CN117761704A - Method and system for estimating relative positions of multiple robots - Google Patents
Method and system for estimating relative positions of multiple robots Download PDFInfo
- Publication number
- CN117761704A CN117761704A CN202311668765.8A CN202311668765A CN117761704A CN 117761704 A CN117761704 A CN 117761704A CN 202311668765 A CN202311668765 A CN 202311668765A CN 117761704 A CN117761704 A CN 117761704A
- Authority
- CN
- China
- Prior art keywords
- robot
- point cloud
- cluster
- information
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000005457 optimization Methods 0.000 claims abstract description 25
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 239000011159 matrix material Substances 0.000 claims abstract description 20
- 230000009466 transformation Effects 0.000 claims abstract description 13
- 238000012897 Levenberg–Marquardt algorithm Methods 0.000 claims abstract description 6
- 238000001914 filtration Methods 0.000 claims description 12
- 238000004891 communication Methods 0.000 claims description 8
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 2
- 230000011218 segmentation Effects 0.000 claims description 2
- 230000001131 transforming effect Effects 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 7
- 230000001149 cognitive effect Effects 0.000 abstract description 2
- 238000002474 experimental method Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 3
- 238000012790 confirmation Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The method comprises the steps of configuring each robot to acquire surrounding point cloud information of a sensor within a certain period of time by using a laser radar, converting the surrounding point cloud information into relative distance and angle information in a homogeneous transformation matrix form, and performing clustering operation on the point cloud information to generate a corresponding identifier; and then matching the relative distance and angle information of each robot through a pairing optimization algorithm, and obtaining the corresponding relation between the cluster identifier and each robot by constructing an optimization model and solving by using a Levenberg-Marquardt algorithm of a nonlinear least square method. The invention can realize efficient task cooperation, more accurate and robust co-location of the multi-robot cluster by using a reliable point cloud cluster identification technology and a nonlinear optimization method. And the overall cognitive level and team efficiency of the multi-robot are improved.
Description
Technical Field
The invention relates to a technology in the field of industrial automation, in particular to a method and a system for estimating the relative positions of multiple robots.
Background
The problem of positioning and co-locating of multiple robots, which is a complex and troublesome industrial problem at present, refers to that a group of robots jointly determine their positions relative to a global coordinate system so as to realize a co-task. In the past technology, the methods used can be summarized as centralized positioning, global positioning, and collective control methods. However, the existing centralized positioning technology can accumulate the obtained errors, so that the accumulated errors are continuously enlarged, and the positioning effect is affected; global positioning technology requires that each robot needs to use a completely consistent algorithm and configuration, which leads to the problem that the overall consistency is difficult to meet; the collective control technology completely fixes parameter setting and has no mobility and maintainability.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method and a system for estimating the relative positions of multiple robots, and a reliable point cloud cluster identification technology and a nonlinear optimization method are used for enabling the multiple robot clusters to realize efficient task cooperation and more accurate and robust cooperative positioning. And the overall cognitive level and team efficiency of the multi-robot are improved.
The invention is realized by the following technical scheme:
the invention relates to a method for estimating the relative positions of multiple robots, which comprises the steps of acquiring surrounding point cloud information of a sensor in a certain time period by configuring each robot through a laser radar, converting the surrounding point cloud information into relative distance and angle information in a homogeneous transformation matrix form, and carrying out clustering operation on the point cloud information to generate a corresponding identifier; and then matching the relative distance and angle information of each robot through a pairing optimization algorithm, and obtaining the corresponding relation between the cluster identifier and each robot by constructing an optimization model and solving by using a Levenberg-Marquardt algorithm of a nonlinear least square method.
The invention relates to a system for realizing the method, which comprises the following steps: the system comprises a point cloud preprocessing module, a point cloud clustering module, a message communication module and a pairing algorithm module, wherein: the point cloud preprocessing module performs ground filtering processing according to the point cloud information obtained by the laser radar sensor to obtain filtered point cloud for filtering ground points, then performs scene screening on the point cloud obtained after ground filtering, and deletes part of discrete points to obtain more concentrated and accurate target point cloud information; the point cloud clustering module performs clustering operation on the target point cloud information to obtain distance and angle information of each cluster from a coordinate system of the point cloud clustering module; the message communication module timely distributes clustering results obtained after self point cloud clustering through a local area network according to information in a message queue in each robot and receives clustering results obtained after point cloud clustering of other robots; the pairing algorithm module constructs a nonlinear optimization function according to the clustering result, and completes pairing among robots according to the distance and angle relation between minimized different robots, so that each robot obtains a relative position relation with other robots.
