Method and arrangement for determining safe vehicle trajectories

Patent metadata

Field	Value
Publication number	US-10037036-B2
Application number	US-201615139554-A
Country	US
Kind code	B2
Filing date	Apr 27, 2016
Priority date	May 5, 2015
Publication date	Jul 31, 2018
Grant date	Jul 31, 2018

Abstract

A method and arrangement for determining safe vehicle trajectories for a vehicle equipped with sensors for monitoring a surrounding environment, taking into account sensing limitations, as well as a vehicle including such an arrangement. The method includes detecting observable obstacles, detecting unobservable areas, adding virtual obstacles in unobservable areas, associating each observable obstacle and each virtual obstacle with one or more hypothetical events and assigning an occurrence probability to each combination of obstacle and one or more events, and determining safe vehicle trajectories based on both observable obstacles and virtual obstacles and the occurrence probability of each combination of obstacle and one or more events.

First claim

What is claimed is:

1. A method for determining safe vehicle trajectories for a vehicle equipped with sensors for monitoring a surrounding environment, taking into account sensing limitations, the method comprising:
detecting observable obstacles using one or more of the sensors;
determining, using a processor, unobservable areas;
adding, using the processor, virtual obstacles in unobservable areas;
associating, using the processor, each observable obstacle and each virtual obstacle with one or more hypotheses for objects concealed outside of an observed area in which obstacle detection is possible, referred to as hypothetical events, and assigning an occurrence probability to each combination of obstacle and one or more events; and
determining, using the processor, safe vehicle trajectories based on both observable obstacles and virtual obstacles and the occurrence probability of each combination of obstacle and one or more events such that, should a first-probability hypothetical event occur, a first margin is applied to the determined safe vehicle trajectories, and should a second-probability hypothetical event occur, a second margin is applied to the determined safe vehicle trajectories, where the first-probability is less than the second-probability and the first margin is smaller than the second margin.

2. The method according to claim 1 wherein unobservable areas are determined as at least one of:
areas outside a sensing range of the vehicle sensors;
areas for which the confidence of obtained sensor readings falls below a predetermined confidence threshold; and
areas occluded by objects in the vehicle surrounding environment.

3. The method according to claim 1 wherein virtual obstacles, the associated hypothetical events of which exceed a predetermined probability to occur, are added to the unobservable areas.

4. The method according to claim 1 wherein an environment category is determined from a set of predetermined environment categories and the one or more hypothetical events selected from a set of predetermined hypothetical events based on the determined environment category, and virtual obstacles having different associated hypothetical events are assigned different probabilities to occur depending on the determined environment category.

5. The method according to claim 4 wherein statistic data for previous occurrences of observed obstacles associated with predetermined environment categories is relied on for determining probabilities for virtual obstacles to occur depending on the determined environment category.

6. The method according claim 1 , wherein the first and second margins are either physical margins to the obstacle or margins in terms of vehicle dynamics.

7. The method according to claim 6 wherein physical margins to an obstacle are distances from a safe vehicle trajectory to the obstacle, and margins in terms of vehicle dynamics are margins in one or more of vehicle longitudinal-, lateral- and yaw-acceleration levels required for following a safe vehicle trajectory.

8. The method according to claim 7 wherein an increased physical margin to an obstacle means an increased distance from a safe vehicle trajectory to the obstacle and a decreased physical margin to an obstacle mean a reduced distance from a safe vehicle trajectory to the obstacle, and an increased margin in terms of vehicle dynamics means reduced vehicle acceleration levels required for following a safe vehicle trajectory and a decreased margin in terms of vehicle dynamics means increased vehicle acceleration levels required for following a safe vehicle trajectory.

9. A system for determining safe vehicle trajectories for a vehicle equipped with sensors for monitoring a surrounding environment, taking into account sensing limitations, the system comprising:
a processor configured to process readings from the vehicle sensors for monitoring the surrounding environment, detect observable obstacles, determine unobservable areas, add virtual obstacles in unobservable areas, associate each observable obstacle and each virtual obstacle with one or more hypotheses for objects concealed outside of an observed area in which obstacle detection is possible, referred to as hypothetical events, and assign an occurrence probability to each combination of obstacle and one or more events, and determine safe vehicle trajectories based on both observable obstacles and virtual obstacles and the occurrence probability of each combination of obstacle and one or more events such that, should a first-probability hypothetical event occur, a first margin is applied to the determined safe vehicle trajectories, and should a second-probability hypothetical event occur, a second margin is applied to the determined safe vehicle trajectories, where the first-probability is less than the second-probability and the first margin is less than the second margin.

10. The system of claim 9 wherein the processor further is configured to determine unobservable areas as at least one of:
areas outside a sensing range of the vehicle sensors;
areas for which the confidence of obtained sensor readings fall below a predetermined confidence threshold; and
areas occluded by objects in the vehicle surrounding environment.

11. The system of claim 9 wherein the processor further is configured to determine an environment category from a set of predetermined environment categories and select the one or more hypothetical events from a set of predetermined hypothetical events based on the determined environment category, and assign to virtual obstacles having different associated hypothetical events different probabilities to occur depending on the determined environment category.

12. The system of claim 11 wherein the processor further is configured to rely on statistic data for previous occurrences of observed obstacles associated with predetermined environment categories for determining probabilities for virtual obstacles to occur depending on the determined environment category.

13. A vehicle comprising:
sensors for monitoring a surrounding environment; and
a system for determining safe vehicle trajectories for the vehicle, taking into account sensing limitations, the system comprising a processor configured to process readings from the vehicle sensors for monitoring the surrounding environment, detect observable obstacles, determine unobservable areas, add virtual obstacles in unobservable areas, associate each observable obstacle and each virtual obstacle with one or more hypotheses for objects concealed outside of an observed area in which obstacle detection is possible, referred to as hypothetical events, and assign an occurrence probability to each combination of obstacle and one or more events, and determine safe vehicle trajectories based on both observable obstacles and virtual obstacles and the occurrence probability of each combination of obstacle and one or more events such that, should a first-probability hypothetical event occur, a first margin is applied to the determined safe vehicle trajectories, and should a second-probability hypothetical event occur, a second margin is applied to the determined safe vehicle trajectories, where the first-probability is less than the second-probability and the first margin is less than the second margin.

14. The method according to claim 1 wherein the vehicle is at least a semi-autonomous vehicle, and the method further comprises controlling the vehicle or outputting information to a vehicle driver based on the safe vehicle trajectories determined.

15. The system of claim 9 wherein the vehicle is at least a semi-autonomo

Assignees

• Volvo Car Corp

Inventors

• Nilsson Jonas
• Brannstrom Mattias Erik
• Ali Mohammad
• Sorstedt Joakim Lin

Classifications

• B60W2754/30

  Longitudinal distance · CPC title

• B60W2554/00

  Input parameters relating to objects · CPC title

• B60W2720/125

  Lateral acceleration · CPC title

• G01S13/931

  of land vehicles · CPC title

• B60W30/095

  Predicting travel path or likelihood of collision · CPC title