CN115328163B - Speed and precision optimization method for inspection robot radar odometer - Google Patents
Speed and precision optimization method for inspection robot radar odometer Download PDFInfo
- Publication number
- CN115328163B CN115328163B CN202211125540.3A CN202211125540A CN115328163B CN 115328163 B CN115328163 B CN 115328163B CN 202211125540 A CN202211125540 A CN 202211125540A CN 115328163 B CN115328163 B CN 115328163B
- Authority
- CN
- China
- Prior art keywords
- point
- odom
- yaw
- lidar
- dimensional
- 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.)
- Active
Links
- 238000007689 inspection Methods 0.000 title claims abstract description 31
- 238000005457 optimization Methods 0.000 title claims abstract description 20
- 238000000034 method Methods 0.000 title claims abstract description 12
- 238000013519 translation Methods 0.000 claims abstract description 21
- 238000006243 chemical reaction Methods 0.000 claims abstract description 10
- 238000012216 screening Methods 0.000 claims abstract description 4
- 238000006073 displacement reaction Methods 0.000 claims description 18
- 230000009466 transformation Effects 0.000 claims description 17
- 238000005259 measurement Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000009434 installation Methods 0.000 claims 1
- 230000006870 function Effects 0.000 description 5
- 238000013507 mapping Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 230000001360 synchronised effect Effects 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
- G05D1/0238—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using obstacle or wall sensors
- G05D1/024—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using obstacle or wall sensors in combination with a laser
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0276—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle
-
- 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
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Optics & Photonics (AREA)
- Electromagnetism (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a speed and precision optimization method of a radar odometer of an inspection robot, which specifically comprises the following steps: collecting point cloud data by using a three-dimensional laser radar, calculating curvature, and screening characteristic points according to the curvature; performing pose conversion of two frames of point clouds, and performing line feature association and surface feature association; convex optimization of the pose translation part is carried out by using a Ceres optimizer, three-dimensional radar data are output, three-dimensional radar odometer data are updated, and then interpolation is carried out. The invention improves the accuracy of the smooth degree of motion control and the matching of the sensor value, and simultaneously improves the efficiency and the accuracy of positioning.
Description
Technical Field
The invention belongs to the technical field of space environment perception, and particularly relates to a speed and precision optimization method for a radar odometer of an inspection robot.
Background
The vehicle bottom inspection robot is used for replacing manpower, carrying out data acquisition on vehicle bottom parts of a vehicle in a maintenance trench of a parking lot or a vehicle section, analyzing through an artificial intelligence algorithm, and automatically giving vehicle maintenance suggestions so as to guarantee the operation safety of the vehicle and reduce the working strength of inspection workers. In order to ensure the accuracy of data acquisition of vehicle bottom parts and ensure that the robot does not collide with the vehicle bottom parts, the speed and the accuracy of positioning a data source in the overhaul ditch of the robot are very important.
The synchronous positioning and mapping technology based on the laser radar is the most common positioning method in the robot field. The LOAM (Lidar Odometry and Mapping) proposed by Ji Zhang in 2014 is a classic three-dimensional radar odometer, and the algorithm divides synchronous positioning and Mapping into different output frequencies so as to improve efficiency and precision. However, due to the existing load-adaptive algorithm based on pure three-dimensional radar point cloud data, the error in the six-degree-of-freedom motion estimation reaches more than 2.5%, and the three-dimensional radar performance is limited by the point cloud acquisition speed and the point cloud input speed, and the output frequency of the motion estimation is usually about 10Hz, which is very unfavorable for the motion control and sensor data reference of the inspection robot based on the positioning data source.
Therefore, it is necessary to provide a wheel speed meter coupled high-frequency high-precision radar odometer solution for the positioning requirement of the vehicle bottom inspection robot under special scenes such as a maintenance trench based on the existing LOAM algorithm.
Disclosure of Invention
In order to achieve the purpose, the invention provides a speed and precision optimization method of a radar odometer of an inspection robot.
The invention discloses a speed and precision optimization method of a radar odometer of an inspection robot, which comprises the following steps of firstly installing a three-dimensional laser radar, a motor wheel speed meter and an inertia measurement unit IMU on the robot:
step 1: scanning by using a three-dimensional laser radar loaded on a robot to obtain laser point cloud data in the inspection trench, wherein the three-dimensional laser radar is at t k And (4) scanning and combining the point cloud data at the moment and recording as p.
