CN114199251B

CN114199251B - Anti-collision positioning method for robot

Info

Publication number: CN114199251B
Application number: CN202111466251.5A
Authority: CN
Inventors: 姜翠环; 骆敏舟; 卢钰
Original assignee: Institute of Intelligent Manufacturing Technology JITRI
Current assignee: Institute of Intelligent Manufacturing Technology JITRI
Priority date: 2021-12-03
Filing date: 2021-12-03
Publication date: 2023-09-15
Anticipated expiration: 2041-12-03
Also published as: CN114199251A

Abstract

The invention discloses an anti-collision positioning method of a robot, which comprises the following steps: acquiring the pose of the robot relative to a world coordinate system based on an odometer; randomly placing n particles carrying pose near the odometer; when moving a distance, selecting particles with highest matching degree with the current scene map to update the particles into the pose of the robot; and when the pose of the robot is updated, synchronously updating a forbidden area of the current scene map, wherein the forbidden area refers to the inside of an obstacle area which is inaccessible to the robot. The method can improve the positioning precision, improve the walking capacity of the robot at the edge of the obstacle, and avoid colliding with the obstacle to damage property.

Description

Anti-collision positioning method for robot

Technical Field

The invention belongs to the technical field of robot positioning, and particularly relates to an anti-collision positioning method of a robot.

Background

In the autonomous walking process of the robot, the robot needs to constantly interact with the environment, so that the pose of the robot in the environment is obtained, and the robot can walk along a known route after the navigation path is planned. Thereby completing corresponding patrol, distribution, guidance and other tasks. Positioning is the basis for navigation and task execution. However, in actual engineering, a large or small positioning error inevitably occurs, and when the robot walks around the edge of the obstacle, the pose of the robot sometimes reaches the inside of the obstacle, and this error can cause a collision or even cannot be executed in subsequent navigation path planning. Repositioning of the robot is required to obtain a more accurate pose of the robot.

At present, the measures at the positioning level generally include a Kalman filtering method, an amcl Monte Carlo method and a graph optimization method.

The Kalman filtering method is to fuse sensor data such as an odometer, imu and GPS to generate Gaussian estimation of the robot pose. This approach is less affected by the scene, but has accumulated errors, since laser/visual data may not be used.

The amcl Monte Carlo is based on a probability method, the initial position of a robot is assumed to be unknown, the robot is enabled to move for a certain distance, and the pose of the robot is determined after the robot is moved for a certain distance through laser map matching and random position resampling calculation of particles. This approach requires the robot to move a distance before it can be positioned gradually to the correct position, but many scenarios do not have the robot to position by this method of movement. Such as small spaces like feature similar scenes and hallways, the probability of positioning failure is greatly improved.

The laser radar signpost (or image key frame) based on the graph optimization is positioned, the robot is required to move, a transfer matrix between the pose and the signpost is obtained, and therefore the current pose of the robot is estimated. It has accumulated errors as in the kalman filtering method. In addition, the pose is adjusted through the return detection of the key frame, but when the key frame features are not good, the return failure is also caused.

Disclosure of Invention

The invention aims at the problems in the prior art and provides an anti-collision positioning method of a robot.

The invention solves the technical problems by the following technical means:

an anti-collision positioning method of a robot, the method comprising:

acquiring the pose of the robot relative to a world coordinate system based on an odometer;

randomly placing n particles carrying pose near the odometer;

when moving a distance, selecting particles with highest matching degree with the current scene map to update the particles into the pose of the robot;

and when the pose of the robot is updated, synchronously updating a forbidden area of the current scene map, wherein the forbidden area refers to the inside of an obstacle area which is inaccessible to the robot.

Further, the method for updating the forbidden area of the current scene map comprises the following steps:

registering the selected radar point cloud of the particle with the highest matching degree with the current scene map, marking the unmatched position of the radar point cloud data and the current scene map as an unreliable factor, further judging whether the unreliable factor is an obstacle area of the current scene map, and if so, adding the current positioning result into the forbidden area.

Further, the method for adding the current positioning result to the forbidden area comprises the following steps:

calculating the measurement result of the radar point cloud of the selected particle with the highest matching degree with the current scene map, mapping the corresponding point to the corresponding position on the current scene map, and adding a forbidden area at each position point to obtain a forbidden area layer;

expanding the obstacle area on the current scene map for a certain distance to obtain an expanded area, and obtaining a scene expansion layer;

and superposing the forbidden area layer, the scene expansion layer and the current scene map to update the forbidden area on the current scene map.

