A pallet handling method and associated apparatus

Abstract

The application discloses a pallet conveying method and a related device, and the method is implemented by using a conveying robot comprising a first fork arm and a second fork arm, and the method comprises the following steps: acquiring target radar data collected by the first fork arm and the second fork arm respectively for a fork-in surface of a target pallet; acquiring a plurality of candidate line segments corresponding to each group of the target radar data after clustering processing; determining a first inner edge line and a second inner edge line matched with a fork joint area of the target pallet based on distances between different candidate line segments; and generating a target path of the conveying robot conveying the target pallet based on the first inner edge line and the second inner edge line. In this way, the application can improve the accuracy of pallet conveying.

Description

Technical Field

[0001] This application relates to the field of robotics, and in particular to a method and apparatus for pallet handling. Background Technology

[0002] In logistics warehousing and industrial production scenarios, forklift handling of pallets is a high-frequency material transfer method. Traditional handling methods mainly rely on manually operated forklifts, where the forks are inserted into the pallet's fork-joint area to lift and move goods. However, with increasing warehouse density and faster work cycles, the traditional model has revealed significant bottlenecks: manual operation requires high driving skills, and blind spots in narrow aisles can easily lead to pallet collisions and cargo tipping; repetitive handling can easily cause fatigue, resulting in fork alignment deviations. In addition, manual recording of handling data is lagging and cannot meet the needs of modern digital warehouse management.

[0003] Therefore, improving the accuracy and efficiency of pallet handling has become an urgent problem to be solved. Summary of the Invention

[0004] The main technical problem addressed by this application is to provide a pallet handling method and related apparatus that can improve the accuracy and efficiency of pallet handling.

[0005] To address the aforementioned technical problems, this application provides a pallet handling method. This method utilizes a handling robot comprising a first fork arm and a second fork arm. The method includes: acquiring target radar data collected by the first and second fork arms respectively, targeting the fork-in surface of the target pallet; acquiring multiple candidate line segments corresponding to each group of target radar data after clustering processing; determining a first inner edge line and a second inner edge line matching the fork-in area of ​​the target pallet based on the distance between different candidate line segments; and generating a target path for the handling robot to handle the target pallet based on the first inner edge line and the second inner edge line.

[0006] To solve the above-mentioned technical problems, another technical solution adopted in this application is to provide an electronic device, including a memory and a processor coupled to each other, wherein the memory stores program instructions, and the processor is used to execute the program instructions to implement the method mentioned in the above technical solution.

[0007] To solve the above-mentioned technical problems, another technical solution adopted in this application is to provide a computer-readable storage medium having program instructions stored thereon, wherein the program instructions, when executed by a processor, implement the method mentioned in the above technical solution.

[0008] The beneficial effects of this application are as follows: Unlike existing technologies, the pallet handling method proposed in this application utilizes the first and second forks of a handling robot to collect target radar data corresponding to the fork-in surface of the target pallet, and extracts multiple candidate line segments from each set of target radar data. Based on the distances of different candidate line segments, a first inner edge line and a second inner edge line matching the fork-in area of ​​the target pallet are determined from the multiple candidate line segments. The target position information of the target pallet in the map coordinate system is determined based on the first and second inner edge lines, and a target path is generated, which greatly reduces the difficulty of locating the target pallet and improves the handling accuracy of the target pallet. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a flowchart illustrating one embodiment of the pallet handling method of this application; Figure 2 yes Figure 1 The flowchart of step S101 corresponds to another embodiment; Figure 3 yes Figure 2 The flowchart of step S201 corresponds to another embodiment; Figure 4 yes Figure 1 The flowchart of step S102 corresponds to another embodiment; Figure 5 yes Figure 1 The flowchart of step S103 corresponds to another embodiment; Figure 6 This is a schematic diagram of the structure of one embodiment of the target stack of this application; Figure 7 yes Figure 5 The flowchart of step S501 corresponds to another embodiment; Figure 8 This is a schematic diagram of another embodiment of the target stack of this application. Figure 9 yes Figure 5 A flowchart corresponding to another embodiment before step S501; Figure 10 yes Figure 1 The flowchart of step S104 corresponds to another embodiment; Figure 11This is a schematic diagram of the structure of one embodiment of the electronic device of this application; Figure 12 This is a schematic diagram of one embodiment of the computer-readable storage medium of this application. Detailed Implementation

[0010] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments, and different embodiments can be adaptively combined. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0011] Please see Figure 1 , Figure 1 This is a flowchart illustrating one embodiment of the pallet handling method of this application. The pallet handling method is implemented using a handling robot including a first fork and a second fork. Specifically, a processor utilizes the handling robot to execute the pallet handling method proposed in this application to transfer the target pallet. The processor can be deployed inside the handling robot or can be set up independently of the handling robot. The pallet handling method specifically includes: S101: For the fork-in surface of the target pallet, acquire the target radar data collected by the first fork arm and the second fork arm respectively.

[0012] In one embodiment, a target pallet to be transported is set in the target scene. The first and second forks of the transport robot collect radar data of the target scene. From the data corresponding to the target time, the target radar data corresponding to the fork-in surface of the target pallet is determined. It should be noted that the first and second forks collect corresponding target radar data respectively; and the target time can be the current time or any time during the transport robot's transport task.

[0013] In some implementation scenarios, radar sensors are respectively installed at the front ends of the first and second forks of the handling robot. The first and second forks use the installed radar sensors to collect radar data of the target scene to obtain target radar data corresponding to the fork-in surface of the target pallet.

[0014] S102: Obtain multiple candidate line segments corresponding to each group of target radar data after clustering.