Step 2: calculating the curvature: calculating the curvature c of each 5 points before and after the current point and the current point, wherein the calculation formula is as follows:
wherein X represents a point on a single scan line of the three-dimensional laser radar, X i Denotes the ith point and | X | denotes the mode of the vector, i.e. the distance of this point to the origin of the three-dimensional lidar coordinate system.
And step 3: screening characteristic points: characteristic points are screened according to the curvature, and are divided into two types: the points with large curvature are line characteristic points, the points with small curvature are surface characteristic points, and the number of the characteristic points is limited, so that the line characteristic points and the surface characteristic points are extracted in the whole three-dimensional space to replace the whole point cloud data.
And 4, step 4: two-frame point cloud pose conversion: the relative pose of the (k + 1) th frame and the (k) th frame is as follows:
wherein R is a rotation matrix and t is a translation vector; point p in the k +1 th frame i Conversion to the kth frame coordinate system: />
/>
Rotational motion estimation R within two frame point cloud sampling time provided using inertial measurement unit IMU IMU As a result R of the point cloud translation transformation.
Translation transformation t in pose transformation using wheel speed meter odom The iterative initial value of the translation vector t in the relative pose for optimization, which is the iterative initial value of the point cloud translation transformation, is given by the transformation of the motor wheel speed meter in this time.
And 5: and (3) feature association:
step 5.1: line feature association: when p is i When the line characteristic point is found, searching for the departure in the last frame
The nearest line characteristic point is marked as a point a, and a line characteristic point is found on the adjacent line and is marked as a point b, the two points form a straight line, and the residual function of the straight line is the distance from the point to the straight line:
wherein p is a 、p b Respectively representing the coordinates of the points a and b in the local coordinate system, and x is the cross product operation of the vector.
Line feature residual Jacobian:
for the inspection groundThe robot running in a straight line in the ditch only needs Jacobian of translational motion, namely only estimates translational motion t and does not estimate angle change, so that a rotation term does not need to be considered, and therefore, the second term on the right of equal sign has
The required angle estimation data R is given by the inertial measurement unit IMU.
And step 5.2: and (3) associating the surface features: when p is i When the feature point is a face feature point, the search is performed in the previous frame
The nearest surface feature point is marked as a point m, two surface feature points are found on an adjacent line and are respectively marked as j and l, the three points form a plane, and a residual function of the plane is the distance from the point to the surface:
wherein p is j 、p l 、p m Respectively representing the coordinates of points j, l, m in the local coordinate system.
Surface feature residual Jacobian:
similarly, the second term on the right of the equal sign has
The rotation term R need not be considered.
Step 6: pose optimization: and combining the solving formulas, and using a Ceres optimizer to perform convex optimization on the pose translation part.
And 7: outputting three-dimensional radar data: and matching every 10 frames of the obtained point cloud data with a feature map, and adding the features of each frame into the feature map, wherein the output frequency of the feature map is 1Hz, and the pose estimation output frequency is 10Hz.
And 8: updating three-dimensional radar odometer data: if the three-dimensional radar odometer issued information is received, updating the three-dimensional radar odometer data (x) lidar ,y lidar ,yaw lidar ) Wherein x is lidar 、y lidar 、yaw lidar Representing three-freedom motion estimation data of x-direction displacement, y-direction displacement and yaw angle of the three-dimensional radar odometer respectively, and taking the pose (x-direction displacement, y-direction displacement and yaw angle) of the motor wheel speed meter at the current time odom ，y odom ，yaw odom ) Recorded as the output interpolation reference (x) odombase ,y odombase ,yaw odombase ) Wherein x is odom 、y odom 、yaw odom The three-degree-of-freedom motion estimation data of x-direction displacement, y-direction displacement and yaw angle, which are issued by a motor wheel speed meter, are respectively, and the updating frequency in the step is 10Hz.
And step 9: interpolation is carried out, the reliability of the wheel speed odometer in the output interval of the two-frame three-dimensional laser radar odometer is fully acquired in the flat linear channel, and when the information issued by the wheel speed odometer of the motor is received, (x) is received odom ，y odom ，yaw odom ) And (x) odombase ，x odombase ，yaw odombase ) And (3) obtaining a conversion relation of the wheel speed meter by taking the difference, wherein the calculation formula is as follows:
wherein k is dir Representing forward or backward movement, the forward movement having a value of 1 and the backward movement having a value of-1; calculating the running distance delta d, and then issuing and updating the pose x lidar ：x lidar ＝x lidar +Δd·cos(yaw lidar ) The updating frequency of the step is 50Hz, and when the information issued by the three-dimensional radar odometer is received again, the position (x) of the motor wheel speed meter at the moment is calculated odom ，y odom ，yaw odom ) Marked as next interpolation (x) odombase ，y odombase ，yaw odombase )。
The beneficial technical effects of the invention are as follows:
the invention uses the improved LOAM algorithm for the robot with single degree of freedom, and improves the problems of low output frequency of the LOAM algorithm and low precision under six-degree-of-freedom motion estimation. By inspection, the issuing frequency of the three-dimensional laser radar odometer is increased from about 10Hz to about 50Hz after the scheme is applied, the precision of the smooth degree of motion control and the numerical value matching of the sensor is improved, the comprehensive positioning error can be as low as +/-0.0075 m, namely 0.0075% in the navigation of 100m in the routing inspection trench, and the positioning precision is optimized.
Drawings
FIG. 1 is a schematic representation of the operation of a robot;
reference numerals: 1-an Inertia Measurement Unit (IMU), 2-routing inspection of trench walls, 3-a three-dimensional laser radar scanning area schematic diagram, 4-a three-dimensional laser radar and 5-a robot body;
FIG. 2 is a general technical flow diagram of the present invention;
FIG. 3 is a flow diagram of a feature-based odometer implementation.
Detailed Description
The invention is described in further detail below with reference to the figures and the detailed description.
The inspection robot at the bottom of the vehicle only has single degree of freedom of linear motion back and forth in a narrow inspection trench. Because most of the interior of the inspection trench is provided with repeated characteristics of concrete columns, wall surfaces and the like, the probability of degradation of a positioning algorithm based on the traditional two-dimensional radar is very high, so that the two-dimensional radar cannot be used for accurately positioning in the inspection trench, the three-dimensional radar is used for positioning, the positioning accuracy is improved, the LOAM algorithm is a commonly used synchronous positioning and mapping algorithm, but the existing LOAM algorithm has the problems of low output frequency and low accuracy under six-degree-of-freedom motion estimation, and the method is combined with the characteristic that an inspection trench inspection robot only runs linearly in the inspection trench, so that the characteristic of positioning on a single shaft only needs to be considered, and the method for optimizing the speed and the accuracy of the radar odometer of the inspection robot is provided.
The invention relates to a differential drive wheel type inspection robot automatically running in a linear inspection trench, which is combined with a figure 1, wherein a robot 5 runs in a flat linear trench by adopting differential drive and is provided with: the robot comprises a three-dimensional laser radar 4, a wheel type encoder, an Inertia Measurement Unit (IMU) 1 and the like, and is based on an ROS system, wherein the robot and the wheel type encoder are automatically positioned and navigated in a patrol trench through the three-dimensional laser radar. 1) The robot moves linearly in the inspection trench, and only an LOAM algorithm is used for one degree of freedom; 2) When two frames of radar point clouds are matched, the pose transformation of the motor wheel speed meter is used as the optimized iterative initial value constraint of the point cloud pose transformation so as to improve the positioning accuracy; 3) The output frequency is improved by interpolation of the motor wheel speed meter, the smooth control degree is improved, and the matching precision of sensor data and the position and posture of the odometer is improved.
The invention discloses a speed and precision optimization method of a radar odometer of an inspection robot, which is shown in figures 2 and 3 and specifically comprises the following steps:
step 1: scanning by using a three-dimensional laser radar loaded on a robot to obtain laser point cloud data in the inspection trench, wherein the three-dimensional laser radar is at t k And point cloud data combined by scanning at all times is recorded as p.
And 2, step: calculating the curvature: and calculating the curvature c of each of 5 points before and after the current point and the current point, wherein the calculation formula is as follows:
wherein X represents a point on a single scan line of the three-dimensional laser radar, X i Represents the ith point and | X | represents the mode of the vector, i.e. the distance of this point to the origin of the three-dimensional lidar coordinate system.
And step 3: screening characteristic points: characteristic points are screened according to the curvature, and are divided into two types: the points with large curvature are line characteristic points, the points with small curvature are surface characteristic points, and the number of the characteristic points is limited, so that the line characteristic points and the surface characteristic points are extracted in the whole three-dimensional space to replace the whole point cloud data, and the data volume needing to be processed in the later period is reduced.
And 4, step 4: two-frame point cloud pose conversion: relative pose of the k +1 th frame and the k frameComprises the following steps:
wherein R is a rotation matrix and t is a translation vector; point p in the k +1 th frame i Conversion to the kth frame coordinate system: />
Considering that the original LOAM algorithm has a large error in the motion estimation of the rotation, when the IMU has a magnetometer function, the rotation motion estimation R in the two-frame point cloud sampling time provided by the inertial measurement unit IMU is used IMU As a result R of point cloud translation transformation, the precision of angle estimation is improved.
During the scanning movement of the three-dimensional laser radar, the robot does not move at a completely uniform speed, so the translation transformation todom in the pose transformation of the wheel speed meter is used as the iteration initial value of the point cloud translation transformation in the time period, namely the iteration initial value of the translation vector t in the relative pose for optimization is given by the transformation of the motor wheel speed meter in the time period. Compared with the constant-speed interpolation motion model of the existing LOAM algorithm, the method provides a better optimized iteration initial value and improves the precision of subsequent translational motion estimation.
And 5: and (3) feature association:
step 5.1: line feature association: when p is i When the line characteristic point is found, searching for the departure in the last frame
The nearest line characteristic point is marked as a point a, and a line characteristic point is found on the adjacent line and is marked as a point b, the two points form a straight line, and the residual function of the straight line is the distance from the point to the straight line:
wherein p is a 、p b Respectively representing the coordinates of the points a and b in the local coordinate system, and x is the cross product of the vector.
Line feature residual Jacobian:
the existing LOAM algorithm needs six-degree-of-freedom motion estimation, so that the rotation item R in the T needs to be optimized, but for a robot which patrols the linear running in a trench, only the Jacobian of translational motion is needed, namely only the translational motion T is estimated, and the angle change is not estimated, so that the rotation item does not need to be considered, and therefore, the second item on the right of equal sign has the rotation item
The required angle estimation data R is given by the inertial measurement unit IMU. Compared with the existing LOAM algorithm, the optimization reduces the data amount required to be processed and greatly improves the positioning precision.
Step 5.2: and (3) associating the surface features: when p is i When the feature point is a face feature point, the search is performed in the previous frame
The nearest surface feature point is marked as a point m, two surface feature points are found on an adjacent line and are respectively marked as j and l, the three points form a plane, and a residual function of the plane is the distance from the point to the surface:
wherein p is j 、p l 、p m Respectively representing the coordinates of points j, l, m in the local coordinate system.
Surface feature residual Jacobian:
similarly, the second term on the right of the equal sign has
The rotation term R need not be considered.
And 6: pose optimization: and combining the solving formulas, and using a Ceres optimizer to perform convex optimization on the pose translation part.
And 7: outputting three-dimensional radar data: with reference to fig. 3, each 10 frames of the obtained point cloud data are matched with a feature map, and the features of each frame are added into the feature map, wherein the output frequency of the feature map is 1Hz, and the pose estimation output frequency is 10Hz.
And 8: updating three-dimensional radar odometer data: if the three-dimensional radar odometer issued information is received, updating the three-dimensional radar odometer data (x) lidar ,y lidar ,yaw lidar ) Wherein x is lidar 、y lidar 、yaw lidar Representing three-freedom motion estimation data of x-direction displacement, y-direction displacement and yaw angle of the three-dimensional radar odometer respectively, and taking the pose (x-direction displacement, y-direction displacement and yaw angle) of the motor wheel speed meter at the current time odom ，y odom ，yaw odom ) Recorded as the output interpolation reference (x) odombase ,y odombase ,yaw odombase ) Wherein x is odom 、y odom 、yaw odom The three-degree-of-freedom motion estimation data of x-direction displacement, y-direction displacement and yaw angle, which are issued by a motor wheel speed meter, are respectively, and the updating frequency in the step is 10Hz.