Technical effects
The invention can realize the confirmation of the position of the robot in a completely unknown environment by pairing based on a nonlinear optimization function, and can realize the interaction of multiple robots so as to help other robots to confirm the positions of the robots, compared with other co-location algorithms, the method can greatly reduce the dependence on environment information and the dependence on sensors, and the actual application scene is that under the complex environment with a plurality of robots, the global positions of all other robots can be obtained by the relative position estimation algorithm of the invention as long as one robot establishes own global position, thereby being capable of executing more complex tasks.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic diagram of the effect of the embodiment.
Detailed Description
As shown in fig. 1, this embodiment relates to a method for estimating relative positions of multiple robots, which specifically includes:
step 1) setting point cloud information obtained by each robot through laser radar, and carrying out ground filtering and scene segmentation on the point cloud information, wherein the method specifically comprises the following steps:
1.1 removing outliers: traversing the obtained original point cloud, and deleting a certain proportion of images with the highest z-axis (height) and the lowest z-axis (height) in the point cloud. Meanwhile, deleting some points far away from other point clouds through nearest neighbor searching, and completing the work of removing isolated points.
1.2 ground filtering: using the PTD algorithm, a bounding box is defined that contains triangles of coordinates x_top, y_top, x_bottom, y_bottom, dividing the entire point cloud data into:wherein: width w and height h.
1.3 PTD traversal starts from the lowest point as seed point: calculating a potential point determination in a triangle, knowing that a potential point can only be on the triangle' S sides and vertices, when the slope S inclination exceeds a threshold, the potential point cannot be on the face, so that the determination of the next point is made after the point outside the triangle is excluded as the ground point, the slope S inclination Wherein: a and B are one sides of a triangle binding box, a point P is an outer point which does not belong to the triangle binding box in the point cloud image, PA, AB is a vector, I is a two-mode form of the vector, and an angle between the point P and a plane formed by the triangle binding box is taken as an inclination angle theta of the triangle PA 。
Step 2) setting each robot to perform clustering operation on the obtained point cloud respectively, and endowing each cluster with a unique identifier, wherein the unique identifier represents other perceived robots, and the method specifically comprises the following steps:
2.1 Using Euclidean distance clustering, taking the distance between the set point cloud clusters as an external threshold value and the internal distance of the point cloud clusters as an internal threshold value;
2.2 Any point P of the point cloud is selected, other points are traversed, intra-cluster points with Euclidean distance smaller than an internal threshold value are used as P points, and after traversing is completed, the points are marked with a cluster center P and a unique ID of a cluster.
2.3 And (3) repeating the process by taking the points outside the clusters as the centers, and traversing the obtained cluster centers.
2.4 If there are other cluster centers whose euclidean distance is smaller than the external threshold value, then the two clusters are combined and the process of finding the points in the clusters is repeated. Until the program cannot find a new cluster, namely, filtering out a proper cluster according to the distance and the number of points in the cluster, and outputting the coordinate sum of the cluster center and the distance to the program to the next module.
Step 3) setting each robot to respectively calculate the relative distance and angle information of the robot and other robots, and each robot respectively exchanges the cluster identifier and the corresponding relative distance and angle information observed by the robot with the other robots by using a message queue communication mechanism.
The relative distance and angle information is represented using a homogeneous transformation matrix.
The message queue communication mechanism specifically comprises:
i) The published message queue: the robot packages the data collected by itself and distributes the data to the ROS local area network at the frequency of 1Hz
ii) a subscription message queue: the robot opens own local area network port, the program can monitor the data issued by other robots through the local area network, a queue data structure is established for storage, the integrity of the data packet is checked every 10s, and if the data is complete, the data is injected into the pairing optimization algorithm.
Step 4) setting each robot to establish constraint conditions through the duality of the relative distance and angle information according to the respective pairing optimization algorithm and solving, and specifically comprising the following steps:
4.1 Constructing an optimization model: euclidean distance establishes an optimization objective: minimum H BX *H XA -H AB || 2 Wherein: the euclidean norm represents the difference between the two matrices; the homogeneous transformation matrix under the view angle of the robot A is H AB Under the view angle of robot B, robot X is observed under the view angle of BHomogeneous transformation matrix is H BX With the homogeneous transformation matrix H of the robot X under the view angle A XA Is a variable.
