CN113688496A - Robot mapping algorithm precision simulation evaluation method - Google Patents
Robot mapping algorithm precision simulation evaluation method Download PDFInfo
- Publication number
- CN113688496A CN113688496A CN202110757015.2A CN202110757015A CN113688496A CN 113688496 A CN113688496 A CN 113688496A CN 202110757015 A CN202110757015 A CN 202110757015A CN 113688496 A CN113688496 A CN 113688496A
- Authority
- CN
- China
- Prior art keywords
- simulation
- algorithm
- robot
- point
- mapping
- 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
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 84
- 238000004088 simulation Methods 0.000 title claims abstract description 78
- 238000013507 mapping Methods 0.000 title claims abstract description 51
- 238000011156 evaluation Methods 0.000 title claims abstract description 23
- 238000000034 method Methods 0.000 claims abstract description 11
- 230000004888 barrier function Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000010408 sweeping Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 241001417527 Pempheridae Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
- G01C21/30—Map- or contour-matching
- G01C21/32—Structuring or formatting of map data
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Manipulator (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention relates to a robot mapping algorithm precision simulation evaluation method, which comprises the steps of building a robot simulation environment, including a simulation robot, a simulation sensor and a simulation scene, and butting a robot mapping algorithm with a simulation system; randomly selecting a plurality of key points in a simulation scene as comparison points; and extracting the position coordinates of the corresponding key points in the algorithm after the graph is built by the graph building algorithm, comparing the position coordinates with the real values in the simulation system, and taking the average value as the graph building precision value of the graph building algorithm. The method realizes the evaluation of the accuracy of the mapping algorithm by using a simulation means, solves the evaluation problem of the accuracy of the mapping algorithm of different kinds of robots, particularly provides a fair scheme and means which can be compared with each other and avoid cheating for different mapping algorithms of different manufacturers or brands of robots of the same type, improves the evaluation efficiency and saves the evaluation time and cost.
Description
Technical Field
The invention relates to a testing technology, in particular to a robot mapping algorithm precision simulation evaluation method.
Background
With the rapid development of the robot industry, robot algorithms are also rapidly developed, robot algorithms of the same type are different, and performances of the robot algorithms may be different in a natural way, for example, a sweeping robot experiences iteration from random movement to ESLAM, and then to different algorithms and technologies such as laser SLAM and VSLAM. The progress and optimization of the algorithm have higher and higher occupation ratio on the improvement of the performance of the robot, and the evaluation work of the algorithm is more and more important.
However, the evaluation means of the algorithm is always lagged compared with the rapid development of the algorithm. Many robotic products perform acceptably in laboratories and factories, but once a real user working environment is reached, various errors occur. On one hand, the evaluation means is backward and incomplete, and on the other hand, enterprises cannot cover complicated and diversified user actual working environments during the test.
In the robot SLAM algorithm, the mapping algorithm is the basis of other algorithms such as navigation, positioning and the like, and is particularly important. For example, a household sweeper only needs to work in a small-area environment in a home, and even if the accuracy of the mapping algorithm is not high, no bad result can be caused, while an industrial robot, such as an industrial AGV, needs to work in a large-area environment such as a factory building and the like, and because the equipment is heavy, the value is high, if the mapping accuracy is low, serious results such as personal safety or property loss can be caused. The method for evaluating the accuracy of the mapping algorithm is a simple, universal and efficient evaluation method, and is a problem which needs to be solved urgently.
Disclosure of Invention
Aiming at the problem of how to accurately evaluate the mapping algorithm after the mapping algorithm is applied, a robot mapping algorithm precision simulation evaluation method is provided.
The technical scheme of the invention is as follows: a robot mapping algorithm precision simulation evaluation method specifically comprises the following steps:
1) the method comprises the steps of building a robot simulation environment which comprises a simulation robot, a simulation sensor and a simulation scene, and butting a robot mapping algorithm with a simulation system;
2) randomly selecting a plurality of key points in a simulation scene as comparison points;
3) the robot in the simulation system moves according to the output information of the mapping algorithm, the simulation system provides information of various sensors for the mapping algorithm, and the mapping algorithm performs motion mapping according to the received information of the sensors and the algorithm;
4) and extracting the position coordinates of the corresponding key points in the algorithm after the graph is built by the graph building algorithm, feeding the coordinates back to the simulation system, comparing the received coordinate values of each key point in the graph building algorithm with the real values by the simulation system, and taking the average value as the graph building accuracy value of the graph building algorithm.