Further, the method further comprises the following steps:

randomly placing particles in a non-obstacle area near an odometer of the updated current scene map;

checking the matching degree of radar data of all particles and a current scene map, recording whether the matching degree and the unmatched area are inside an obstacle area of a static map, if yes, marking the matching degree and the unmatched area as an unreliable factor, carrying out weighted summation on the unreliable factor and the matching degree to obtain the confidence degree of all particles, wherein the confidence degree represents the position and pose credibility degree of the corresponding particles, selecting the particle with the largest credibility degree, and obtaining the optimal position and pose of the particles, wherein the position and pose are positioning results.

The beneficial effects of the invention are as follows: the method can improve the positioning precision, improve the walking capacity of the robot at the edge of the obstacle, and avoid colliding with the obstacle to damage property.

Drawings

FIG. 1 is a flow chart of an anti-collision positioning method of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described in conjunction with the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Examples

Referring to fig. 1, the anti-collision positioning method provided in this embodiment includes: and (3) carrying out global positioning based on the amcl, in the global positioning process, converting the pose of the odometer relative to a robot coordinate system into a world coordinate system, thus obtaining the pose of the radar coordinate system relative to the world coordinate system, randomly placing n (such as n=1000) particles carrying the pose near the odometer, moving a distance, recording the matching degree of each particle and the scene map 1, selecting the particle (namely the pose) with the highest matching degree from the particles, updating the pose as the pose of the odometer in the global positioning, and further processing and updating the map scene based on the pose updating.

The method for processing and updating the map scene comprises the following steps:

and selecting particles with highest matching degree from the particles to register the radar point cloud with the scene map 1, and performing secondary processing on the scene map 1 in the point cloud registration process. And checking the position where the radar data is not matched with the scene map 1, marking the position as an unreliable factor, further checking whether the unreliable factor is in the obstacle of the scene map 1, if so, adding the current positioning result into the forbidden area, and if not, finishing positioning updating.

More specifically:

step 1), adding a forbidden area layer. And calculating a measurement result of the current radar point cloud based on the pose obtained by positioning, mapping corresponding points to corresponding positions on the scene map, and adding forbidden areas at each position point.

Step 2), adding a scene expansion layer. And (3) expanding the obstacle area on the scene map for a certain distance to obtain an expanded area according to the principle of the step 1).

Step 3), performing superposition calculation on the forbidden area layer, the scene expansion layer and the scene map in the step 1) and the step 2) to obtain a new scene map 2, wherein the new scene map 2 is adopted in subsequent calculation.

Step 4) randomly placing particles in the non-obstacle area of the scene map 2 in the vicinity of the odometer, each particle being to represent a different odometer pose.

Step 5), checking the matching degree of radar data of all particles and the scene map 1, recording whether the matching degree and the unmatched area are inside an obstacle of the static map (which is unreasonable measurement), if so, marking the matching degree and the unmatched area as an untrusted factor, carrying out weighted summation on the untrusted factor and the matching degree to obtain the confidence degree of all particles, wherein the confidence degree represents the position and pose credibility degree of the corresponding particles, selecting the particle with the largest credibility degree, and obtaining the optimal position and pose of the particles, wherein the position and pose is a positioning result.

It is noted that relational terms such as first and second, and the like, if any, are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for collision-resistant positioning of a robot, the method comprising:

randomly placing n particles carrying pose near the odometer;

when the pose of the robot is updated, synchronously updating a forbidden area of the current scene map, wherein the forbidden area refers to the inside of an inaccessible obstacle area of the robot;

the method for updating the forbidden area of the current scene map comprises the following steps:

registering the selected radar point cloud of the particle with the highest matching degree with the current scene map, marking the unmatched position of the radar point cloud data and the current scene map as an unreliable factor, further judging whether the unreliable factor is an obstacle area of the current scene map, and if so, adding the current positioning result into the forbidden area;

the method for adding the current positioning result to the forbidden area is as follows:

superposing and calculating the forbidden area layer, the scene expansion layer and the current scene map to update the forbidden area on the current scene map;

further comprises:

checking the matching degree of radar data of all particles and a current scene map, recording whether an unmatched area is in an obstacle area of a static map while recording the matching degree, if so, marking the unmatched area as an untrusted factor, carrying out weighted summation on the untrusted factor and the matching degree to obtain the confidence degree of all the particles, wherein the confidence degree represents the position and pose credibility of the corresponding particles, selecting the particle with the largest confidence degree, and obtaining the optimal position and pose of the particles, wherein the position and pose are positioning results.