-201916439682-A
Country	US
Kind code	B2
Filing date	Jun 12, 2019
Priority date	Jun 12, 2019
Publication date	Feb 8, 2022
Grant date	Feb 8, 2022

Abstract

Autonomous vehicle interactive decision making may include identifying two or more traffic participants and gaps between the traffic participants, selecting a gap and identifying a traffic participant based on a coarse probability of a successful merge between the autonomous vehicle and a corresponding traffic participant, generating an intention prediction associated with the identified traffic participant based on vehicle dynamics of the identified traffic participant, predicted behavior of the identified traffic participant in the absence of the autonomous vehicle, and predicted behavior of the identified traffic participant in the presence of the autonomous vehicle making a maneuver creating an interaction between the identified traffic participant and the autonomous vehicle, generating an intention prediction associated with the autonomous vehicle, calculating an updated probability of a successful interaction between the identified traffic participant and the autonomous vehicle based on the intention prediction associated with the identified traffic participant and the autonomous vehicle.

First claim

The invention claimed is:

1. A system for autonomous vehicle interactive decision making, comprising:
a sensor identifying three or more traffic participants within a same lane and two or more gaps between the three or more traffic participants;
an action predictor, implemented via a processor, selecting a gap from two or more of the gaps via a first coarse tree search and identifying a traffic participant from the three or more traffic participants associated with the selected gap based on calculating a coarse probability of a successful merge for the autonomous vehicle at each one of the two or more gaps;
an intention predictor, implemented via the processor, generating an intention prediction, via a second tree search, associated with the identified traffic participant based on one or more vehicle dynamics of the identified traffic participant, predicted behavior of the identified traffic participant in the absence of the autonomous vehicle, and predicted behavior of the identified traffic participant in the presence of the autonomous vehicle making a maneuver creating an interaction between the identified traffic participant and the autonomous vehicle;
wherein the intention predictor generates an intention prediction associated with the autonomous vehicle;
a model updater, implemented via the processor, calculating an updated probability of a successful merge between the identified traffic participant and the autonomous vehicle based on the intention prediction associated with the identified traffic participant and the intention prediction associated with the autonomous vehicle only for the vehicles associated with the selected gap; and
an autonomous action selector, implemented via the processor, implementing the maneuver based on the updated probability of the successful interaction between the identified traffic participant and the autonomous vehicle,
wherein the action predictor calculates the coarse probability of the successful merge between the autonomous vehicle and the corresponding traffic participant of the three or more traffic participants for each of the two or more gaps, and
wherein the coarse probability of the successful merge is calculated based on an estimated willingness to yield of the corresponding traffic participant associated with each of the two or more gaps, a size of a corresponding gap, and a distance between the corresponding gap and a current position of the autonomous vehicle.

2. The system for autonomous vehicle interactive decision making of claim 1 , wherein the intention predictor generates the intention prediction associated with the identified traffic participant based on continuity associated with the identified traffic participant, wherein continuity is a likelihood of the identified traffic participant continuing a current course of action.

3. The system for autonomous vehicle interactive decision making of claim 1 , wherein the intention predictor generates the intention prediction associated with the identified traffic participant based on influencibility associated with the identified traffic participant, wherein influencibility is a likelihood of the identified traffic participant changing a current course of action based on a state of the autonomous vehicle.

4. The system for autonomous vehicle interactive decision making of claim 1 , wherein the coarse probability of the successful merge m is represented by:

P
⁡

(

m
|
y

,
d
,
g

)

∝

P
⁡

(

m
|
y

)

⁢

P
⁡

(

m
|
d

)

⁢

P
⁡

(

m
|
g

)

P
⁡

(
m
)

,

wherein P(m|y) is a personality model which governs an estimated willingness to yield y of the corresponding traffic participant associated with each of the two or more gaps, a size of a corresponding gap g, and a distance d between the corresponding gap and a current position of the autonomous vehicle, wherein P(m) is a non-zero value.

5. The system for autonomous vehicle interactive decision making of claim 4 , wherein

P
⁡

(

m
|
d

)

∝

exp
⁡

(

-

(

d
-

d
0

)

2

σ
2

)

,

wherein σ 2 is a non-zero value and σ is indicative of a standard deviation of measurement noise.

6. A method for autonomous vehicle interactive decision making, comprising:
identifying three or more traffic participants within a same lane and two or more gaps between the three or more traffic participants;
selecting a gap from two or more of the gaps via a first coarse tree search and identifying a traffic participant from the three or more traffic participants associated with the selected gap based on calculating a coarse probability of a successful merge for the autonomous vehicle at each one of the two or more gaps;
generating an intention prediction, via a second tree search, associated with the identif

Assignees

• Honda Motor Co Ltd

Inventors

• Isele David Francis

Classifications

• G08G1/167

  Driving aids for lane monitoring, lane changing, e.g. blind spot detection · CPC title

• B60W30/0956 Primary

  the prediction being responsive to traffic or environmental parameters · CPC title

• B60W30/18163

  Lane change; Overtaking manoeuvres · CPC title

• B60W60/00276

  for two or more other traffic participants · CPC title

• B60W30/16 Primary

  Control of distance between vehicles, e.g. keeping a distance to preceding vehicle · CPC title