Concrete floor edge finishing robot and its finishing trajectory planning method

Abstract

This application relates to the field of construction robot technology, specifically to a concrete floor edge finishing robot and its finishing trajectory planning method, including a tracked mobile chassis, a multi-degree-of-freedom robotic arm, a plate-type finishing actuator, a machine vision perception module, and control components for realizing finishing trajectory planning; the tracked mobile chassis is used to move the entire machine to the edge area to be finished and to stop stably, providing a basis for finishing trajectory execution; the machine vision perception module is used to collect image information of the edge area to be finished, identify the target area, and output the target coordinates or target pose, providing position parameters for finishing trajectory planning; the multi-degree-of-freedom robotic arm is used to drive the plate-type finishing actuator to move according to the planned finishing trajectory, and the plate-type finishing actuator is used to perform edge finishing action along the planned trajectory; the control components include an industrial computer, a robotic arm controller, and an AGV chassis controller.

Description

Technical Field

[0001] This application relates to the field of construction robot technology, specifically to a concrete floor edge finishing robot and its finishing trajectory planning method. Background Technology

[0002] Concrete floor construction is a crucial part of building projects, and the quality of the floor finishing directly affects the flatness, wear resistance, and service life of the surface. Traditional floor finishing mainly relies on manual operation of power trowels, which is inefficient, labor-intensive, and the finishing quality is greatly affected by the operator's experience. With the development of construction automation technology, large-area floor finishing has gradually become mechanized and automated, but finishing edge areas remains a challenge for automated construction.

[0003] In building construction, the edge areas of the floor typically include wall edges, column bases, construction joint edges, and narrow passageways. These areas have limited space, making it difficult for traditional large-scale finishing equipment to operate effectively. Current technology still primarily relies on manual labor for edge finishing, which presents the following technical problems: First, workers must bend over for extended periods in confined spaces, easily leading to occupational health issues; second, manual labor is inefficient, and edge finishing becomes a bottleneck restricting the overall construction progress when constructing large-area floors; third, inconsistent manual operation results in varying finishing quality from different operators or at different times; and fourth, the accuracy of manual finishing is difficult to guarantee, often resulting in unsatisfactory finishing quality in edge areas, affecting the overall quality of the floor.

[0004] In recent years, scholars and enterprises both domestically and internationally have conducted extensive research in the field of construction robots, resulting in the emergence of construction robot products such as floor finishing robots and wall plastering robots. However, existing construction robot products are mainly designed for the operational needs of large, standard areas, lacking effective solutions for unstructured, space-constrained working environments such as edge areas. Specifically, existing floor finishing robots have the following shortcomings: First, their mobile chassis is too large to enter narrow edge working areas; second, they lack dedicated end effector designs for finishing operations in edge areas; third, the finishing trajectory planning methods do not consider the special characteristics of edge areas, and the planning accuracy and operational efficiency need to be improved; fourth, they lack effective edge recognition and positioning technology, making it difficult for them to autonomously perceive edge positions and boundary information. Summary of the Invention

[0005] A robot for collecting light at the edge of a concrete floor and its light collection trajectory planning method include a tracked mobile chassis, a multi-degree-of-freedom robotic arm, a plate-type light collection actuator, a machine vision perception module, and control components for realizing light collection trajectory planning.

[0006] The tracked mobile chassis is used to move the whole machine to the edge area to be collected light and stop stably, providing a basis for the execution of the light collection trajectory;

[0007] The machine vision perception module is used to collect image information of the edge area to be collected, identify the target area and output the target coordinates or target pose, providing position parameters for the planning of the light collection trajectory; the multi-degree-of-freedom robotic arm is used to drive the plate-type light collection actuator to move according to the planned light collection trajectory, and the plate-type light collection actuator is used to perform edge-following light collection action along the planned trajectory;

[0008] The control components include an industrial computer, a robotic arm controller, and an AGV chassis controller. The industrial computer is used to receive the output information of the machine vision perception module and complete the light-receiving trajectory planning. The robotic arm controller is used to control the multi-degree-of-freedom robotic arm to move according to the planned trajectory. The AGV chassis controller is used to control the tracked mobile chassis to complete the docking and positioning.

[0009] Preferably, the tracked mobile chassis is provided with an installation platform at its top, and a support column is provided on one side of the upper end of the installation platform. The robotic arm base is installed on the upper end of the support column, and the multi-degree-of-freedom robotic arm is installed on the robotic arm base. The multi-degree-of-freedom robotic arm can adjust its posture in multiple directions according to the planned light-collecting trajectory, so that the plate-type light-collecting actuator can complete the light-collecting action along the target trajectory, which is suitable for the trajectory execution requirements of complex edge areas such as walls and column bases.