4.2 Solving the target in the step 4.1 by using a Levenberg-Marquardt algorithm of a nonlinear least square method, and obtaining an optimal inverse transformation matrix H_XA, namely indicating that X is an A robot under the view angle of the robot, wherein the method specifically comprises the following steps of: transforming the optimization problem into minimum J (H XA )＝||H BX *H XA -H AB || 2 Wherein: j (H) XA ) An objective function representing an optimization problem; first initialize H XA Then calculates the initial value of the objective function with respect to H XA And updating and adjusting the estimated value by using a Levenberg-Marquardt rule to enable the objective function to reach the requirement.
The Levenberg-Marquardt rule is specifically as follows: wherein: />H for updated estimation XA A matrix; />For H in the last iteration XA Estimating a matrix; h () is the objective function J (H XA ) Concerning H XA To describe the curvature of the objective function; grad () is the objective function J (J) XA ) Concerning H XA Gradient of (i.e. J vs. H) XA Is the first derivative of (a); lambda is a regularization parameter, namely the Levenberg-Marquardt parameter, and I is an identity matrix.
When λ is smaller, the algorithm is more prone to newton's method, and when λ is larger, the algorithm is more prone to gradient descent method.
Step 5) setting the pairing relation between the cluster identification and the robot id which are issued by each robot to the local area network, and issuing pairing information to inform the relative positions and angle information of the robot and other robots for the subsequent nodes to optimize and schedule.
The pairing information refers to: for example, when the robot a receives the position information h_bx of the cluster sent by the robot B, the robot a knows that one of the point cloud clusters h_ax of the robot a is the representative of the robot B through the pairing algorithm, and then the robot a issues the ID of the point cloud cluster paired with the robot B and the ID of the point cloud cluster belonging to the robot a in the point cloud sent by the robot B.
Through specific practical experiments, an automobile chassis and a small autonomous robot are adopted as two robot platforms, the two platforms are respectively provided with a laser radar and can communicate through a local area network, the experimental environment is a robot A and a robot B, the experiment is that the robot A is used as a reference, and the information obtained through point cloud clustering does not pass through paired position information and is used as the position of the observed robot B; and in the second experiment, after the two are subjected to pairing optimization after point cloud clustering, the positions of the robot B are obtained after the distances and the directions of the pair of observed opponents are averaged. Comparing the positions obtained in the first experiment and the second experiment with absolute coordinates converted from GPS true value positions of the robot B, wherein the absolute coordinates are as follows: the distance and the azimuth under the two visual angles obtained after pairing optimization are subjected to average processing, and the positioning deviation generated under long-time running is remarkably reduced for the self-positioning autonomous robot only through the automobile chassis. In the experiment, the system can stably run until the electric quantity of the autonomous robot is insufficient.
As shown in fig. 2, the average error using the present method is 0.0263, whereas the average error of the conventional method is 0.0270. The error has obvious rising trend along with the increase of the operation time in the traditional method; but the error of the algorithm applied by the invention is relatively stable. It follows that a stable positioning error can be maintained during long operation by using the method.
The foregoing embodiments may be partially modified in numerous ways by those skilled in the art without departing from the principles and spirit of the invention, the scope of which is defined in the claims and not by the foregoing embodiments, and all such implementations are within the scope of the invention.
Claims (8)
1. The method is characterized in that after each robot is configured to acquire surrounding point cloud information of a sensor within a certain period of time by adopting a laser radar, the surrounding point cloud information is converted into relative distance and angle information in a homogeneous transformation matrix form, and clustering operation is performed on the point cloud information to generate a corresponding identifier; and then matching the relative distance and angle information of each robot through a pairing optimization algorithm, and obtaining the corresponding relation between the cluster identifier and each robot by constructing an optimization model and solving by using a Levenberg-Marquardt algorithm of a nonlinear least square method.