[0015] In one embodiment, the target radar data collected by the radar sensors of the first and second forks are clustered to obtain radar point cloud clusters corresponding to each group of target radar data. Based on each radar point cloud cluster, corresponding candidate line segments are extracted.

[0016] In some implementation scenarios, Euclidean clustering is performed on the target radar data collected by the first fork arm to obtain multiple radar point cloud clusters corresponding to the first fork arm. Line segment fitting is then performed on each radar point cloud cluster to extract corresponding candidate line segments. Similarly, Euclidean clustering is performed on the target radar data collected by the second fork arm to obtain multiple radar point cloud clusters corresponding to the second fork arm. Line segment fitting is then used to obtain multiple candidate line segments corresponding to the first fork arm. Euclidean clustering classifies multiple radar point clouds according to a preset clustering threshold. This clustering threshold can be obtained through estimation or determined based on the geometric resolution and recognition distance of the radar sensor. The higher the geometric resolution or the greater the recognition distance, the larger the clustering threshold becomes.

[0017] In one embodiment, the implementation process of step S102 may further include: clustering the target radar data collected by the first and second forks to obtain radar point cloud clusters corresponding to each group of target radar data. Based on each radar point cloud cluster, corresponding line segments are extracted, and the extracted line segments are filtered, with line segments whose lengths fall within a preset length range serving as candidate line segments. The preset length can be determined through estimation or based on the size information of the target pallet, to avoid interference from excessively short or long interfering line segments that could disrupt subsequent steps.

[0018] S103: Based on the distance between different candidate line segments, determine the first inner edge line and the second inner edge line that match the crossover area of ​​the target pallet.

[0019] In one embodiment, the distance between different candidate line segments is calculated, and based on the distance between different candidate line segments and the width of the crossover area of ​​the target stack, a first inner edge line and a second inner edge line matching the crossover area are determined.

[0020] In some implementation scenarios, the target pallet includes a fork joint area for accommodating the first and second forks. During transport, the first and second forks fully engage in this fork joint area, raising the fork height to support the movement of the target pallet. The distances between different candidate line segments are calculated, and the absolute value of the difference between the candidate line segment distance and the width of the fork joint area is determined. The two candidate line segments corresponding to the absolute values ​​of the differences less than a reference value are selected as the first and second inner edge lines matching the fork joint area.

[0021] In some implementation scenarios, in response to the rectangular cross-section of the aforementioned forked area, the aforementioned first inner edge line is an edge line located on one side inside the forked area, and the aforementioned second inner edge line is an edge line located on the other side inside the forked area, and the heights of the first inner edge line and the second inner edge line are the same.

[0022] In some implementation scenarios, the equation of each candidate line segment in the robot coordinate system is obtained, and the corresponding distance is calculated based on the equations of different candidate line segments. Alternatively, after determining the candidate line segments, the centroid coordinates of each candidate line segment in the robot coordinate system are determined, and the distance between different candidate line segments is calculated based on the centroid coordinates.

[0023] S104: Based on the first inner edge line and the second inner edge line, generate the target path for the transport robot to transport the target pallet.

[0024] In one embodiment, after determining the first inner edge line and the second inner edge line, the target pose of the transport robot at the target time is obtained. Combining the target pose, the first inner edge line, and the second inner edge line, the target position information of the fork-in surface of the target pallet in the map coordinate system is determined. Based on the target position information, path planning is performed to generate the target path for the transport robot to transport the target pallet.

[0025] The pallet handling method proposed in this application utilizes the first and second forks of a handling robot to collect target radar data corresponding to the fork-in surface of the target pallet, and extracts multiple candidate line segments from each set of target radar data. Based on the distances of different candidate line segments, a first inner edge line and a second inner edge line matching the fork-in area of ​​the target pallet are determined from the multiple candidate line segments. The target position information of the target pallet in the spatial coordinate system is determined based on the first inner edge line and the second inner edge line, and a target path is generated. This greatly reduces the difficulty of locating the target pallet and improves the handling accuracy of the target pallet.

[0026] Please see Figure 2 , Figure 2 yes Figure 1 The flowchart of step S101 corresponds to another embodiment. Specifically, the implementation process of step S101 includes: S201: Acquire the reference radar data collected by the first and second forks from the target pallet at the target moment.

[0027] In one embodiment, at the same target time, the first fork arm and the second fork arm respectively collect reference radar data of the target pallet.

[0028] S202: Obtain the initial position of the target pallet to be transported in the target scene in the spatial coordinate system, transform the initial position to the robot coordinate system, and obtain the reference position of at least part of the cross-joining area in the robot coordinate system; wherein, the initial position is determined based on the coordinates of the reference point and the reference angle in the cross-joining surface.

[0029] In one embodiment, the initial position of the target pallet to be transported in the target scene in the spatial coordinate system is obtained, as well as the target pose of the transport robot at the target time. Based on the target pose, the initial position is transformed to the robot coordinate system to obtain the initial position of the fork entry surface in the robot coordinate system.

[0030] In some implementation scenarios, the reference point coordinates and reference angle of the target pallet in the target scene in the spatial coordinate system are obtained from the processor. Based on the above reference point coordinates and reference angle, a homogeneous transformation matrix is ​​constructed, and this homogeneous transformation matrix is ​​used as the initial position. The reference point coordinates are the coordinates of the midpoint of the lower edge line of the target pallet's fork-in face in the spatial coordinate system, and the reference angle is used to characterize the orientation of the target pallet in the spatial coordinate system.