[0010] Preferably, the tracked mobile chassis includes a frame, a left tracked walking mechanism, a right tracked walking mechanism, a differential drive mechanism, and a battery pack. The differential drive mechanism includes a left drive module and a right drive module, which are respectively connected to the left and right tracked walking mechanisms. This allows for precise adjustment of the docking position, ensuring that the movement trajectory of the multi-degree-of-freedom robotic arm covers the edge area to be collected, thus preventing trajectory deviation. The battery pack is located in the lower part of the tracked mobile chassis to power the entire machine, while also lowering the center of gravity and improving the chassis's docking stability. This provides a reliable foundation for trajectory execution. By adjusting the output speeds of the left and right drive modules, the tracked mobile chassis can reach the predetermined docking position and provide a stable base for the subsequent precise collection of light by the robotic arm.

[0011] Preferably, the plate-type light-collecting actuator includes a flat plate and a plate mounting base. The plate mounting base is used to fix the flat plate to the end of the multi-degree-of-freedom robotic arm. The flat plate has a flat structure, which can adapt to the planned reciprocating light-collecting trajectory, fit the edge area to be collected, and complete uniform light collection, ensuring the effectiveness of trajectory execution and making up for the edge blind spots of traditional rotary plate.

[0012] Preferably, the machine vision perception module includes a vision sensor and a sensor bracket. The sensor bracket is used to fix the vision sensor and adjust its field of view direction to ensure that the vision sensor can accurately acquire image information of the edge area to be received. The vision sensor adopts an industrial camera or a depth camera to transmit the acquired image information to an industrial control computer, which performs target recognition, edge extraction, and coordinate calculation to provide reliable data support for the accurate planning of the light-receiving trajectory. The control component also includes a chassis drive module, a power supply module, and a communication module, which are integrated into a sheet metal control box on the other side of the upper part of the mounting platform. A cooling fan is installed on the outside of the sheet metal control box to manage the heat of the control component and ensure the stable operation of trajectory planning and execution.

[0013] The method for planning the trajectory of a robot for edge finishing of concrete floors includes the following steps:

[0014] Step 1: Receive the movement command sent by the industrial computer through the AGV chassis controller, control the differential drive mechanism to drive the tracked mobile chassis to move to the vicinity of the edge area to be collected and complete the docking;

[0015] Step 2: The machine vision perception module collects image information of the edge area of ​​the light to be collected through the vision sensor and transmits it to the industrial control computer. The industrial control computer performs edge extraction, target recognition and coordinate calculation on the image to obtain the homogeneous coordinates of the target point in the visual coordinate system. Then, it transforms the target to the coordinate system of the robot arm base through the homogeneous transformation matrix to obtain the target coordinates or target pose.

[0016] Step 3: Generate the execution trajectory of the multi-degree-of-freedom robotic arm end effector based on the target coordinates or target pose;

[0017] Step 4: Smooth the end effector trajectory of the multi-degree-of-freedom robotic arm using fifth-order polynomial trajectory planning, trajectory parameters. Within the time interval [0, T], it satisfies Corresponding speed acceleration By constraining the starting position, ending position, starting and ending speeds and starting and ending accelerations, the polynomial coefficients are solved to control the multi-degree-of-freedom robotic arm to drive the plate-type polishing actuator to move, so that the plate-type polishing plate performs reciprocating polishing along the edge area.

[0018] Step 5: Re-inspect the light collection results and perform compensatory light collection for areas that are not covered or have insufficient light collection quality.

[0019] Preferably, in step one, the linear velocity of the tracked mobile chassis... and angular velocity satisfy:

[0020]

[0021]

[0022] in, The linear velocity of the left track traveling mechanism; The linear velocity of the right track traveling mechanism; The distance between the center points of the left and right tracks is given by the following kinematic equation in the ground coordinate system: ;in, These are the position coordinates of the tracked mobile chassis in the ground coordinate system. This is the chassis heading angle.

[0023] Preferably, in step two, the homogeneous coordinates of the target point in the visual sensor coordinate system are: c p= Its homogeneous coordinates in the robot arm base coordinate system are: b P The homogeneous transformation matrix is b T c = ;

[0024] in, , , This represents the three-dimensional position coordinates of the target point in the camera coordinate system; 1 indicates the homogeneous coordinate form, used for multiplication with the 4×4 transformation matrix; b T c Represents the homogeneous transformation matrix from the camera coordinate system to the robot arm base coordinate system; This represents a 3×3 rotation matrix that describes the camera's orientation relative to the robotic arm base. This represents a 3×1 translation vector, describing the position of the camera relative to the robotic arm base.

[0025] Preferably, in step three, the target path curve of the edge region is set as follows: ,in, For path parameters;

[0026] The ideal end position of a plate-type trowel is ,in, For the board-type trowel at the path point The pose matrix at the location, the pose satisfies , Let be the normal vector of the slab-type trowel. The joint variable vector is obtained by solving the inverse kinematics problem, where the normal vector is the local plane of the ground. ,satisfy This means obtaining the joint angles of the multi-degree-of-freedom robotic arm at different path points, thereby driving the robotic arm to complete the edge light collection trajectory.

