Method for simulating obstacle in unmanned simulation scene

Patent metadata

Field	Value
Publication number	US-11840259-B2
Application number	US-202017119662-A
Country	US
Kind code	B2
Filing date	Dec 11, 2020
Priority date	May 14, 2020
Publication date	Dec 12, 2023
Grant date	Dec 12, 2023

Abstract

A method for simulating an obstacle in an unmanned simulation scene includes: for obstacle information in a three-dimensional scene map, determining Gaussian distribution information of obstacle position detected by using a perception algorithm to be tested based on actual perception performance of the perception algorithm; adjusting each position of the simulated obstacle in the initial motion trajectory sequence, such that a position deviation between each adjusted position point in the target motion trajectory sequence and the corresponding position point in the initial motion trajectory sequence follows a Gaussian distribution; and adding the target motion trajectory sequence of the simulated obstacle to the three-dimensional scene map.

First claim

What is claimed is:

1. A method for simulating an obstacle in an unmanned simulation scene, comprising:
obtaining obstacle information to be added to a three-dimensional scene map, the obstacle information comprising an initial motion trajectory sequence of a simulated obstacle;
determining Gaussian distribution information of obstacle position detected by using a perception algorithm to be tested based on actual perception performance of the perception algorithm, the Gaussian distribution information comprising a preset expectation value and a preset variance value of a Gaussian distribution;
adjusting the initial motion trajectory sequence based on the Gaussian distribution to obtain a target motion trajectory sequence of the simulated obstacle, for each position point in the target motion trajectory sequence, a position deviation between the position point in the target motion trajectory sequence and a corresponding position point in the initial motion trajectory sequence following the Gaussian distribution; and
adding the target motion trajectory sequence of the simulated obstacle to the three-dimensional scene map;
wherein the method further comprises:
determining an effective decision region of an unmanned vehicle model to be tested based on a velocity of the unmanned vehicle model, wherein the effective decision region refers to a region that has an impact on a driving plan of the unmanned vehicle;
obtaining an exponential distribution model corresponding to a type of the simulated obstacle, and determining residence period information of the type of simulated obstacle in the effective decision region based on the exponential distribution model; and
adding the residence period information of the type of simulated obstacle to the three-dimensional scene map.

2. The method of claim 1 , wherein the obstacle information further comprises an initial contour of the simulated obstacle at each position point in the initial motion trajectory sequence, the simulated obstacle has a same initial contour at each position point in the initial motion trajectory sequence, and the method further comprises:
adjusting the initial contour of the simulated obstacle at each position point in the initial motion trajectory sequence based on the Gaussian distribution to obtain a target contour of the simulated obstacle at each position point in the target motion trajectory sequence, for each position point in the target motion trajectory sequence, a contour deviation between a target contour of the simulated obstacle at the position point in the target motion trajectory sequence and an initial contour of the simulated obstacle at a corresponding position point in the initial motion trajectory sequence following the Gaussian distribution; and
adding the target contour of the simulated obstacle at each position point in the target motion trajectory sequence to the three-dimensional scene map.

3. The method of claim 1 , wherein the obstacle information further comprises a type and an initial position point of the simulated obstacle, before obtaining the obstacle information to be added to the three-dimensional scene map, the method further comprises:
obtaining a position range specified in the three-dimensional scene map;
determining the type and the initial position point of the simulated obstacle distributed within the position range by utilizing a preset obstacle distribution model;
adding the type and the initial position point of the simulated obstacle to the three-dimensional scene map.

4. The method of claim 2 , wherein the obstacle information further comprises a type and an initial position point of the simulated obstacle, before obtaining the obstacle information to be added to the three-dimensional scene map, the method further comprises:
obtaining a position range specified in the three-dimensional scene map;
determining the type and the initial position point of the simulated obstacle distributed within the position range by utilizing a preset obstacle distribution model;
adding the type and the initial position point of the simulated obstacle to the three-dimensional scene map.

5. The method of claim 3 , further comprising:
obtaining a behavior model of the simulated obstacle, the behavior model being established by analyzing behavior data in actual driving scene data made by an actual obstacle corresponding to the simulated obstacle when the actual obstacle encounters a n unmanned vehicle; and
determining the initial motion trajectory sequence of the simulated obstacle based on the initial position point and the behavior model.

6. An electronic device, comprising:
at least one processor; and
a memory, communicatively coupled to the at least one processor,
wherein the memory is configured to store instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is caused to implement a method for simulating an obstacle in an unmanned simulation scene, the method comprising:
obtaining obstacle information to be added to a three-dimensional scene map, the obstacle information comprising an initial motion trajectory sequence of a simulated obstacle;
determining Gaussian distribution information of obstacle position detected by using a perception algorithm to be tested based on actual perception performance of the perception algorithm, the Gaussian distribution information comprising a preset expectation value and a preset variance value of a Gaussian distribution;
adjusting the initial motion trajectory sequence based on the Gaussian distribution to obtain a target motion trajectory sequence of the simulated obstacle, for each position point in the target motion trajectory sequence, a position deviation between the position point in the target motion trajectory sequence and a corresponding position point in the initial motion trajectory sequence following the Gaussian distribution; and
adding the target motion trajectory sequence of the simulated obstacle to the three-dimensional scene map;
wherein the method further comprises:
determining an effective decision region of an unmanned vehicle model to be tested based on a velocity of the unmanned vehicle model, wherein the effective decision region refers to a region that has an impact on a driving plan of the unmanned vehicle;
obtaining an exponential distribution model corresponding to a type of the simulated obstacle, and determining residence period information of the type of simulated obstacle in the effective decision region based on the exponential distribution model; and
adding the residence period information of the type of simulated obstacle to the three-dimensional scene map.

7. The electronic device of claim 6 , wherein the obstacle information further comprises an initial contour of the simulated obstacle at each position point in the initial motion trajectory sequence, the simulated obstacle has a same initial contour at each position point in the initial motion trajectory sequence, and the method further comprises:
adjusting the initial contour of the simulated obstacle at each position point in the initial motion trajectory sequence based on the Gaussian distribution to obtain a target contour of the simulated obstacle at each position point in the target motion trajectory sequence, for each position point in the target motion trajectory sequence, a contour deviation between a target contour of the simulated obstacle at the position point in the target motion trajectory sequence and an initial contour of the simulated obstacle at a corresponding position point in the initial motion trajectory sequence following the Gaussian distribution; and
adding the target contour of the simulated obstacle at each positi

Assignees

• Apollo Intelligent Driving Tech Beijing Co Ltd

Inventors

• Yu Ning
• Yang Fan
• Mao Jiming

Classifications

• B60W60/0027 Primary

  using trajectory prediction for other traffic participants · CPC title

• B60W40/02

  related to ambient conditions · CPC title

• G06T19/20

  Editing of three-dimensional [3D] images, e.g. changing shapes or colours, aligning objects or positioning parts · CPC title

• B60W2554/402

  Type · CPC title

• B60W2554/4041

  Position · CPC title