[0031] Further, based on the depth of the fork region of the target pallet, the relative pose from the fork-in surface of the target pallet to a fixed depth inside the target pallet is determined. Based on the initial position and relative pose of the fork-in surface in the spatial coordinate system and the robot coordinate system, the reference position of at least a portion of the fork region of the target pallet in the robot coordinate system is determined. The reference position is calculated by the following formula (1).

[0032] (1) in, This indicates the reference position of at least part of the crossover area in the robot coordinate system. This indicates the initial position of the fork-in face in the robot coordinate system. This indicates the initial position of the fork in the spatial coordinate system. Indicates relative pose.

[0033] S203: Based on the reference position, extract the target radar data corresponding to the intrusion area from each set of reference radar data.

[0034] In one embodiment, a point cloud filtering region corresponding to the fork-in region of the target pallet is determined based on the reference position corresponding to the fork-in region. This point cloud filtering region is then used to filter the reference radar data collected by the first and second forks respectively, to obtain the target radar region corresponding to the fork-in surface.

[0035] In some implementation scenarios, the reference radar data acquired by the first and second forks respectively includes multiple reference radar point clouds. The reference positions corresponding to the forked areas are appropriately expanded to obtain point cloud filtering regions. For the reference radar data acquired by the first fork, all reference radar point clouds located within the aforementioned point cloud filtering region are used as target radar data. Similarly, for the reference radar data acquired by the second fork, all reference radar point clouds located within the aforementioned point cloud filtering region are used as target radar data.

[0036] The above scheme improves the efficiency of target path generation by identifying the target radar data corresponding to the entry area and filtering out radar data with low reference value, thereby reducing subsequent computational costs.

[0037] Please see Figure 3 , Figure 3 yes Figure 2 The flowchart of step S201 corresponds to another embodiment. Specifically, the implementation process of step S201 includes: S301: Acquire multiple sets of initial radar data within the target scene collected by the first and second forks respectively.

[0038] In one embodiment, based on the initial position set by the processor, the transport robot is driven to move to a position close to the target pallet. The first and second forks are used to continuously acquire radar data of the target scene, resulting in multiple sets of initial radar data collected by the first and second forks respectively.

[0039] In some implementation scenarios, the initial radar data collected by the first and second forks each time are matched with corresponding timestamps.

[0040] S302: Based on the timestamp of the initial radar data, determine the reference radar data collected by the first and second forks at the target time.

[0041] In one embodiment, the acquisition time difference between the first initial radar data acquired by the first fork arm and the second initial radar data acquired by the second fork arm is obtained by matching the timestamps of the first initial radar data acquired by the first fork arm and the second initial radar data acquired by the second fork arm. If the acquisition time difference is less than a preset time threshold, the corresponding first initial radar data and second initial radar data are used as reference radar data acquired by the first fork arm and the second fork arm respectively at the target time.

[0042] In one embodiment, the implementation process of step S302 may further include: determining the maximum and minimum timestamps from multiple sets of initial radar data collected from the first and second forks respectively. The difference between the maximum and minimum timestamps is used as the total time span corresponding to the multiple sets of initial radar data. A preset time threshold is obtained, and the ratio of the total time span to the time threshold is rounded up to obtain multiple data containers, and a reference storage index is determined for each data container. For the first and second forks, all collected initial radar data are sorted according to the timestamp from latest to earliest to obtain the initial radar sequences corresponding to the first and second forks respectively. The initial radar sequences are traversed sequentially, and the ratio of the timestamp of each initial radar data to the time threshold is obtained, and the ratio is rounded down to determine the target storage index of the corresponding initial radar data. Based on the target storage index and the reference storage index corresponding to each data container, the initial radar data is stored in the corresponding data container. In response to any data container simultaneously including initial radar data collected from the first and second forks respectively, the initial radar data in that data container is used as the reference radar data for subsequent data in terms of time dimension. The target time is determined based on the timestamps corresponding to the two sets of reference radar data.

[0043] The aforementioned time threshold can be obtained based on estimation, or it can be inferred from the acquisition interval of the radar sensors on the first and second forks.

[0044] In some implementation scenarios, after determining the time dimension of the reference radar data, the average of the timestamps corresponding to the two sets of reference radar data is used as the target time to help determine the pose of the handling robot. Alternatively, the earlier timestamp of the two sets of reference radar data can be used as the target time. Or, either of the timestamps of the two sets of reference radar data can be used as the target time.

[0045] In a specific application scenario, the maximum timestamp in all initial radar data is 220ms, and the minimum timestamp is 100ms. When the determined time threshold is 25ms, the first difference between the maximum and minimum timestamps is calculated. The ratio of this first difference to the time threshold is rounded up to construct five data containers. Corresponding reference storage indices are assigned to these five data containers: "Bin 0", "Bin 1", "Bin 2", "Bin 3", and "Bin 4". When the first timestamp of the first initial radar data in the initial radar sequence corresponding to the first fork arm is 220ms, the second difference between the first and minimum timestamps (i.e., 120ms) is obtained. The ratio of this second difference to the time threshold is calculated and rounded down (i.e., 4) to determine the reference storage index of the first initial radar data in the initial radar sequence corresponding to the first fork arm as "Bin 4", and this initial radar data is stored in the corresponding data container. When the first timestamp corresponding to the first initial radar data in the initial radar sequence corresponding to the second fork arm is 210ms, the second difference (i.e., 110ms) between the first timestamp and the minimum timestamp is obtained. The ratio of the above second difference to the time threshold is calculated and rounded down (i.e., 4) to determine that the reference storage index corresponding to the first initial radar data in the initial radar sequence corresponding to the second fork arm is also "Bin 4", and this initial radar data is stored in the corresponding data container. At this time, the data container corresponding to "Bin 4" simultaneously includes the initial radar data collected by the first and second fork arms respectively. The two sets of initial radar data in this data container are then used as reference radar data for the time dimension. The target time is determined based on the timestamps of the two sets of reference radar data.