[0027] Preferably, in step five, the ideal contour point set of the target edge region is set as follows: The actual processing area point set after re-inspection is Uncovered areas If the area not covered > , Preset area threshold, or edge deviation ,satisfy , If a preset deviation threshold is set, then compensation for light reception is determined.

[0028] Preferably, in step one, the parking position of the tracked mobile chassis maintains a preset distance from the edge area to be cleaned, so as to avoid interference between the chassis and the wall or column base, and at the same time prevent the tracks from driving into the wet and soft edge area.

[0029] Preferably, in step two, if it is necessary to transform the target point to the world coordinate system, a pose transformation matrix of the robotic arm base relative to the world coordinate system needs to be introduced. w T b The homogeneous coordinates of the target point in the world coordinate system w P= w T b` b T c` c p.

[0030] Preferably, in step four, the reciprocating motion direction of the trowel is parallel to the contour of the edge area, and the working surface of the trowel remains in contact with the concrete floor surface during the motion.

[0031] Preferably, in step five, the re-inspection involves acquiring image information of the area after light reception using a machine vision perception module, and the industrial control computer then completes the actual processing of the area point set. The identification and calculation of the data enable automated re-inspection.

[0032] The beneficial effects of this invention are as follows:

[0033] 1. This invention adopts a tracked mobile chassis with a differential drive structure, which has a large walking ground contact area and can travel smoothly on soft, wet concrete surfaces without slipping or sinking. At the same time, it can turn and adjust its position flexibly on the spot. Combined with the low-positioned battery pack, it effectively lowers the center of gravity of the whole machine, greatly improving the docking stability and anti-overturning ability of the equipment during operation. It can be adapted to various narrow construction sites to complete precise docking and positioning.

[0034] 2. It adopts a collaborative mode of coarse positioning of the chassis and fine operation of multi-degree-of-freedom robotic arms. With the help of six-degree-of-freedom robotic arms, the working posture can be flexibly adjusted. With the help of flat plate trowel, it can carry out reciprocating finishing work. It can reach the edge blind spots that traditional finishing equipment cannot reach, such as wall edges, column bases, and construction joints. It completely eliminates the manual finishing operation mode, effectively improves the construction coverage of the floor edge, unifies the overall finishing quality, and reduces the problems of uneven construction height and uneven surface texture.

[0035] 3. Relying on machine vision to automatically identify and locate edge areas, and combining with a dedicated trajectory planning algorithm to generate smooth operation paths, a fully automated construction closed loop is constructed, consisting of visual perception, trajectory calculation, mechanical execution, and visual re-inspection and compensation. There is no need for manual on-site layout and manual alignment, which not only greatly reduces the labor intensity of on-site construction and improves the efficiency of floor edge finishing, but also allows for quality self-inspection and correction based on quantitative judgment standards, thus fully ensuring the standardization and consistency of concrete floor edge finishing. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0037] Figure 2 This is a schematic diagram of a tracked mobile chassis structure.

[0038] Figure 3 A schematic diagram of a multi-degree-of-freedom robotic arm and its mounting structure;

[0039] Figure 4 This is a schematic diagram of a plate-type light-collecting actuator.

[0040] Figure 5 A schematic diagram showing the layout of the machine vision perception module;

[0041] Figure 6 This is a schematic diagram of the internal cross-section of a tracked mobile chassis;

[0042] Figure 7 This is a block diagram of the control system.

[0043] Figure 8 This is a schematic diagram of the work status;

[0044] Figure 9 This is a flowchart of the light collection method.

[0045] In the picture:

[0046] 1. Tracked mobile chassis; 11. Frame; 12. Left tracked travel mechanism; 13. Right tracked travel mechanism; 14. Differential drive mechanism; 141. Left drive module; 142. Right drive module; 15. Battery pack; 16. Mounting platform; 17. Support column; 21. Robotic arm base; 22. Multi-degree-of-freedom robotic arm; 23. Plate-type light-collecting actuator; 231. Plate-type trowel; 232. Trowel mounting base; 31. Machine vision perception module; 311. Vision sensor; 312. Sensor bracket; 4. Sheet metal control box; 42. Cooling fan. Detailed Implementation

[0047] The application principle of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0048] Example 1:

[0049] Please see Figure 1-9 The concrete floor edge finishing robot of the present invention includes a tracked mobile chassis 1, a mounting platform 16, a supporting column 17, a robotic arm base 21, a multi-degree-of-freedom robotic arm 22, a plate-type finishing actuator 23, a machine vision perception module 31, and a sheet metal control box 4. The whole machine adopts a chassis-supported combined with upper-level execution arrangement. The tracked mobile chassis 1 serves as the foundation for the whole machine's movement. The upper mounting platform 16 is used to house the robotic arm base 21, the supporting column 17, and the sheet metal control box 4. The multi-degree-of-freedom robotic arm 22 is mounted on the robotic arm base 21, and its end is connected to the plate-type finishing actuator 23 for actually performing finishing actions on the edge area of ​​the floor. The sheet metal control box 4 is located at the rear of the mounting platform 16, and a cooling fan 42 is installed on the outside of the box for cooling the various electronic control modules inside the box.