2. The method for estimating relative positions of multiple robots according to claim 1, comprising:
step 1) setting point cloud information obtained by each robot through laser radar, and carrying out ground filtering and scene segmentation on the point cloud information, wherein the method specifically comprises the following steps:
1.1 removing outliers: traversing the obtained original point cloud, and deleting images with a certain proportion of the highest z-axis (height) and the lowest z-axis (height) in the point cloud; meanwhile, deleting some points far away from other point clouds through nearest neighbor searching, and completing the work of removing isolated points;
1.2 ground filtering: using the PTD algorithm, a bounding box is defined that contains triangles of coordinates x_top, y_top, x_bottom, y_bottom, dividing the entire point cloud data into:wherein: width w and height h;
1.3 PTD traversal starts from the lowest point as seed point: calculating a potential point determination in a triangle, knowing that a potential point is only on the triangle' S sides and vertices, when the slope S dip exceeds a threshold, the potential point is not possible on this face, so points outside this triangle are excludedThe inclination of the slope S is determined for the next point after the ground pointWherein: a and B are one sides of a triangle binding box, a point P is an outer point which does not belong to the triangle binding box in the point cloud image, PA, AB is a vector, I is a two-mode form of the vector, and an angle between the point P and a plane formed by the triangle binding box is taken as an inclination angle theta of the triangle PA ；
Step 2) setting each robot to perform clustering operation on the obtained point cloud respectively, and endowing each cluster with a unique identifier, wherein the unique identifier represents other perceived robots, and the method specifically comprises the following steps:
2.1 Using Euclidean distance clustering, taking the distance between the set point cloud clusters as an external threshold value and the internal distance of the point cloud clusters as an internal threshold value;
2.2 Selecting any point P of the point cloud, traversing other points, marking the points with a clustering center P and a unique ID of a cluster after traversing, wherein the points are used as intra-cluster points of the P points with Euclidean distance smaller than an internal threshold value;
2.3 Repeating the above process by taking the points outside the clusters as the centers, and traversing the obtained cluster centers;
2.4 When other cluster centers with Euclidean distance smaller than the external threshold value exist in the cluster centers, combining the two clusters, and repeatedly searching the points in the clusters; until a program cannot find a new cluster, namely filtering out a proper cluster according to the distance and the number of points in the cluster, and outputting the coordinate sum of the cluster center and the distance to the program to a next module;
step 3) each robot is set to respectively calculate the relative distance and angle information of the robot and other robot bodies, and each robot respectively exchanges the cluster identifier and the corresponding relative distance and angle information observed by the robot with other robots by using a message queue communication mechanism;
step 4) setting each robot to establish constraint conditions through the duality of the relative distance and angle information according to the respective pairing optimization algorithm and solving, and specifically comprising the following steps:
4.1 Constructing an optimization model: euclidean distance establishes an optimization objective: minimum H BX *H XA -H AB || 2 Wherein: the euclidean norm represents the difference between the two matrices; the homogeneous transformation matrix under the view angle of the robot A is H AB The homogeneous transformation matrix of the robot X under the B visual angle is observed to be H under the B visual angle of the robot B BX With the homogeneous transformation matrix H of the robot X under the view angle A XA Is a variable;
4.2 Solving the target in the step 4.1 by using a Levenberg-Marquardt algorithm of a nonlinear least square method, and obtaining an optimal inverse transformation matrix H_XA, namely indicating that X is an A robot under the view angle of the robot, wherein the method specifically comprises the following steps of: transforming the optimization problem into a minimizer J (H XA )＝||H BX *H XA -H AB || 2 Wherein: j (H) XA ) An objective function representing an optimization problem; first initialize H XA Then calculates the initial value of the objective function with respect to H XA The gradient and Hessian matrix of the model (C) are updated and the estimated value is adjusted by using the Levenberg-Marquardt rule, so that the objective function reaches the requirement;
step 5) setting the pairing relation between the cluster identification and the robot id which are issued by each robot to the local area network, and issuing pairing information to inform the relative positions and angle information of the robot and other robots for the subsequent nodes to optimize and schedule.
3. The method of estimating relative positions of multiple robots according to claim 2, wherein the relative distance and angle information is represented using a homogeneous transformation matrix.
4. The method for estimating relative positions of multiple robots according to claim 2, wherein said message queue communication mechanism comprises:
i) The published message queue: the robot packs the data collected by the robot and distributes the data to the local area network of the ROS at the frequency of 1 Hz;
ii) a subscription message queue: the robot opens own local area network port, the program can monitor the data issued by other robots through the local area network, a queue data structure is established for storage, the integrity of the data packet is checked every 10s, and if the data is complete, the data is injected into the pairing optimization algorithm.