[0046] Optionally, in other embodiments, after determining multiple data containers based on the total time span corresponding to the timestamps in multiple sets of initial radar data, a timestamp range corresponding to each data container is determined. The initial radar data is stored in the corresponding data container based on the timestamp of each initial radar data and each timestamp range. In response to any data container simultaneously including initial radar data collected by the first and second forks respectively, the initial radar data in that data container is used as reference radar data for subsequent radar data along the time dimension. The target time is determined based on the timestamps corresponding to the two sets of reference radar data. The specific process for determining the data containers can be referred to the corresponding embodiments described above.

[0047] In a specific application scenario, the maximum timestamp in all initial radar data is 220ms, the minimum timestamp is 100ms, and the determined time threshold is 25ms. Five data containers are constructed, and the timestamp ranges corresponding to the six data containers are as follows: , , , and When the first timestamp corresponding to the first initial radar data in the initial radar sequence corresponding to the second fork arm is 210ms, the initial radar data is stored in the fifth data container. Similarly, when the first timestamp corresponding to the first initial radar data in the initial radar sequence corresponding to the second fork arm is 210ms, the initial radar data is also stored in the fifth data container. Since the fifth data container now contains the initial radar data collected by the first and second fork arms respectively, these two sets of initial radar data are used as reference radar data to determine the target time.

[0048] In the above solution, the handling robot continuously uses the first and second forks to collect initial radar data during the handling task. By determining two sets of reference radar data that are aligned in time dimension based on the timestamp of each set of initial radar data, the robot avoids the reference radar data being out of sync, which would reduce the accuracy of target path generation.

[0049] Please see Figure 4 , Figure 4 yes Figure 1 The flowchart of step S102 corresponds to another embodiment. Specifically, the implementation process of step S102 includes: S401: Cluster the target radar data collected by the first fork arm and the second fork arm respectively to obtain the clustering result corresponding to each group of target radar data, and obtain the initial line segment corresponding to each clustering result.

[0050] In one embodiment, Euclidean clustering is performed on the target radar data collected by the radar sensors of the first and second forks to obtain a clustering result corresponding to each group of target radar data. This clustering result includes multiple radar point cloud clusters. Based on each radar point cloud cluster, the corresponding initial line segment is extracted. The specific process of performing Euclidean clustering on the target radar data can be referred to the corresponding embodiment described above, and will not be elaborated further here.

[0051] S402: Obtain the length of each initial line segment and the reference angle between the initial line segment and the moving direction of the handling robot. Based on the line segment length and the reference angle, select multiple candidate line segments from all the initial line segments corresponding to the first fork arm and the second fork arm respectively.

[0052] In one embodiment, the length of each initial line segment and the reference angle between each initial line segment and the moving direction of the handling robot are obtained. Initial line segments whose lengths are within a preset length range and whose reference angles are less than an angle threshold are selected as candidate line segments. It should be noted that the angle threshold is mentioned several times in this application; however, the angle threshold may be the same or different in different embodiments.

[0053] In some implementation scenarios, the aforementioned reference angle can be 40°, 41°, or 45°, which can be obtained through estimation or by back-calculation through multiple experiments.

[0054] The above solution, because the first and second forks of the handling robot need to fork into the fork area of ​​the target pallet along the extension direction of the through hole during the pallet handling process, greatly improves the execution efficiency of subsequent steps by combining the line segment length and reference angle to remove initial line segments that are too long, too short, or have too large reference angle deviations in order to facilitate the subsequent screening of the first and second inner edge lines.

[0055] Please see Figure 5 , Figure 5 yes Figure 1 The flowchart of step S103 corresponds to another embodiment. In response to the cross-joint region including at least one through-hole, the specific implementation process of step S103 includes: S501: Based on the geometric relationship information between different candidate line segments and the width of the through hole, determine the reference line segment group corresponding to each through hole; wherein, the geometric relationship information includes at least one of distance, angle difference and endpoint coordinate component difference.

[0056] In one embodiment, geometric relationship information between different candidate line segments is obtained, including at least one of distance, angle difference, and endpoint coordinate component difference. Based on the determined geometric relationship information and the via width, a reference line segment group corresponding to the via is determined from all candidate line segments.

[0057] In some implementation scenarios, a first candidate line segment is selected from all candidate line segments corresponding to the first fork arm, and a second candidate line segment is selected from all candidate line segments corresponding to the second fork arm. The distance and angle difference between the first and second candidate line segments are obtained, and the coordinate component difference of the endpoints of the first and second candidate line segments closest to the handling robot is calculated. In response to the following conditions: the difference between the distance between the first and second candidate line segments and the width of the through hole is less than a distance threshold; the angle difference between the first and second candidate line segments is less than an angle threshold; and the coordinate component difference of the endpoints of the first and second candidate line segments closest to the handling robot is less than a coordinate component threshold, then the first and second candidate line segments are constructed as a reference line segment group corresponding to the through hole. The aforementioned endpoint coordinate component difference can be the height difference between the endpoints of the first and second candidate line segments closest to the handling robot, i.e., the coordinate component difference along the Z-axis perpendicular to the horizontal plane, or it can be other coordinate component differences, such as the coordinate component difference corresponding to the X-axis or Y-axis. It should be noted that the distance threshold is mentioned multiple times in this application; in different implementations, the distance threshold can be the same or different.