[0050] In the overall workflow, the tracked mobile chassis 1 is responsible for movement and docking, the machine vision perception module 31 is responsible for identifying the target edge area, the multi-degree-of-freedom robotic arm 22 is responsible for executing trajectory movements based on the calculation results, and the plate-type light-retrieving actuator 23 is responsible for the final light-retrieving of the floor edge. The overall structure integrates movement, recognition, calculation, and execution, making it suitable for the automated operation needs of narrow spaces and unstructured edge areas in floor construction scenarios.

[0051] Please see Figure 2 The tracked mobile chassis 1 includes a frame 11, a left track traveling mechanism 12, a right track traveling mechanism 13, and a mounting platform 16. The frame 11 serves as the main load-bearing structure of the chassis, supporting and connecting the left track traveling mechanism 12, the right track traveling mechanism 13, the mounting platform 16, and various internal functional components. The left track traveling mechanism 12 and the right track traveling mechanism 13 are respectively located on both sides of the frame 11 to achieve stable movement of the entire machine.

[0052] The mounting platform 16 is located on the upper surface of the tracked mobile chassis 1 and serves as the mounting base for the upper equipment. The robotic arm base 21, support column 17, and sheet metal control box 4 are all connected to the tracked mobile chassis 1 via the mounting platform 16, thus forming an integrated structural layout. Due to its large installation area and good planar rigidity, the mounting platform 16 can provide stable installation support for the movement of the multi-degree-of-freedom robotic arm 22.

[0053] Please see Figure 6 The tracked mobile chassis 1 is internally equipped with a differential drive mechanism 14 and a battery pack 15. The differential drive mechanism 14 includes a left drive module 141 and a right drive module 142, wherein the left drive module 141 is driven by the left track walking mechanism 12, and the right drive module 142 is driven by the right track walking mechanism 13. When the left drive module 141 and the right drive module 142 drive at the same speed, the tracked mobile chassis 1 moves forward or backward in a straight line; when there is a speed difference between the left drive module 141 and the right drive module 142, the tracked mobile chassis 1 can turn; when the left and right tracks drive in opposite directions, the whole machine can turn on the spot. The differential drive mechanism 14 gives the whole machine good mobility, making it easy to adjust its posture and stop near the target area in narrow construction areas.

[0054] The battery pack 15 is located inside the lower part of the tracked mobile chassis 1. This arrangement brings the power unit as close to the bottom as possible, thereby lowering the overall center of gravity and improving the anti-tipping ability and overall stability of the tracked mobile chassis 1 when the robotic arm is extended. The battery pack 15 supplies power to the tracked mobile chassis 1, the machine vision perception module 31, and the industrial computer, robotic arm controller, AGV chassis controller, power module, and communication module inside the sheet metal control box 4. Placing the battery pack 15 inside the chassis also helps to enhance the centralization and protection of the overall wiring.

[0055] Please see Figure 3 A multi-degree-of-freedom robotic arm 22 is mounted on a robotic arm base 21, which is fixedly mounted on an installation platform 16. A support column 17 is mounted on the installation platform 16 and is used to support related installation structures. In this invention, the support column 17 mainly serves to provide auxiliary support, improve local installation rigidity, and provide a structural foundation for the arrangement of related accessories.

[0056] The multi-degree-of-freedom robotic arm 22 is preferably a six-degree-of-freedom robotic arm, capable of adjusting its posture and changing its spatial position in multiple directions, enabling its end-effector, the light-collecting actuator 23, to reach the target area along a predetermined trajectory and maintain a suitable working posture. Compared to fixed or low-degree-of-freedom actuators, the multi-degree-of-freedom robotic arm 22 is more suitable for edge-fitting operations in complex edge areas such as walls, column bases, construction joints, and narrow corners, significantly improving the accessibility of the light-collecting range.

[0057] In actual construction, the tracked mobile chassis 1 can first complete the rough movement and docking, and then the multi-degree-of-freedom robotic arm 22 is responsible for the fine execution of local parts. This avoids interference caused by the chassis being too close to the wall or column base, and also avoids the tracks driving directly into soft, wet areas that are not suitable for load-bearing. This mode of chassis coarse positioning combined with robotic arm fine execution is one of the important features of this embodiment for adapting to complex construction environments.