And step 9: interpolation is carried out, the reliability of the wheel speed odometer in the output interval of the two-frame three-dimensional laser radar odometer is fully acquired in the flat linear channel, and when the information issued by the wheel speed odometer of the motor is received, (x) is received odom ，y odom ，xaw odom ) And (x) odombase ，y odombase ，yaw odombase ) And (3) obtaining a conversion relation of the wheel speed meter by taking the difference, wherein the calculation formula is as follows:
wherein k is dir Indicating forward or backward movement, forward movement having a value of 1, backward movementThe dynamic value is-1; calculating the running distance delta d, and then issuing the updated pose x lidar ：x lidar ＝x lidar +Δd·cos(yaw lidar ) The updating frequency of the step is 50Hz, and when the information issued by the three-dimensional radar odometer is received again, the position (x) of the motor wheel speed meter at the moment is calculated odom ，y odom ，yaw odom ) Marked as next interpolation (x) odombase ，y odombase ，yaw odombase )。
Compared with the prior art, the invention improves the problems of low output frequency of the LOAM algorithm and low precision under six-degree-of-freedom motion estimation by using the improved LOAM algorithm for the robot with single degree of freedom. By inspection, the issuing frequency of the three-dimensional laser radar odometer is increased from about 10Hz to about 50Hz after the scheme is applied, the precision of the smooth degree of motion control and the numerical value matching of the sensor is improved, the comprehensive positioning error can be as low as +/-0.0075 m, namely 0.0075% in the navigation of 100m in the routing inspection trench, and the positioning precision is optimized.
Claims (1)
1. The utility model provides a speed and precision optimization method of patrolling and examining robot radar odometer which characterized in that, three-dimensional laser radar (4), motor wheel speedometer and inertia measurement unit IMU (1) of installation on robot (5), the optimization step specifically is:
step 1: scanning by using a three-dimensional laser radar loaded on the robot to acquire laser point cloud data in the inspection trench, wherein the three-dimensional laser radar is arranged at t k Scanning and combining the point cloud data at any moment and recording the point cloud data as p;
step 2: calculating the curvature: calculating the curvature c of each 5 points before and after the current point and the current point, wherein the calculation formula is as follows:
wherein X represents a point on a single scan line of the three-dimensional laser radar, X i Represents the ith point, | X | represents the mode of the vector, i.e. the distance from this point to the origin of the three-dimensional lidar coordinate system；
And step 3: screening characteristic points: characteristic points are screened according to the curvature, and are divided into two types: the points with large curvature are line characteristic points, the points with small curvature are surface characteristic points, and the number of the characteristic points is limited, so that the line characteristic points and the surface characteristic points are extracted in the whole three-dimensional space to replace the whole point cloud data;
and 4, step 4: two-frame point cloud pose conversion: the relative pose of the (k + 1) th frame and the (k) th frame is as follows:
wherein R is a rotation matrix and t is a translation vector; point p in the k +1 th frame i Conversion to the kth frame coordinate system: />
Rotational motion estimation R within two frame point cloud sampling time provided using inertial measurement unit IMU IMU As a result of the point cloud translation transformation R;
translation transformation t in pose transformation using wheel speed meter odom As an iteration initial value of the point cloud translation transformation, namely an iteration initial value of a translation vector t in the relative pose for optimization is given by the transformation of the motor wheel speed meter in the time;
and 5: and (3) feature association:
step 5.1: line feature association: when p is i When the line characteristic point is found, searching for the departure in the last frame
The nearest line characteristic point is marked as a point a, and a line characteristic point is found on the adjacent line and is marked as a point b, the two points form a straight line, and the residual function of the straight line is the distance from the point to the straight line:
wherein p is a 、p b Are respectively provided withRepresenting the coordinates of the points a and b in a local coordinate system, wherein x is the cross product operation of the vector;
line feature residual Jacobian:
for the robot which patrols and examines the linear running in the trench, only the Jacobian of the translation motion is needed, namely only the translation motion t is estimated, the angle change is not estimated, and therefore, the rotation item does not need to be considered, and therefore, the second item on the right of the equal sign has
The required angle estimation data R is given by an inertial measurement unit IMU;
and step 5.2: and (3) associating the surface features: when p is i When the feature point is a face feature point, the search is performed in the previous frame
The nearest surface feature point is marked as a point m, two surface feature points are found on an adjacent line and are respectively marked as j and l, the three points form a plane, and the residual function is the distance from the point to the surface:
wherein p is j 、p l 、p m Respectively representing the coordinates of the points j, l and m in a local coordinate system;
surface feature residual Jacobian:
for the same reason, the second item on the right of the equal sign has
The rotation term R does not need to be considered;
step 6: pose optimization: combining the solving formula, and using a Ceres optimizer to perform convex optimization on the pose translation part;
and 7: outputting three-dimensional radar data: matching every 10 frames of the obtained point cloud data with a feature map, and adding the features of each frame into the feature map, wherein the output frequency of the feature map is 1Hz, and the pose estimation output frequency is 10Hz;
and step 8: updating three-dimensional radar odometry data: if the three-dimensional radar odometer issuing information is received, updating the three-dimensional radar odometer data (x) lidar ，y lidar ，yaw lidar ) Wherein x is lidar 、y lidar 、yaw lidar Representing three-freedom motion estimation data of x-direction displacement, y-direction displacement and yaw angle of the three-dimensional radar odometer respectively, and taking the pose (x-direction displacement, y-direction displacement and yaw angle) of the motor wheel speed meter at the current time odom ，y odom ，yaw odom ) Recorded as the output interpolation reference (x) odombase ，y odombase ，yaw odombase ) Wherein x is odom 、y oaom 、yaw odom Three-degree-of-freedom motion estimation data of x-direction displacement, y-direction displacement and yaw angle, which are issued by a motor wheel speed meter, are respectively, and the updating frequency in the step is 10Hz;
and step 9: interpolation is carried out, the reliability of the wheel speed odometer in the output interval of the two-frame three-dimensional laser radar odometer is fully acquired in the flat linear channel, and when the information issued by the wheel speed odometer of the motor is received, (x) is received odom ，y odom ，yaw odom ) And (x) odombase ，y odombase ，yaw odombase ) And (3) obtaining a conversion relation of the wheel speed meter by taking the difference, wherein the calculation formula is as follows:
wherein k is dir Representing forward or backward movement, the forward movement having a value of 1 and the backward movement having a value of-1; the distance of travel deltad is calculated,then issuing the updated pose x lidar ：x lidar ＝x lidar +Δd·cos(yaw lidar ) The updating frequency of the step is 50Hz, and when the three-dimensional radar odometer issued information is received again, the position and the attitude (x) of the motor wheel speed meter at the moment are determined odom ，y odom ，yaw odom ) Marked as next interpolated (x) odombase ，y odombase ，yaw odombase )。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211125540.3A CN115328163B (en) | 2022-09-16 | 2022-09-16 | Speed and precision optimization method for inspection robot radar odometer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211125540.3A CN115328163B (en) | 2022-09-16 | 2022-09-16 | Speed and precision optimization method for inspection robot radar odometer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115328163A CN115328163A (en) | 2022-11-11 |
| CN115328163B true CN115328163B (en) | 2023-03-28 |
Family
ID=83930816