[0058] Specifically, the equations of the first and second candidate line segments in the robot coordinate system are obtained. Based on the equations, the distance and angle difference between them are calculated. Furthermore, the coordinate component difference is determined based on the coordinate components of the endpoints of the first and second candidate line segments closest to the handling robot on the Z-axis.

[0059] In some implementation scenarios, geometric relationship information may only include a portion of the distance, angle difference, and endpoint coordinate component difference. For example, when the geometric relationship information includes distance and angle difference, if the difference between the distance between the first candidate line segment and the second candidate line segment and the via width is less than a distance threshold, and the angle difference between the first candidate line segment and the second candidate line segment is less than an angle threshold, then the first candidate line segment and the second candidate line segment are constructed as a reference line segment group corresponding to the via.

[0060] For a specific application scenario, please refer to Figure 6 , Figure 6 This is a schematic diagram of one embodiment of the target stack of this application. Figure 6 The target pallet shown has a first side bottom beam and a second side bottom beam at its bottom, forming a unique through hole between them. This through hole is the fork joint area of ​​the target pallet. Any first candidate line segment is selected from all candidate line segments corresponding to the first fork arm, and any second candidate line segment is selected from all candidate line segments corresponding to the second fork arm. In response to the following conditions: the difference between the distance between the first and second candidate line segments and the through hole width H1 is less than a distance threshold; the angle difference between the first and second candidate line segments is less than an angle threshold; and the difference in the coordinate components of the endpoints of the first and second candidate line segments closest to the handling robot is less than a coordinate component threshold, the first and second candidate line segments are constructed as a reference line segment group corresponding to the through hole. Figure 6 The first reference line segment and the second reference line segment are shown in the figure.

[0061] In one embodiment, after selecting any first candidate line segment from all candidate line segments corresponding to the first fork arm and any second candidate line segment from all candidate line segments corresponding to the second fork arm, the linear equations of the first and second candidate line segments in the robot coordinate system are obtained. The angle difference between the first and second candidate line segments can also be the parallel deviation angle calculated based on the coefficients of the linear equation. In response to the distance between the first and second candidate line segments being less than a distance threshold, the parallel deviation angle being less than an angle threshold, and the difference between the endpoint coordinate components being less than a coordinate component threshold, the first and second candidate line segments are constructed as a reference line segment group corresponding to the through hole. The parallel deviation angle is calculated by the following formula (2).

[0062] (2) in, This represents the parallel deviation angle between the first candidate line segment and the second candidate line segment, i.e., the angle difference. and The coefficients of the equation of the first candidate line segment are represented. and This represents the coefficients of the straight line equation for the second candidate line segment.

[0063] S502: Based on the distance between different reference line segments, determine the first inner edge line and the second inner edge line corresponding to the crossover area formed by all through holes.

[0064] In one embodiment, at least one reference line segment is selected from the reference line segment group corresponding to each through hole, and the difference between the distance between the two selected reference line segments and the width of the crossover area is obtained. If the absolute value of the difference is less than a preset threshold, the two reference line segments are used as the first inner edge line and the second inner edge line corresponding to the crossover area.

[0065] For a specific application scenario, please refer to [link / reference]. Figure 6 Since a unique through-hole is formed between the first and second side bottom beams of the target pallet, this through-hole constitutes the cross-joining area of ​​the target pallet, and the width H1 of the through-hole is the width of the cross-joining area. After determining the first and second reference line segments corresponding to the unique through-hole through the above-described corresponding implementation method, it is further determined whether the absolute value of the difference between the distance between the first and second reference line segments and the width H1 of the cross-joining area is less than a preset threshold. If so, it indicates that the distance between the first and second reference line segments is close to the actual width of the cross-joining area, and the first and second reference line segments constitute the first inner edge line and the second inner edge line corresponding to the cross-joining area.

[0066] The above solution, by combining at least one of the distance, angle difference, and endpoint coordinate component difference, selects a group of reference line segments corresponding to the bottom edge of each through hole from all candidate line segments, and further selects the first inner edge line and the second inner edge line corresponding to the crossover area from all reference line segment groups. This helps to reduce the difficulty of locating the target pallet in the target scene and improve the efficiency of pallet handling.

[0067] Please see Figure 7 and Figure 8 , Figure 7 yes Figure 5 The flowchart of step S501 corresponds to another embodiment. Figure 8 This is a schematic diagram of another embodiment of the target stack of this application. Figure 8The target pallet shown includes a first side bottom beam, a central bottom beam, and a second side bottom beam, with a first through hole formed between the first side bottom beam and the central bottom beam, and a second through hole formed between the second side bottom beam and the central bottom beam. Based on this, the specific implementation process of step S501 includes: S601: Select a first reference line segment and a second reference line segment corresponding to the first through hole from all candidate line segments; wherein, the angle difference between the first reference line segment and the second reference line segment is less than an angle threshold, the difference between the distance between the first reference line segment and the second reference line segment and the width of the first through hole is less than a distance threshold, and the difference between the coordinate components of the endpoints of the first reference line segment and the second reference line segment that are close to the handling robot is less than a coordinate component threshold.

[0068] In one embodiment, a first candidate line segment and a second candidate line segment are arbitrarily selected from all candidate line segments. The angle difference, distance, and endpoint coordinate component difference between the selected first candidate line segment and the second candidate line segment are obtained. In response to the absolute value of the difference between the distance between the first candidate line segment and the second candidate line segment and the width of the first through hole being less than a distance threshold, the angle difference being less than an angle threshold, and the endpoint coordinate component difference being less than a coordinate component threshold, the first candidate line segment and the second candidate line segment are designated as the first reference line segment and the second reference line segment corresponding to the first through hole. The specific methods for obtaining the angle difference, distance, and endpoint coordinate component difference in this embodiment can be referred to the corresponding embodiments described above, and will not be elaborated upon here.

[0069] S602: Select the third reference line segment and the fourth reference line segment corresponding to the second through hole from all candidate line segments; wherein, the angle difference between the third reference line segment and the fourth reference line segment is less than the angle threshold, the difference between the distance between the third reference line segment and the fourth reference line segment and the width of the second through hole is less than the distance threshold, and the difference between the coordinate components of the endpoints of the third reference line segment and the fourth reference line segment that are close to the handling robot is less than the coordinate component threshold.

[0070] In one embodiment, a third candidate line segment and a fourth candidate line segment are arbitrarily selected from all candidate line segments, and the angle difference, distance, and endpoint coordinate component difference between the selected third candidate line segment and the fourth candidate line segment are obtained. In response to the absolute value of the difference between the distance between the third candidate line segment and the fourth candidate line segment and the width of the second via being less than a distance threshold, the angle difference being less than an angle threshold, and the endpoint coordinate component difference being less than a coordinate component threshold, the third candidate line segment and the fourth candidate line segment are designated as the third reference line segment and the fourth reference line segment corresponding to the second via.

[0071] It should be noted that in practical applications, some target pallets include multiple spaced central bottom beams, and the area between adjacent central bottom beams is small enough to accommodate the first or second fork arm. In such cases, the multiple central bottom beams are combined and regarded as one central bottom beam.

[0072] The above solution, for a cross-joint area including multiple through holes, helps to improve the accuracy of determining the first inner edge line and the second inner edge line corresponding to the cross-joint area by screening a group of reference line segments corresponding to the bottom edge line inside each through hole from all candidate line segments.

[0073] Please see Figure 9 , Figure 9 yes Figure 5 The flowchart before step S501 corresponds to another embodiment. Specifically, the process before step S501 includes: S701: Obtain the first reference distance between any two candidate line segments corresponding to the first fork arm, and the second reference distance between any two candidate line segments corresponding to the second fork arm.

[0074] In one embodiment, for all candidate line segments corresponding to the target radar data acquired by the first fork arm, any two candidate line segments are selected, and a first reference distance between them is calculated. Similarly, for all candidate line segments corresponding to the target radar data acquired by the second fork arm, any two candidate line segments are selected, and a second reference distance between them is calculated.

[0075] S702: Based on the difference between the first reference distance and the first actual distance and the difference between the second reference distance and the second actual distance, determine the positioning relationship of all candidate line segments; wherein, when the positioning relationship of at least some candidate line segments meets the preset distance condition, perform the step of determining the reference line segment group corresponding to each through hole based on the geometric relationship information between different candidate line segments and the through hole width.

[0076] In one embodiment, a preset first actual distance and a preset second actual distance are obtained. The absolute value of the difference between the first reference distance and the first actual distance, and the absolute value of the difference between the second reference distance and the second actual distance are determined. The positioning relationship between all candidate line segments is determined based on the absolute values ​​of these differences. When the positioning relationship of some candidate line segments meets the preset distance condition, the step of determining the reference line segment group corresponding to each through hole based on the geometric relationship information between different candidate line segments and the through hole width is performed. The first actual distance and the second actual distance can be obtained through estimation or by back-calculation through multiple trials.

[0077] In one embodiment, a first actual distance and a second actual distance, determined based on the size information of the target pallet, are acquired. The absolute value of the difference between the first reference distance and the first actual distance, and the absolute value of the difference between the second reference distance and the second actual distance, are determined. The positioning relationship between all candidate line segments is determined based on these absolute values. When the positioning relationship of some candidate line segments meets a preset distance condition, the step of determining a reference line segment group corresponding to each through-hole based on the geometric relationship information between different candidate line segments and the through-hole width is performed. The aforementioned size information includes at least the width of the bottom beam of the target pallet.

[0078] In some implementation scenarios, the target pallet includes a first side bottom beam, a central bottom beam, and a second side bottom beam. During the handling process, the first fork corresponds to a first through hole formed between the first side bottom beam and the central bottom beam, and the second fork corresponds to a second through hole formed between the second side bottom beam and the central bottom beam. A first actual distance is determined based on the width of the first side bottom beam, the width of the central bottom beam, and the width of the first through hole, and a second actual distance is determined based on the width of the central bottom beam, the width of the second side bottom beam, and the width of the second through hole. It is determined whether the absolute value of the difference between the first reference distance and the first actual distance is within a preset value range, and whether the absolute value of the difference between the second reference distance and the second actual distance is within a preset value range. If so, it indicates that the first fork and the second fork have respectively collected candidate line segments corresponding to the bottom edges inside the two through holes, that is, at least some candidate line segments satisfy the preset distance condition. Return to the step of determining the reference line segment group, wherein the specific determination process of the reference line segment group can refer to the corresponding implementation method described above.

[0079] For a specific application scenario, please refer to [link / reference]. Figure 8 , Figure 8 The target pallet shown has a first side bottom beam width of w1, a central bottom beam width of w2, a second side bottom beam width of w3, a first through hole width of h1, and a second through hole width of h2. A first actual distance m1 is determined based on the widths w1, w2, and h1 of the first side bottom beam and the central bottom beam; and a second actual distance m2 is determined based on the widths w2, w3, and h2 of the central bottom beam, the second side bottom beam, and the second through hole. If the absolute value of the difference between the first reference distance and the first actual distance between the two candidate line segments corresponding to the first fork arm is within a preset numerical range, it indicates that the measured distance between the two candidate line segments is close to the first actual distance, meaning that the two candidate line segments are the bottom edges within the first through hole. Similarly, if the absolute value of the difference between the second reference distance and the second actual distance between the two candidate line segments corresponding to the second fork arm is within a preset numerical range, then the two candidate line segments are the bottom edges within the second through hole. The first actual distance m1 and the second actual distance m2 are calculated using the following formulas (3) and (4).

[0080] (3) (4) The above scheme determines whether the preset distance conditions are met between different candidate line segments by targeting the through holes corresponding to the first and second forks, thereby judging whether the target radar data acquisition is accurate and eliminating interference from erroneous radar data acquisition on subsequent path planning, thus further improving the stability of the generated target path.

[0081] Additionally, it should be noted that when the target pallet only includes the first side bottom beam and the second side bottom beam, the process of determining whether the positioning relationship of some candidate line segments meets the preset distance condition before step S501 includes: obtaining the first reference distance between any two candidate line segments from all candidate line segments corresponding to the first fork arm and the second fork arm, and determining the positioning relationship of all candidate line segments based on the difference between the first reference distance and the first actual distance. When the positioning relationship of at least some candidate line segments meets the preset distance condition, the step of determining the reference line segment group corresponding to each through hole based on the geometric relationship information between different candidate line segments and the through hole width is executed. Among them, the first actual distance is determined according to the width of the first side bottom beam, the width of the second side bottom beam and the through hole width, and the specific calculation formula can be referred to the above formula (3) or formula (4).

[0082] Please see Figure 10 , Figure 10 yes Figure 1 The flowchart of step S104 corresponds to another embodiment. Specifically, the implementation process of step S104 includes: S801: Determine the center coordinates of the fork-in surface in the robot coordinate system based on the first inner edge line and the second inner edge line.

[0083] In one embodiment, a first target straight line is determined between the first inner edge line and the second inner edge line, wherein the opposite direction of the first target straight line is the average of the directions of the first and second inner edge lines. A second target straight line is determined passing through the endpoint of either the first or second inner edge line near the transport robot, and the second target straight line is perpendicular to the first target straight line. The intersection of the first and second target straight lines is taken as the center point of the target pallet fork-in surface. The center coordinates of the center point in the robot coordinate system are obtained.

[0084] In some implementation scenarios, the equation coefficients of the straight line corresponding to the first inner edge are a1, a2 and a3, and the equation coefficients of the straight line corresponding to the second inner edge are b1, b2 and b3. Then the equation coefficients and angle of the first target straight line are obtained by the following formulas (5) to (8).

[0085] (5) (6) (7) (8) in, , and The coefficients of the equation for the first target line are represented. This represents the angle of the first target line.

[0086] S802: Based on the center coordinates, determine the coordinates of the transport endpoint corresponding to the target pallet in the spatial coordinate system.

[0087] In one embodiment, the depth of the forking region extending along the fork-in direction of the transport robot's fork arm is obtained. Based on the depth of the forking region and the actual depth of the fork-in region at the target time, the reference relative pose from the center point of the fork-in surface to the plane containing the next forking task point within the forking region is determined. The target pose of the transport robot at the target time is obtained, and based on the center coordinates in the robot coordinate system, the target pose, and the reference relative pose, the coordinates of the transport endpoint corresponding to the target pallet in the spatial coordinate system are determined.

[0088] In some implementation scenarios, the coordinates of the transport endpoint corresponding to the target pallet in the spatial coordinate system are calculated by the following formula (9).

[0089] (9) in, Indicates the coordinates of the destination of the transport. This represents the center coordinates in the robot's coordinate system. This indicates the target pose of the transport robot. This indicates the reference relative pose.

[0090] S803: Generate the target path based on the coordinates of the transport endpoint.

[0091] In one embodiment, a transport path is planned based on the coordinates of the transport endpoint to generate a target path for the transport robot to transport the target pallet, and the transport robot is driven to run along the target path.

[0092] In some implementation scenarios, to improve the accuracy of handling tasks, radar data is collected in real time using the first and second forks during the handling process of the handling robot, and any of the above-described implementation methods are executed cyclically to make real-time corrections to the target path. During real-time correction, the reference relative pose is adjusted in real time according to the actual depth of the fork insertion area of ​​the handling robot. For example, when the fork has inserted 2 / 3 of the depth of the insertion area, the next insertion point is determined as the position of the last side of the target pallet.

[0093] The above solution, by executing the pallet handling method proposed in this application in real time during the handling robot's handling task to make real-time corrections to the target path, helps to improve the accuracy and stability of the handling task.

[0094] Please see Figure 11 , Figure 11 This is a schematic diagram of one embodiment of the electronic device of this application. The electronic device includes a memory 10 and a processor 20 coupled to each other. The memory 10 stores program instructions, and the processor 20 executes the program instructions to implement the methods mentioned in any of the above embodiments. Specifically, the electronic device includes, but is not limited to, desktop computers, laptops, tablets, servers, etc., and is not limited thereto. In addition, the processor 20 may also be called a CPU (Center Processing Unit). The processor 20 may be an integrated circuit chip with signal processing capabilities. The processor 20 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor may be a microprocessor or any conventional processor. In addition, the processor 20 may be implemented by integrated circuit chips.

[0095] Please see Figure 12 , Figure 12 This is a schematic diagram of a computer-readable storage medium according to an embodiment of the present application. The computer-readable storage medium 30 stores program instructions 40 that can be executed by a processor. When the program instructions 40 are executed by the processor, they implement the methods mentioned in any of the above embodiments.

[0096] In the several embodiments provided in this application, it should be understood that the disclosed methods and apparatus can be implemented in other ways. For example, the apparatus implementations described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0097] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0098] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0099] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0100] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for transporting pallets, characterized in that, The method utilizes a handling robot including a first fork and a second fork, and the method includes: For the fork-in surface of the target pallet, acquire the target radar data collected by the first fork arm and the second fork arm respectively; After clustering, obtain multiple candidate line segments corresponding to each group of target radar data; Based on the distance between the different candidate line segments, a first inner edge line and a second inner edge line that match the crossover area of ​​the target pallet are determined; Based on the first inner edge line and the second inner edge line, a target path is generated for the transport robot to transport the target pallet.

2. The pallet handling method according to claim 1, characterized in that, The step of acquiring target radar data collected by the first fork and the second fork respectively, targeting the fork-in surface of the target pallet, includes: Acquire reference radar data of the target pallet collected by the first fork arm and the second fork arm at the target time; The initial position of the target pallet to be transported in the target scene is obtained in the spatial coordinate system, and the initial position is transformed to the robot coordinate system to obtain the reference position of at least part of the fork joint area in the robot coordinate system; wherein, the initial position is determined based on the coordinates of the reference point and the reference angle in the fork joint surface; Based on the reference position, the target radar data corresponding to the crossing area is extracted from each set of reference radar data.

3. The pallet handling method according to claim 2, characterized in that, The step of acquiring reference radar data of the target pallet collected by the first fork arm and the second fork arm at the target time includes: Acquire multiple sets of initial radar data within the target scene collected by the first fork arm and the second fork arm respectively; Based on the timestamp of the initial radar data, the reference radar data collected by the first fork arm and the second fork arm at the target time are determined respectively.

4. The pallet handling method according to claim 1, characterized in that, The acquisition of multiple candidate line segments corresponding to each group of target radar data after clustering processing includes: Cluster the target radar data collected by the first fork arm and the second fork arm respectively to obtain the clustering result corresponding to each group of target radar data, and obtain the initial line segment corresponding to each clustering result; Obtain the length of each initial line segment and the reference angle between the initial line segment and the moving direction of the handling robot. Based on the line segment length and the reference angle, select multiple candidate line segments from all the initial line segments corresponding to the first fork arm and the second fork arm respectively.

5. The pallet handling method according to claim 1, characterized in that, The crossover region includes at least one through-hole. Determining the first inner edge line and the second inner edge line matching the crossover region of the target pallet based on the distance between different candidate line segments includes: Based on the geometric relationship information between different candidate line segments and the through-hole width, a reference line segment group corresponding to each through-hole is determined; wherein, the geometric relationship information includes at least one of distance, angle difference, and endpoint coordinate component difference; Based on the distance between the different reference line segments, the first inner edge line and the second inner edge line corresponding to the crossover area formed by all the through holes are determined.

6. The pallet handling method according to claim 5, characterized in that, The target pallet includes a first side bottom beam, a central bottom beam, and a second side bottom beam. A first through hole is formed between the first side bottom beam and the central bottom beam, and a second through hole is formed between the second side bottom beam and the central bottom beam. The step of determining a reference segment group corresponding to each of the candidate line segments and the via width includes: From all the candidate line segments, select a first reference line segment and a second reference line segment corresponding to the first through hole; wherein, the angle difference between the first reference line segment and the second reference line segment is less than an angle threshold, the difference between the distance between the first reference line segment and the second reference line segment and the width of the first through hole is less than a distance threshold, and the difference between the coordinate components of the endpoints of the first reference line segment and the second reference line segment closest to the handling robot is less than a coordinate component threshold; and... Select a third reference line segment and a fourth reference line segment corresponding to the second through hole from all the candidate line segments; wherein, the angle difference between the third reference line segment and the fourth reference line segment is less than an angle threshold, the difference between the distance between the third reference line segment and the fourth reference line segment and the width of the second through hole is less than a distance threshold, and the difference between the coordinate components of the endpoints of the third reference line segment and the fourth reference line segment closest to the handling robot is less than a coordinate component threshold.

7. The pallet handling method according to claim 5, characterized in that, Before determining the reference segment group corresponding to each through hole based on the geometric relationship information between different candidate line segments and the through hole width, the process includes: Obtain the first reference distance between any two candidate line segments corresponding to the first fork arm, and the second reference distance between any two candidate line segments corresponding to the second fork arm; Based on the difference between the first reference distance and the first actual distance, and the difference between the second reference distance and the second actual distance, the positioning relationship of all the candidate line segments is determined; wherein, when the positioning relationship of at least some of the candidate line segments meets the preset distance condition, the step of determining the reference line segment group corresponding to each through hole based on the geometric relationship information between different candidate line segments and the through hole width is executed.

8. The pallet handling method according to claim 1, characterized in that, The step of generating a target path for the transport robot to transport the target pallet based on the first inner edge line and the second inner edge line includes: Based on the first inner edge line and the second inner edge line, determine the center coordinates of the fork entry surface in the robot coordinate system; Based on the center coordinates, determine the transport endpoint coordinates corresponding to the target pallet in the spatial coordinate system; The target path is generated based on the coordinates of the transport endpoint.

9. An electronic device, characterized in that, include: A memory and a processor are coupled to each other, the memory storing program instructions, and the processor executing the program instructions to implement the method as described in any one of claims 1-8.

10. A computer-readable storage medium having program instructions stored thereon, characterized in that, When the program instructions are executed by the processor, they implement the method as described in any one of claims 1-8.

Images