[0058] Please see Figure 4 The plate-type finishing actuator 23 is located at the end of the multi-degree-of-freedom robotic arm 22, and includes a plate-type trowel 231 and a trowel mounting base 232. The plate-type trowel 231 is a flat finishing component, whose working surface is in direct contact with the concrete floor surface during use, and achieves smoothing and finishing of the edge area through reciprocating motion. The trowel mounting base 232 is located on the upper part of the plate-type trowel 231 and is used to fix the plate-type trowel 231 to the end flange or connecting component of the multi-degree-of-freedom robotic arm 22.

[0059] Unlike the rotary trowel structure commonly used for large-area flooring, the 231 flat trowel is flatter and more suitable for edge finishing near walls or column bases. Since edge areas typically have distinct linear boundaries or partially closed contours, the 231 flat trowel, through its reciprocating linear movement along the edge, can more effectively extend the work area to locations that traditional rotary finishing tools struggle to reach, thereby improving the integrity and consistency of edge finishing.

[0060] Please see Figure 5 The machine vision perception module 31 includes a vision sensor 311 and a sensor bracket 312. The vision sensor 311 is mounted on the sensor bracket 312 and oriented towards the area to be captured. The sensor bracket 312 is used to maintain the installation stability and field of view of the vision sensor 311 on the whole machine. The vision sensor 311 can be an industrial camera, a depth camera, or other vision sensing devices suitable for identifying the edge areas of the floor, and its function is to acquire image information of the wall edges, column bases, or construction boundary areas.

[0061] After the image information collected by the vision sensor 311 is transmitted to the industrial control computer, the computer processes the image, extracts edges, and identifies the target to obtain the target coordinates or pose of the area to be illuminated. This method allows the robot to automatically identify the spatial position parameters of the edge area without manual line laying or position estimation, providing a basis for subsequent robotic arm trajectory generation. Compared to purely manual teaching or construction methods that rely entirely on manual alignment, this invention improves the efficiency and consistency of target localization through the machine vision perception module 31.

[0062] Please see Figure 7The control system of this invention includes a camera, an industrial computer, a robotic arm controller, a robotic arm, an AGV chassis controller, a chassis drive module, a Buke motor, and a power supply module. The camera is used to acquire image data and send it to the industrial computer; the industrial computer receives image information output by the machine vision perception module 31, processes the image information, and generates target coordinates, target pose, or execution trajectory; the robotic arm controller receives the target coordinates, target pose, or execution trajectory sent by the industrial computer and controls the multi-degree-of-freedom robotic arm 22 to complete the corresponding movements; the AGV chassis controller receives movement commands sent by the industrial computer and controls the chassis drive module and the Buke motor to realize the movement and docking of the tracked mobile chassis 1; the power supply module supplies power to the above control and drive modules.

[0063] Please continue reading. Figure 1 The sheet metal control box 4 houses the industrial computer, robotic arm controller, AGV chassis controller, chassis drive module, power supply module, and communication module. A cooling fan 42 is located on the outside of the sheet metal control box 4 to facilitate airflow within the box, reduce heat buildup, and thus improve the stability of the control system during long-term operation. Centralized arrangement through the sheet metal control box 4 not only facilitates the integration and maintenance of the entire machine's electrical control system but also aids in protection and transportation during on-site construction.

[0064] The control logic of this embodiment is as follows: the camera or vision sensor 311 acquires images, which are then processed by the industrial control computer. On one hand, the industrial control computer sends the target coordinates or execution trajectory to the robotic arm controller to control the multi-degree-of-freedom robotic arm 22 to drive the plate-type light-collecting actuator 23 to complete edge light collection; on the other hand, it sends a chassis movement command to the AGV chassis controller to drive the tracked mobile chassis 1 to move to the next working position. This constitutes a closed-loop construction process of "visual recognition - coordinate calculation - trajectory execution - re-inspection and compensation".

[0065] Please see Figure 8 When the robot is working, the tracked mobile chassis 1 first moves to the vicinity of the edge area to be finished and stops. After stopping, the multi-degree-of-freedom robotic arm 22 extends towards the target area, bringing the plate-type finishing actuator 23 close to the wall or column base area. The plate-type trowel 231 is close to the concrete floor surface and reciprocates along the edge direction to finish the edge area. Since the tracked mobile chassis 1 does not need to directly press against the wall, it can reduce mechanical interference between the chassis and the boundary and also reduce the risk of the chassis driving into a locally wet and soft edge area.

[0066] The operation mode of this embodiment is mainly reflected in the collaborative mode of vehicle-stopping combined with boom movement. That is, the tracked mobile chassis 1 completes regional-level movement and docking, while the multi-degree-of-freedom robotic arm 22 completes local fine construction. This mode retains the construction range advantage of the mobile platform and makes full use of the flexibility and accessibility of the robotic arm, which is particularly suitable for construction tasks in narrow locations and with high posture requirements, such as the edge areas of the ground.

[0067] Example 2:

[0068] Based on Example 1, a method for planning the light collection trajectory of a robot for collecting light at the edge of a concrete floor is provided, including the following steps:

[0069] Step 1: Move the tracked mobile chassis 1 to the vicinity of the edge area to be collected and stop it;

[0070] Step 2: Collect image information of the edge region of the light to be collected through the machine vision perception module 31, identify the edge region, and obtain the target coordinates or target pose;

[0071] Step 3: Generate the execution trajectory of the end effector of the multi-degree-of-freedom robotic arm 22 based on the target coordinates or target pose;

[0072] Step 4: Control the multi-degree-of-freedom robotic arm 22 to drive the plate-type light-collecting actuator 23 to move, so that the plate-type wiping plate 231 performs reciprocating light collection along the edge area;

[0073] Step 5: Re-inspect the light collection results and perform compensatory light collection for areas that are not covered or have insufficient light collection quality.

[0074] Further, in step two, the image information collected by the vision sensor 311 is transmitted to the industrial control computer, which performs edge extraction, target recognition, and coordinate calculation on the image information to determine the position parameters of the area to be covered. Further, in step four, the plate-type trowel 231, driven by the multi-degree-of-freedom robotic arm 22, performs reciprocating linear motion along the wall or column edge to complete the edge-covering and light-collecting operation. Further, in step five, if the re-inspection results indicate the presence of uncovered areas or areas with insufficient light-collecting quality, steps three and four are re-executed to achieve compensation trajectory generation and secondary light collection.

[0075] To further illustrate the motion control principle of the tracked multi-degree-of-freedom concrete floor edge-collecting robot in this embodiment during automatic edge-collecting, the following section describes its kinematic relationships, coordinate transformation relationships, and trajectory planning methods in conjunction with the working process of the tracked mobile chassis, multi-degree-of-freedom robotic arm, and machine vision perception module.

[0076] In this embodiment, the tracked mobile chassis 1 uses a differential drive system with left and right tracks to achieve forward, backward, turning, and in-place turning. Let the linear velocity of the left track traveling mechanism 12 be... The linear velocity of the right track traveling mechanism 13 is The center distance between the left and right tracks is Then the linear velocity of the tracked mobile chassis 1 and angular velocity It can be represented as:

[0077]

[0078]

[0079] Therefore, the kinematic equations of the tracked mobile chassis 1 in the ground coordinate system can be obtained as follows: ;in, These are the position coordinates of the tracked mobile chassis in the ground coordinate system. This refers to the chassis heading angle. By adjusting the output speeds of the left drive module 141 and the right drive module 142, the tracked mobile chassis 1 can reach the predetermined stopping position and provide a stable base for the subsequent fine light collection by the robotic arm.

[0080] In this embodiment, the machine vision perception module 31 is used to identify wall edges, column bases, or construction boundary areas, and output the position parameters of the target point in the visual coordinate system. In order for the multi-degree-of-freedom robotic arm 22 to perform the corresponding actions, the target point in the visual coordinate system needs to be transformed into the robotic arm base coordinate system.

[0081] Let the homogeneous coordinates of the target point in the 311 coordinate system of the vision sensor be: c p= Then its homogeneous coordinates in the robot arm base coordinate system can be expressed as: b P The homogeneous transformation matrix from the visual coordinate system to the robot arm base coordinate system is: b T c = Furthermore, if it is necessary to transform the target point to the world coordinate system or the chassis coordinate system, it can also be written as: w P= w T b` b T c` c p in which w T b is the pose transformation matrix of the robotic arm base 21 relative to the world coordinate system.

[0082] After obtaining the target coordinates or pose of the edge region to be collected, it is necessary to generate a reciprocating execution trajectory along the edge direction for the plate-type light-collecting actuator 23. Let the target path curve of the edge region be:

[0083]

[0084] Where s is the path parameter. Then, when the plate-type wiping plate 231 performs reciprocating light collection, its ideal end pose can be expressed as:

[0085]

[0086] in Indicates that the plate-type trowel 231 is at the path point The orientation matrix at the location. To ensure good adhesion between the slab 231 and the concrete floor surface, its orientation should ensure that the normal direction of the slab surface is as consistent as possible with the normal direction of the floor surface, that is: ,in, This is the normal vector of the plate-type trowel 231. This is the normal vector of the local plane of the floor. This posture constraint can improve the stability and uniformity of edge finishing.

[0087] In actual control, the multi-degree-of-freedom robotic arm 22 can be controlled according to the end-effector pose matrix. By performing inverse kinematics, we obtain the joint variable vector: kinetic relationship: This allows the joint angles of the multi-degree-of-freedom robotic arm 22 at different path points to be obtained, thereby driving the robotic arm to complete the edge light collection trajectory.

[0088] When the multi-degree-of-freedom robotic arm 22 drives the plate-type trowel 231 to perform reciprocating motion, if the speed change at the reversal point is too rapid, it can easily cause mechanical impact and uneven finishing marks. To reduce the reversal impact, this invention can use fifth-order polynomial trajectory planning to continuously constrain displacement, velocity, and acceleration. Let the trajectory parameters be... Within the time interval [0, T], the following conditions are met:

[0089]

[0090] The corresponding velocities and accelerations are as follows:

[0091]

[0092]

[0093] By constraining the starting position, ending position, starting and ending speeds, and starting and ending accelerations, polynomial coefficients can be solved, thereby achieving a smooth speed transition of the plate-type trowel 231 during the reversal process. This method helps reduce the impact of the plate-type trowel 231 at the edges and improves the consistency of the finished surface.

[0094] In step five of this embodiment, the light acquisition result is re-examined, and compensation light acquisition is performed on uncovered or low-quality areas. Let the ideal edge contour point set obtained from the target edge region identification be... The actual processing area point set obtained after re-inspection is: The uncovered area can be represented as: If the area not covered satisfy: > This indicates the existence of areas requiring light compensation. Furthermore, the edge deviation can also be defined as: ,when If the light reception deviation in the current area exceeds the allowable range, it is necessary to regenerate the compensation trajectory and perform secondary light reception. The above-mentioned judgment method enables quantitative analysis of the light collection effect in the edge area, rather than relying solely on manual visual inspection, thus improving the rationality and scientific nature of automated compensation.

[0095] Example 3:

[0096] To further illustrate the practical application process of Embodiments 1 and 2, an example of an edge light collection implementation process is given below:

[0097] First, the tracked mobile chassis 1 moves to the vicinity of the area to be collected under the control of the AGV chassis controller, and determines the stopping position according to the construction boundary and the working space of the robotic arm;

[0098] Subsequently, the vision sensor 311 acquires images of the target area, and the industrial control computer extracts the edge contours and calculates the target coordinates and target pose.

[0099] Afterwards, the industrial control computer generates the edge-attaching reciprocating execution trajectory based on the geometric recognition results and the dimensions of the plate-type trowel 231, and sends it to the robotic arm controller;

[0100] The robotic arm controller controls the multi-degree-of-freedom robotic arm 22 to drive the plate-type light-collecting actuator 23 to reciprocate along the edge direction to complete the light collection;

[0101] After the light collection is completed, the machine vision perception module 31 collects and analyzes the processed area again. If a missed area or deviation exceeds the limit is detected, the compensation light collection step is executed again.

[0102] As can be seen from the above process, this embodiment not only achieves the organic integration of chassis, robotic arm and vision module in terms of structure, but also establishes a closed-loop operation process of "identification-calculation-execution-re-inspection-compensation" in terms of method, making the edge light collection process more automated and scientific.

Claims

1. A robot for finishing the edges of concrete floors, characterized in that: It includes a tracked mobile chassis (1), a multi-degree-of-freedom robotic arm (22), a plate-type light-collecting actuator (23), a machine vision perception module (31), and control components for realizing light-collecting trajectory planning; The tracked mobile chassis (1) is used to move the whole machine to the edge area to be collected and stop stably, providing a basis for the execution of the light collection trajectory; The machine vision perception module (31) is used to collect image information of the edge area to be collected, identify the target area and output the target coordinates or target pose, and provide position parameters for the planning of the light collection trajectory; the multi-degree-of-freedom robotic arm (22) is used to drive the plate light collection actuator (23) to move according to the planned light collection trajectory, and the plate light collection actuator (23) is used to perform edge-following light collection action along the planned trajectory; The control components include an industrial computer, a robotic arm controller, and an AGV chassis controller. The industrial computer is used to receive the output information of the machine vision perception module (31) and complete the light-receiving trajectory planning. The robotic arm controller is used to control the multi-degree-of-freedom robotic arm (22) to move according to the planned trajectory. The AGV chassis controller is used to control the tracked mobile chassis (1) to complete the docking and positioning.

2. The concrete floor edge finishing robot according to claim 1, characterized in that: The tracked mobile chassis (1) is provided with an installation platform (16) at its top. A support column (17) is provided on one side of the upper end of the installation platform (16). The robotic arm base (21) is installed on the upper end of the support column (17). The multi-degree-of-freedom robotic arm (22) is installed on the robotic arm base (21). The multi-degree-of-freedom robotic arm (22) adjusts its attitude in multiple directions according to the planned light-collecting trajectory, so that the plate-type light-collecting actuator (23) completes the light-collecting action along the target trajectory.

3. The concrete floor edge finishing robot according to claim 1, characterized in that: The tracked mobile chassis (1) includes a frame (11), a left tracked walking mechanism (12), a right tracked walking mechanism (13), a differential drive mechanism (14), and a battery pack (15); the differential drive mechanism (14) includes a left drive module (141) and a right drive module (142), which are respectively connected to the left tracked walking mechanism (12) and the right tracked walking mechanism (13); the battery pack (15) is located in the lower part of the tracked mobile chassis (1) and is used to power the whole machine.