5. The method for estimating relative positions of multiple robots according to claim 1 or 2, wherein the Levenberg-Marquardt algorithm is specifically:wherein:h for updated estimation XA A matrix; />For H in the last iteration XA Estimating a matrix; h () is the objective function J (H XA ) Concerning H XA To describe the curvature of the objective function; grad () is the objective function J (H) XA ) Concerning H XA Gradient of (i.e. J vs. H) XA Is the first derivative of (a); lambda is a regularization parameter, namely the Levenberg-Marquardt parameter, and I is an identity matrix.
6. The method of estimating relative positions of multiple robots according to claim 5, wherein the algorithm is more prone to newton's method when λ is smaller and to gradient descent when λ is larger.
7. The method for estimating relative positions of multiple robots according to claim 1, wherein the pairing information is: for example, when the robot a receives the position information h_bx of the cluster sent by the robot B, the robot a knows that one of the point cloud clusters h_ax of the robot a is the representative of the robot B through the pairing algorithm, and then the robot a issues the ID of the point cloud cluster paired with the robot B and the ID of the point cloud cluster belonging to the robot a in the point cloud sent by the robot B.
8. A multi-robot relative position estimation system implementing the method of any one of claims 1-7, comprising: the system comprises a point cloud preprocessing module, a point cloud clustering module, a message communication module and a pairing algorithm module, wherein: the point cloud preprocessing module performs ground filtering processing according to the point cloud information obtained by the laser radar sensor to obtain filtered point cloud for filtering ground points, then performs scene screening on the point cloud obtained after ground filtering, and deletes part of discrete points to obtain more concentrated and accurate target point cloud information; the point cloud clustering module performs clustering operation on the target point cloud information to obtain distance and angle information of each cluster from a coordinate system of the point cloud clustering module; the message communication module timely distributes clustering results obtained after self point cloud clustering through a local area network according to information in a message queue in each robot and receives clustering results obtained after point cloud clustering of other robots; the pairing algorithm module constructs a nonlinear optimization function according to the clustering result, and completes pairing among robots according to the distance and angle relation between minimized different robots, so that each robot obtains a relative position relation with other robots.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311668765.8A CN117761704B (en) | 2023-12-07 | 2023-12-07 | Method and system for estimating relative positions of multiple robots |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311668765.8A CN117761704B (en) | 2023-12-07 | 2023-12-07 | Method and system for estimating relative positions of multiple robots |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117761704A true CN117761704A (en) | 2024-03-26 |
| CN117761704B CN117761704B (en) | 2024-08-13 |
Family
ID=90317245
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311668765.8A Active CN117761704B (en) | 2023-12-07 | 2023-12-07 | Method and system for estimating relative positions of multiple robots |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117761704B (en) |
Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130245824A1 (en) * | 2012-03-15 | 2013-09-19 | Gm Global Technology Opeations Llc | Method and system for training a robot using human-assisted task demonstration |
| US20160059419A1 (en) * | 2014-09-03 | 2016-03-03 | Canon Kabushiki Kaisha | Robot apparatus and method for controlling robot apparatus |
| CN109709801A (en) * | 2018-12-11 | 2019-05-03 | 智灵飞（北京）科技有限公司 | A kind of indoor unmanned plane positioning system and method based on laser radar |
| CA3036241A1 (en) * | 2018-03-09 | 2019-09-09 | Jena-Optronik Gmbh | Method for initializing a tracking algorithm, method for training an artificial neural network, computer program product, computer-readable storage medium, and data carrier signal for carrying out such methods and device for data processing |