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211125540.3A Active CN115328163B (en) | 2022-09-16 | 2022-09-16 | Speed and precision optimization method for inspection robot radar odometer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115328163B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115540875B (en) * | 2022-11-24 | 2023-03-07 | 成都运达科技股份有限公司 | Method and system for high-precision detection and positioning of train vehicles in station track |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111929699A (en) * | 2020-07-21 | 2020-11-13 | 北京建筑大学 | Laser radar inertial navigation odometer considering dynamic obstacles and mapping method and system |
| WO2022003179A1 (en) * | 2020-07-03 | 2022-01-06 | Five AI Limited | Lidar mapping |
| CN113985429A (en) * | 2021-09-23 | 2022-01-28 | 天津大学 | Unmanned aerial vehicle environment scanning and reconstructing method based on three-dimensional laser radar |
| CN114004869A (en) * | 2021-09-18 | 2022-02-01 | 浙江工业大学 | Positioning method based on 3D point cloud registration |
| CN114526745A (en) * | 2022-02-18 | 2022-05-24 | 太原市威格传世汽车科技有限责任公司 | Drawing establishing method and system for tightly-coupled laser radar and inertial odometer |
-
2022
- 2022-09-16 CN CN202211125540.3A patent/CN115328163B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022003179A1 (en) * | 2020-07-03 | 2022-01-06 | Five AI Limited | Lidar mapping |
| CN111929699A (en) * | 2020-07-21 | 2020-11-13 | 北京建筑大学 | Laser radar inertial navigation odometer considering dynamic obstacles and mapping method and system |
| CN114004869A (en) * | 2021-09-18 | 2022-02-01 | 浙江工业大学 | Positioning method based on 3D point cloud registration |
| CN113985429A (en) * | 2021-09-23 | 2022-01-28 | 天津大学 | Unmanned aerial vehicle environment scanning and reconstructing method based on three-dimensional laser radar |
| CN114526745A (en) * | 2022-02-18 | 2022-05-24 | 太原市威格传世汽车科技有限责任公司 | Drawing establishing method and system for tightly-coupled laser radar and inertial odometer |
Non-Patent Citations (2)
| Title |
|---|
| 柯艳国 ; 王雄奇 ; 魏新 ; 吴贤斌 ; 朱仲贤 ; 董翔宇 ; 黄杰 ; 董二宝 ; .一种基于图优化的实时3D激光SLAM算法.机电一体化.2020,(Z1),全文. * |
| 殷江 ; 林建德 ; 孔令华 ; 邹诚 ; 游通飞 ; 易定容 ; .基于激光雷达的移动机器人三维建图与定位.福建工程学院学报.2020,(04),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115328163A (en) | 2022-11-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109709801B (en) | Indoor unmanned aerial vehicle positioning system and method based on laser radar | |
| CN112083725B (en) | Structure-shared multi-sensor fusion positioning system for automatic driving vehicle | |
| CN112083726B (en) | Park-oriented automatic driving double-filter fusion positioning system | |
| Zhao et al. | A robust laser-inertial odometry and mapping method for large-scale highway environments | |
| CN114526745B (en) | Drawing construction method and system for tightly coupled laser radar and inertial odometer | |
| CN114018248B (en) | Mileage metering method and image building method integrating code wheel and laser radar | |
| CN114001733B (en) | Map-based consistent efficient visual inertial positioning algorithm | |
| CN115328163B (en) | Speed and precision optimization method for inspection robot radar odometer | |
| CN115479598A (en) | Positioning and mapping method based on multi-sensor fusion and tight coupling system | |
| CN112781594A (en) | Laser radar iteration closest point improvement algorithm based on IMU coupling | |
| CN115272596A (en) | Multi-sensor fusion SLAM method oriented to monotonous texture-free large scene | |
| CN116929338B (en) | Map construction method, device and storage medium | |
| CN115540850A (en) | Unmanned vehicle mapping method combining laser radar and acceleration sensor | |
| CN116429116A (en) | Robot positioning method and equipment | |
| CN113129377B (en) | Three-dimensional laser radar rapid robust SLAM method and device | |
| CN114067210A (en) | Mobile robot intelligent grabbing method based on monocular vision guidance | |
| CN115046543A (en) | Multi-sensor-based integrated navigation method and system | |
| CN112945233A (en) | Global drift-free autonomous robot simultaneous positioning and map building method | |
| CN116124161B (en) | LiDAR/IMU fusion positioning method based on priori map | |
| WO2021109166A1 (en) | Three-dimensional laser positioning method and system | |
| CN114543786B (en) | Wall climbing robot positioning method based on visual inertial odometer | |
| Wang et al. | SW-LIO: A Sliding Window Based Tightly Coupled LiDAR-Inertial Odometry | |
| CN112947481B (en) | Autonomous positioning control method for home service robot | |
| CN114638902B (en) | On-line estimation method for external parameters of vehicle-mounted camera | |
| Yuan et al. | LIWO: LiDAR-Inertial-Wheel Odometry |
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 |