4. The concrete floor edge finishing robot according to claim 1, characterized in that: The plate-type light-collecting actuator (23) includes a flat plate (231) and a plate mounting base (232). The plate mounting base (232) is used to fix the flat plate (231) to the end of the multi-degree-of-freedom robotic arm (22). The flat plate (231) has a flat structure.

5. The concrete floor edge finishing robot according to claim 1, characterized in that: The machine vision perception module (31) includes a vision sensor (311) and a sensor bracket (312). The sensor bracket (312) is used to fix the vision sensor (311) and adjust its field of view. The visual sensor (311) transmits the collected image information to the industrial control computer, which then performs target recognition, edge extraction and coordinate calculation. The control component also includes a chassis drive module, a power supply module and a communication module, which are integrated into a sheet metal control box (4) on the other side of the upper end of the mounting platform (16). A cooling fan (42) is installed on the outside of the sheet metal control box (4) for thermal management of the control component.

6. The method for planning the light-collecting trajectory of the concrete floor edge-collecting robot according to any one of claims 1-5, characterized in that: Includes the following steps: Step 1: Receive the movement command sent by the industrial computer through the AGV chassis controller, control the differential drive mechanism (14) to drive the tracked mobile chassis (1) to move to the vicinity of the edge area to be collected and complete the docking; Step 2: Collect image information of the edge area to be received by the machine vision perception module (31) through the vision sensor (311), and transmit it to the industrial control computer. The industrial control computer performs edge extraction, target recognition and coordinate calculation on the image, obtains the homogeneous coordinates of the target point in the visual coordinate system, and then transforms it to the coordinate system of the robot arm base through the homogeneous transformation matrix to obtain the target coordinates or target pose. Step 3: Generate the execution trajectory of the end effector of the multi-degree-of-freedom robotic arm (22) based on the target coordinates or target pose; Step 4: Use fifth-order polynomial trajectory planning to smooth the end effector trajectory of the multi-degree-of-freedom robotic arm (22), and the trajectory parameters are... Within the time interval [0, T], it satisfies Corresponding speed acceleration By constraining the starting position, ending position, starting and ending speeds and starting and ending accelerations, the polynomial coefficients are solved, and the multi-degree-of-freedom robotic arm (22) is controlled to drive the plate-type light-collecting actuator (23) to move, so that the plate-type wiping plate (231) performs reciprocating light collection along the edge area; Step 5: Re-inspect the light collection results and perform compensatory light collection for areas that are not covered or have insufficient light collection quality.

7. The method for planning the light-collecting trajectory of the concrete floor edge-collecting robot according to claim 6, characterized in that: In step one, the linear velocity of the tracked mobile chassis (1) and angular velocity satisfy: in, The linear velocity of the left track traveling mechanism; The linear velocity of the right track traveling mechanism; The distance between the center points of the left and right tracks is given by the following kinematic equation in the ground coordinate system: ;in, These are the position coordinates of the tracked mobile chassis in the ground coordinate system. This is the chassis heading angle.

8. The method for planning the light-collecting trajectory of the concrete floor edge-collecting robot according to claim 6, characterized in that: In step two, the homogeneous coordinates of the target point in the visual sensor coordinate system are: c p= Its homogeneous coordinates in the robot arm base coordinate system are: b P The homogeneous transformation matrix is b T c = ; in, , , This represents the three-dimensional position coordinates of the target point in the camera coordinate system; 1 indicates the homogeneous coordinate form, used for multiplication with the 4×4 transformation matrix; b T c Represents the homogeneous transformation matrix from the camera coordinate system to the robot arm base coordinate system; This represents a 3×3 rotation matrix that describes the camera's orientation relative to the robotic arm base. This represents a 3×1 translation vector, describing the position of the camera relative to the robotic arm base.

9. The method for planning the light-collecting trajectory of the concrete floor edge-collecting robot according to claim 6, characterized in that: In step three, let the target path curve of the edge region be... ,in, For path parameters; The ideal end position of a plate-type trowel is ,in, For the board-type trowel at the path point The pose matrix at the location, the pose satisfies , Let be the normal vector of the slab-type trowel. The joint variable vector is obtained by solving the inverse kinematics problem, where the normal vector is the local plane of the ground. ,satisfy That is, to obtain the joint angles of the multi-degree-of-freedom robotic arm (22) at different path points, thereby driving the robotic arm to complete the edge light collection trajectory.

10. The method for planning the light-collecting trajectory of the concrete floor edge-collecting robot according to claim 6, characterized in that: In step five, let the ideal contour point set of the target edge region be... The actual processing area point set after re-inspection is Uncovered areas If the area not covered > , Preset area threshold, or edge deviation ,satisfy , If a preset deviation threshold is set, then compensation for light reception is determined.

Images