| CN111563442A (en) * | 2020-04-29 | 2020-08-21 | 上海交通大学 | Slam method and system for fusing point cloud and camera image data based on laser radar |
| WO2021109167A1 (en) * | 2019-12-06 | 2021-06-10 | 苏州艾吉威机器人有限公司 | Three-dimensional laser mapping method and system |
| WO2021238306A1 (en) * | 2020-05-25 | 2021-12-02 | 华为技术有限公司 | Method for processing laser point cloud and related device |
| CN114004869A (en) * | 2021-09-18 | 2022-02-01 | 浙江工业大学 | Positioning method based on 3D point cloud registration |
| CN114419147A (en) * | 2021-11-16 | 2022-04-29 | 新兴际华集团有限公司 | Rescue robot intelligent remote human-computer interaction control method and system |
| WO2022095302A1 (en) * | 2020-11-05 | 2022-05-12 | 湖南大学 | Hierarchical gaussian mixture model-based fast and robust robot three-dimensional reconstruction method |
| US20220291692A1 (en) * | 2021-03-15 | 2022-09-15 | Omron Corporation | Method and Apparatus for Updating an Environment Map Used by Robots for Self-localization |
| US20220309835A1 (en) * | 2021-03-26 | 2022-09-29 | Harbin Institute Of Technology, Weihai | Multi-target detection and tracking method, system, storage medium and application |
| CN115741720A (en) * | 2022-12-13 | 2023-03-07 | 合肥工业大学 | Zero calibration system and method for robot based on binocular vision technology and LM algorithm |
| CN115855062A (en) * | 2022-12-07 | 2023-03-28 | 重庆理工大学 | Autonomous mapping and path planning method for indoor mobile robot |
| CN116128966A (en) * | 2023-02-28 | 2023-05-16 | 上海交通大学 | Semantic positioning method based on environmental object |
| CN116184431A (en) * | 2022-12-04 | 2023-05-30 | 之江实验室 | Unmanned aerial vehicle autonomous navigation positioning map construction method, device and system |
| CN116758153A (en) * | 2023-05-31 | 2023-09-15 | 重庆大学 | Multi-factor graph-based back-end optimization method for accurate pose acquisition of robot |
-
2023
- 2023-12-07 CN CN202311668765.8A patent/CN117761704B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130245824A1 (en) * | 2012-03-15 | 2013-09-19 | Gm Global Technology Opeations Llc | Method and system for training a robot using human-assisted task demonstration |
| US20160059419A1 (en) * | 2014-09-03 | 2016-03-03 | Canon Kabushiki Kaisha | Robot apparatus and method for controlling robot apparatus |
| CA3036241A1 (en) * | 2018-03-09 | 2019-09-09 | Jena-Optronik Gmbh | Method for initializing a tracking algorithm, method for training an artificial neural network, computer program product, computer-readable storage medium, and data carrier signal for carrying out such methods and device for data processing |
| CN109709801A (en) * | 2018-12-11 | 2019-05-03 | 智灵飞（北京）科技有限公司 | A kind of indoor unmanned plane positioning system and method based on laser radar |
| WO2021109167A1 (en) * | 2019-12-06 | 2021-06-10 | 苏州艾吉威机器人有限公司 | Three-dimensional laser mapping method and system |
| CN111563442A (en) * | 2020-04-29 | 2020-08-21 | 上海交通大学 | Slam method and system for fusing point cloud and camera image data based on laser radar |
| WO2021238306A1 (en) * | 2020-05-25 | 2021-12-02 | 华为技术有限公司 | Method for processing laser point cloud and related device |
| WO2022095302A1 (en) * | 2020-11-05 | 2022-05-12 | 湖南大学 | Hierarchical gaussian mixture model-based fast and robust robot three-dimensional reconstruction method |
| US20220291692A1 (en) * | 2021-03-15 | 2022-09-15 | Omron Corporation | Method and Apparatus for Updating an Environment Map Used by Robots for Self-localization |
| US20220309835A1 (en) * | 2021-03-26 | 2022-09-29 | Harbin Institute Of Technology, Weihai | Multi-target detection and tracking method, system, storage medium and application |
| CN114004869A (en) * | 2021-09-18 | 2022-02-01 | 浙江工业大学 | Positioning method based on 3D point cloud registration |
| CN114419147A (en) * | 2021-11-16 | 2022-04-29 | 新兴际华集团有限公司 | Rescue robot intelligent remote human-computer interaction control method and system |
| CN116184431A (en) * | 2022-12-04 | 2023-05-30 | 之江实验室 | Unmanned aerial vehicle autonomous navigation positioning map construction method, device and system |
| CN115855062A (en) * | 2022-12-07 | 2023-03-28 | 重庆理工大学 | Autonomous mapping and path planning method for indoor mobile robot |
| CN115741720A (en) * | 2022-12-13 | 2023-03-07 | 合肥工业大学 | Zero calibration system and method for robot based on binocular vision technology and LM algorithm |