Further, the key points in the step 2) are determined according to the simulation scene space and the position of the obstacle, all the key points are randomly selected from the following three types,
the first kind is: the intersection points or points on the intersection lines of the two wall bodies;
the second kind is: the intersection point or the point on the intersection line of the two right-angle outer planes of the rectangular barrier;
the third category: the center of a circle or a point on the circumference of a circular obstacle.
Further, the method for calculating the mapping accuracy value of the mapping algorithm in the step 4) is as follows:
calculating the distance difference between the algorithm coordinate and the actual coordinate of each key point according to the following formula:
in the formula: the Xi point is the real coordinate value of the X axis of the point i; the Yi point is a Y-axis real coordinate value of the point i; the Xi' point is the coordinate value of the X-axis map of the point i; yi is a Y-axis map coordinate value of the point i;
the map-building precision value E of the map-building algorithm is calculated according to the following formula:
in the formula: and N is the total number of the randomly selected key points.
The invention has the beneficial effects that: the robot mapping algorithm precision simulation evaluation method realizes the evaluation of the mapping algorithm precision by using a simulation means, solves the evaluation problem of the mapping algorithm precision of different kinds of robots, particularly provides a fair scheme and means for comparing different mapping algorithms of different manufacturers or brands of robots of the same type, avoids cheating, improves the evaluation efficiency and saves the evaluation time and cost.
Drawings
FIG. 1 is a schematic diagram of a simulation scenario according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of key point selection locations and markers according to the present invention.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
The robot simulation refers to the simulation of the body, the sensor and the physical environment of the robot, the simulation environment is connected with the real algorithm of the robot in a butt joint mode, and the algorithm is used for controlling the simulation robot to move in the simulation environment. Since the robot simulation technique is not the focus of the present invention, it will be briefly described only when referring to the present invention.
The method comprises the steps of setting up a robot simulation environment, wherein the robot simulation environment comprises a simulation robot, a simulation sensor and a simulation scene, connecting a robot mapping algorithm with a simulation system, enabling the robot in the simulation system to move according to output information of the mapping algorithm, providing information of various sensors for the mapping algorithm by the simulation system, and enabling the mapping algorithm to move to map according to received sensor information and an algorithm.
Randomly selecting a plurality of key points in a simulation scene as comparison points, extracting the coordinates of the corresponding key points in the algorithm after the graph building of the graph building algorithm, feeding the coordinates back to the simulation system, carrying out precision comparison on the received coordinate values of each key point in the graph building algorithm and the real values by the simulation system, and taking the average value as the precision value of the graph building algorithm. It can be seen that the more key points are selected, the closer the precision value is to the overall precision value.
Aiming at robots with different working environments and purposes, the mapping algorithm can be used for precision evaluation according to the method as long as a proper simulation scene and a proper key point number are selected. The mapping algorithm of the same type of robot, such as a sweeping robot, is evaluated by using the same simulation scene and the same number of key points at the same position, and the result can be used for comparing the advantages and disadvantages of different mapping algorithms of the same type of robot.
The technical scheme of selecting key points in a simulation scene, calculating the coordinates of the key points by an algorithm and comparing the coordinates with real coordinates in a simulation environment is as follows:
in a simulation scene, as shown in the schematic diagram of the simulation scene in the embodiment of fig. 1, a certain number of key points are determined according to the space of the simulation scene and the positions of obstacles. All key points are selected only among the following three types, as shown in fig. 2:
1. the intersection points or points on the intersection lines of the two wall bodies;
2. the intersection point or the point on the intersection line of the two right-angle outer planes of the rectangular furniture or the object;
3. the center or point on the circumference of a circular piece of furniture or object.
Keypoints are marked in the scene. The labeling method is as follows: for example, key points located on the intersecting lines of two walls, the walls within a certain range (for example, within 10 cm) near the intersection point are marked, as shown in fig. 2 by the area formed by the two intersecting lines of the point 1 and the thick line of the corner. When the sensor of the robot scans the wall with the mark near the key point, the sensor sends the mark information to the algorithm together with normal data in a JSON format.
Taking a 2D laser radar as an example, the scanning data of the sensor with the mark is:
{ "Angle": -1.570796, "distance": 0, "information": 1, "angle": -1.562070, "distance": 0.665, "information": 0, … … }
The angle value in the JSON format data is the scanning angle of the current laser point, the distance value is the barrier distance value fed back by the laser point, and the information value is the information of the key point. If the scanned area is a non-key point mark area, the information value is 0, otherwise, the information value is the number of the key point. For example, 1 in the data, indicates that the current scanning point is near key point number 1.
After the robot finishes moving in the simulation scene and completes the mapping, the algorithm can take the information values (distance and angle) of all key point positions, calculate the coordinates of the key points based on the information values, then feed the calculated coordinates of the key points back to the simulation system, the simulation system obtains the distance difference between the calculated coordinates and the real coordinates of the key points in the scene, and then calculates the average value, namely the precision of the mapping algorithm.
In the process, because the scene and the key point position are randomly selected in the simulation scene by the simulation system, when the method is used for comparing different algorithms, the algorithm cannot acquire the accurate coordinates of the key point in advance, and cheating cannot be performed.
According to the invention, a family simulation scene is built in a robot simulation system, and 12 key points are set in the simulation scene, as shown in figure 1. Marking 12 key points and a wall or a facade in an area which is not more than 10cm nearby, and butting the robot mapping algorithm with the simulation system through a network interface to realize data intercommunication.
And installing a laser sensor for the robot, controlling the robot to move along the wall in the simulation scene, moving and simultaneously acquiring the data of the laser sensor, and synchronously transmitting the data to a mapping algorithm. The robot moves for two circles in the scene, and the robot stops moving after the laser scans all walls and objects.
The mapping algorithm completes mapping according to the received laser data, and sends the coordinate values of 12 key points to the simulation system, the simulation system calculates the distance difference between the 12 coordinate values fed back by the algorithm and the real coordinate values in the scene, and then the averaging step is as follows:
calculating the distance difference between the algorithm coordinate and the actual coordinate of each key point according to the following formula (1):
the final mapping accuracy result is calculated according to the following formula (2):
in the formula: the Xi point is the real coordinate value of the X axis of the point i; the Yi point is a Y-axis real coordinate value of the point i; the Xi' point is the coordinate value of the X-axis map of the point i; yi is a Y-axis map coordinate value of the point i; and E is the drawing establishing precision.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (3)
1. A robot mapping algorithm precision simulation evaluation method is characterized by comprising the following steps:
1) the method comprises the steps of building a robot simulation environment which comprises a simulation robot, a simulation sensor and a simulation scene, and butting a robot mapping algorithm with a simulation system;
2) randomly selecting a plurality of key points in a simulation scene as comparison points;
3) the robot in the simulation system moves according to the output information of the mapping algorithm, the simulation system provides information of various sensors for the mapping algorithm, and the mapping algorithm performs motion mapping according to the received information of the sensors and the algorithm;
4) and extracting the position coordinates of the corresponding key points in the algorithm after the graph is built by the graph building algorithm, feeding the coordinates back to the simulation system, comparing the received coordinate values of each key point in the graph building algorithm with the real values by the simulation system, and taking the average value as the graph building accuracy value of the graph building algorithm.
2. The robot mapping algorithm precision simulation evaluation method according to claim 1, wherein the key points in step 2) are determined according to the simulation scene space and the position of the obstacle, all the key points are randomly selected from the following three types,
the first kind is: the intersection points or points on the intersection lines of the two wall bodies;
the second kind is: the intersection point or the point on the intersection line of the two right-angle outer planes of the rectangular barrier;
the third category: the center of a circle or a point on the circumference of a circular obstacle.
3. The robot mapping algorithm precision simulation evaluation method according to claim 1 or 2, characterized in that the mapping precision value calculation method of the mapping algorithm in the step 4) is as follows:
calculating the distance difference between the algorithm coordinate and the actual coordinate of each key point according to the following formula:
in the formula: the Xi point is the real coordinate value of the X axis of the point i; the Yi point is a Y-axis real coordinate value of the point i; the Xi' point is the coordinate value of the X-axis map of the point i; yi is a Y-axis map coordinate value of the point i;
the map-building precision value E of the map-building algorithm is calculated according to the following formula:
in the formula: and N is the total number of the randomly selected key points.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110757015.2A CN113688496B (en) | 2021-07-05 | 2021-07-05 | Precision simulation evaluation method for robot mapping algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110757015.2A CN113688496B (en) | 2021-07-05 | 2021-07-05 | Precision simulation evaluation method for robot mapping algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113688496A true CN113688496A (en) | 2021-11-23 |
CN113688496B CN113688496B (en) | 2024-04-12 |
Family
ID=78576651
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110757015.2A Active CN113688496B (en) | 2021-07-05 | 2021-07-05 | Precision simulation evaluation method for robot mapping algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113688496B (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110262283A (en) * | 2019-06-11 | 2019-09-20 | 远形时空科技(北京)有限公司 | A kind of the vision robot's emulation platform and method of more scenes |
CN110587606A (en) * | 2019-09-18 | 2019-12-20 | 中国人民解放军国防科技大学 | Open scene-oriented multi-robot autonomous collaborative search and rescue method |
CN111496848A (en) * | 2020-03-19 | 2020-08-07 | 中山大学 | Mobile robot repeated positioning precision test based on Euclidean distance |
WO2021016807A1 (en) * | 2019-07-29 | 2021-02-04 | 西门子股份公司 | Context awareness device simulation method, device, and system |
CN112947081A (en) * | 2021-02-05 | 2021-06-11 | 浙江大学 | Distributed reinforcement learning social navigation method based on image hidden variable probability model |
-
2021
- 2021-07-05 CN CN202110757015.2A patent/CN113688496B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110262283A (en) * | 2019-06-11 | 2019-09-20 | 远形时空科技(北京)有限公司 | A kind of the vision robot's emulation platform and method of more scenes |
WO2021016807A1 (en) * | 2019-07-29 | 2021-02-04 | 西门子股份公司 | Context awareness device simulation method, device, and system |
CN110587606A (en) * | 2019-09-18 | 2019-12-20 | 中国人民解放军国防科技大学 | Open scene-oriented multi-robot autonomous collaborative search and rescue method |
CN111496848A (en) * | 2020-03-19 | 2020-08-07 | 中山大学 | Mobile robot repeated positioning precision test based on Euclidean distance |
CN112947081A (en) * | 2021-02-05 | 2021-06-11 | 浙江大学 | Distributed reinforcement learning social navigation method based on image hidden variable probability model |
Non-Patent Citations (1)
Title |
---|
熊有伦等编著: "机器人学:建模、控制与视觉 第2版", 华中科技大学出版社, pages: 402 - 404 * |
Also Published As
Publication number | Publication date |
---|---|
CN113688496B (en) | 2024-04-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Becker et al. | Reliable navigation using landmarks | |
US20190202067A1 (en) | Method and device for localizing robot and robot | |
CN111149072A (en) | Magnetometer for robot navigation | |
CN111522022B (en) | Dynamic target detection method of robot based on laser radar | |
CN109917790A (en) | It is a kind of independently to guide vehicle and its travel control method and control device | |
Weber et al. | APR-global scan matching using anchor point relationships | |
Zhou | A closed-form algorithm for the least-squares trilateration problem | |
CN108801259A (en) | A kind of indoor navigation method, device and mobile terminal | |
CN113688496B (en) | Precision simulation evaluation method for robot mapping algorithm | |
CN109960260A (en) | A kind of independently guiding vehicle and its air navigation aid and control device | |
Zunino | Simultaneous localization and mapping for navigation in realistic environments | |
Shaffer et al. | Comparison of two range-based pose estimators for a mobile robot | |
CN111595328A (en) | Real obstacle map construction and navigation method and system based on depth camera | |
US20230056652A1 (en) | Robotic Grasping Via RF-Visual Sensing And Learning | |
Jae-Bok | Mobile robot localization using range sensors: Consecutive scanning and cooperative scanning | |
Jung et al. | Structured light 2D range finder for simultaneous localization and map-building (SLAM) in home environments | |
CN114995459A (en) | Robot control method, device, equipment and storage medium | |
Rupp et al. | Optimized landmark arrangement for absolute localization-a practical approach | |
Kita et al. | 3D map building using mobile robot with scan device | |
Almanza-Ojeda et al. | Occupancy map construction for indoor robot navigation | |
Chen et al. | A simple and generic method for real-time accurate hand-eye calibration using a 3D vision sensor | |
CN112731434A (en) | Positioning method and system based on laser radar and marker | |
Sandberg et al. | A robot localization method based on laser scan matching | |
Miah et al. | Mobile robot navigation using direction-sensitive RFID reader | |
Xiaowei et al. | The localization of mobile robot based on laser scanner |
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 |