| CN116128966A (en) * | 2023-02-28 | 2023-05-16 | 上海交通大学 | Semantic positioning method based on environmental object |
| CN116758153A (en) * | 2023-05-31 | 2023-09-15 | 重庆大学 | Multi-factor graph-based back-end optimization method for accurate pose acquisition of robot |
Non-Patent Citations (9)
| Title |
|---|
| SHUO WANG 等: "Distributed Relative Localization Algorithms for Multi-Robot Networks: A Survey", 《SENSORS》, 21 February 2023 (2023-02-21), pages 1 - 28 * |
| TING YUAN 等: "Lane Detection and Estimation from Surround View Camera Sensing Systems", 《2023 IEEE SENSORS》, 28 November 2023 (2023-11-28), pages 1 - 4 * |
| YAN MENG 等: "Autonomous robot calibration using vision technology", 《ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING》, 31 August 2007 (2007-08-31), pages 436 - 446 * |
| ZHIYUAN ZHU 等: "Indoor Multi-Robot Cooperative Mapping Based on Geometric Features", 《IEEE ACCESS》, 17 May 2021 (2021-05-17), pages 74574 - 74588, XP011856389, DOI: 10.1109/ACCESS.2021.3081252 * |
| 曹文祺 等: "具有可参数化不确定性系统的对偶自适应模型预测控制", 《控制理论与应用》, 9 April 2019 (2019-04-09), pages 1197 - 1206 * |
| 李杨 等: "基于Levenberg-Marquardt算法的串联协作机器人精度标定研究", 《黑龙江工业学院学报(综合版)》, 20 March 2021 (2021-03-20), pages 112 - 117 * |
| 杨丽 等: "不确定环境下多机器人的动态编队控制", 《机器人》, 15 March 2010 (2010-03-15), pages 283 - 288 * |
| 蒋丰: "移动机器人SLAM技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》, 15 March 2022 (2022-03-15), pages 1 - 79 * |
| 陈士凯: "自主定位导航技术在商用服务机器人中的应用与挑战", 《人工智能》, 10 June 2018 (2018-06-10), pages 103 - 115 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117761704B (en) | 2024-08-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109725327B (en) | Method and system for building map by multiple machines | |
| JP6150531B2 (en) | Terrain information acquisition device, terrain information acquisition system, terrain information acquisition method and program | |
| CN109100731B (en) | Mobile robot positioning method based on laser radar scanning matching algorithm | |
| CN109541997A (en) | It is a kind of towards the quick, intelligent programmed method of plane/almost plane workpiece spray robot | |
| JP5018458B2 (en) | Coordinate correction method, coordinate correction program, and autonomous mobile robot | |
| JP5480667B2 (en) | Position / orientation measuring apparatus, position / orientation measuring method, program | |
| CN110908374B (en) | Mountain orchard obstacle avoidance system and method based on ROS platform | |
| CN113034593B (en) | 6D pose labeling method, system and storage medium | |
| CN111273312A (en) | Intelligent vehicle positioning and loop-back detection method | |
| WO2021195939A1 (en) | Calibrating method for external parameters of binocular photographing device, movable platform and system | |
| CN116091724A (en) | Building digital twin modeling method | |
| CN114792338A (en) | Vision fusion positioning method based on prior three-dimensional laser radar point cloud map | |
| CN117830521A (en) | Virtual park construction method and management method based on digital twin | |
| JP2022138037A (en) | Information processor, information processing method, and program | |
| CN114972640B (en) | Cable three-dimensional reconstruction and parameter calculation method based on point cloud | |
| CN115965790A (en) | Oblique photography point cloud filtering method based on cloth simulation algorithm | |
| CN117761704B (en) | Method and system for estimating relative positions of multiple robots | |
| Dorokhov et al. | Recognition of cow teats using the 3D-ToF camera when milking in the “herringbone” milking parlor | |
| Breitenmoser et al. | Distributed coverage control on surfaces in 3d space | |
| Petersen et al. | Target tracking and following from a multirotor UAV | |
| CN111815684A (en) | Space multivariate feature registration optimization method and device based on unified residual error model | |
| CN115239899B (en) | Pose map generation method, high-precision map generation method and device | |
| CN115026834B (en) | Method for realizing correction function based on robot template program | |
| Mukhopadhyay et al. | Multi-robot Map Exploration Based on Multiple Rapidly-Exploring Randomized Trees | |
| CN115166686A (en) | Multi-unmanned aerial vehicle distributed cooperative positioning and mapping method in satellite rejection environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |