A method and apparatus for laser cleaning of a rail-type robot

Abstract

The application provides a guide rail type mechanical arm laser cleaning method and device, and relates to the technical field of robot automation. The method comprises the following steps: acquiring at least one cleaning area determined in advance for a to-be-cleaned workpiece and a target pose of a mechanical arm at each cleaning point of the to-be-cleaned workpiece; generating a plurality of random stations for each cleaning area, and for each station, determining the coverage rate of the current station at the corresponding cleaning point in the cleaning area to which the current station belongs according to the pose of the mechanical arm at the current station and the target pose of the mechanical arm at each cleaning point in the cleaning area to which the current station belongs, wherein the station is the position of the mechanical arm on the guide rail; screening the target station corresponding to each cleaning area from the plurality of stations according to the coverage rate of the cleaning point; and controlling the mechanical arm to clean the to-be-cleaned workpiece according to the target station. The method can effectively solve the problems of low cleaning efficiency or high cleaning cost in the prior art.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of robot automation, and in particular to a guide rail type mechanical arm laser cleaning method and device. BACKGROUND

[0002] In the field of robot automation cleaning, compared with the traditional handheld scheme, the method of using a laser carried by the end of a mechanical arm to clean can accurately control the distance between the mechanical arm and the cleaning surface, and the cleaning is uniform and effective, and is widely used in the cleaning industry of metal plates. For small workpieces, robots usually use a fixed base installation method, and through reasonable placement, the robot can cover the entire cleaning area. However, for large workpieces (such as large parts in the aerospace industry), a fixed mechanical arm cannot cover the entire cleaning area. To solve this problem, the following methods are usually used in the prior art: first, the mechanical arm is installed on a movable trolley, and experienced workers specify fixed stations for cleaning; second, a flow line is used, and the mechanical arm is still fixed, but multiple lasers are used for cleaning. However, the first method requires human experience and has low work efficiency, and the second method requires multiple devices and has high cleaning cost. SUMMARY

[0003] The present application provides a guide rail type mechanical arm laser cleaning method and device to solve the problem of low cleaning efficiency or high cleaning cost of the mechanical arm cleaning operation in the prior art.

[0004] The present application provides a guide rail type mechanical arm laser cleaning method, comprising the following steps:

[0005] Obtaining at least one cleaning area determined in advance for a workpiece to be cleaned, and a target pose of a mechanical arm at each cleaning point of the workpiece to be cleaned;

[0006] Generating a plurality of random stations for each cleaning area, and for each station, determining the coverage rate of the cleaning point corresponding to the station in the cleaning area to which the station belongs according to the pose of the mechanical arm at the current station and the target pose of the mechanical arm at each cleaning point in the cleaning area to which the station belongs, the station being the position of the mechanical arm on the guide rail;

[0007] According to the coverage rate of the cleaning point, the target station corresponding to each cleaning area is selected from the plurality of stations;

[0008] According to the target station, controlling the mechanical arm to clean the workpiece to be cleaned.

[0009] According to the rail type mechanical arm laser cleaning method provided in the application, for each station, according to the pose of the mechanical arm at the current station and the target poses of the cleaning points in the cleaning area to which the current station belongs, the coverage of the cleaning points corresponding to the current station in the cleaning area is determined, comprising:

[0010] For each station, according to the pose of the mechanical arm at the current station and the target poses of the cleaning points in the cleaning area to which the current station belongs, the cleaning points that can be reached by the end effector of the mechanical arm in the cleaning area to which the current station belongs are determined.

[0011] According to the ratio of the reachable cleaning points to the total amount of cleaning points in the cleaning area to which the current station belongs, the coverage of the cleaning points corresponding to the current station in the cleaning area is determined.

[0012] According to the rail type mechanical arm laser cleaning method provided in the application, for each station, according to the pose of the mechanical arm at the current station and the target poses of the cleaning points in the cleaning area to which the current station belongs, the cleaning points that can be reached by the end effector of the mechanical arm in the cleaning area to which the current station belongs are determined, comprising:

[0013] The initial pose of the mechanical arm is obtained;

[0014] The interpolation trajectory of the mechanical arm from the initial pose to each target pose through each station is determined, and the stations with collision risk on the interpolation trajectory are removed from the plurality of stations to obtain remaining stations;

[0015] For each station in the remaining stations, according to the pose of the mechanical arm at the current station and the target poses of the cleaning points in the cleaning area to which the current station belongs, the cleaning points that can be reached by the end effector of the mechanical arm in the cleaning area to which the current station belongs are determined.

[0016] According to the rail type mechanical arm laser cleaning method provided in the application, each cleaning point in the cleaning area to which the current station belongs includes a target cleaning point, and the cleaning points that can be reached by the end effector of the mechanical arm in the cleaning area to which the current station belongs are determined according to the pose of the mechanical arm at the current station and the target poses of the cleaning points in the cleaning area to which the current station belongs, comprising:

[0017] The relative pose of the target pose of the target cleaning point in the cleaning area to which the current station belongs and the pose of the mechanical arm at the current station is determined.

[0018] Solving joint angles of the mechanical arm according to the relative pose through an inverse kinematics model of the mechanical arm;

[0019] If it is determined that at least a set of joint angles can be solved, it is determined that the end effector of the mechanical arm can reach the target cleaning point in the cleaning area to which the current station belongs.

[0020] According to the guide rail type mechanical arm laser cleaning method provided in the present application, the inverse kinematics model is determined through the following steps:

[0021] Obtaining a unified robot description format file of the mechanical arm;

[0022] According to the unified robot description format file, obtaining the geometric parameters and kinematics parameters of the mechanical arm;

[0023] According to the geometric parameters and the kinematics parameters, constructing an inverse kinematics model of the mechanical arm.

[0024] According to the guide rail type mechanical arm laser cleaning method provided in the present application, the station positions with collision risks on the interpolation trajectory are removed from the plurality of station positions to obtain the remaining station positions, including:

[0025] Obtaining a bounding box hierarchy tree corresponding to the mechanical arm, and a plurality of directional bounding box models are stored in the bounding box hierarchy tree, and the plurality of directional bounding box models are generated in advance for each joint, connecting rod, end effector of the mechanical arm and fixed obstacles in a working area where the mechanical arm is located;

[0026] Based on the bounding box hierarchy tree, a flexible collision detection library is used to determine the station positions with collision risks on the interpolation trajectory;

[0027] The station positions with collision risks are removed from the plurality of station positions to obtain the remaining station positions.

[0028] According to the guide rail type mechanical arm laser cleaning method provided in the present application, the cleaning area includes adjacent first and second cleaning areas, and an overlapping area is arranged between the first and second cleaning areas, and the overlapping area includes at least one cleaning point in the first cleaning area and at least one cleaning point in the second cleaning area, and for each station position, the coverage rate of the corresponding cleaning point in the cleaning area to which the current station belongs is determined according to the pose of the mechanical arm at the current station and the target pose of each cleaning point in the cleaning area to which the current station belongs, including:

[0029] For each of the plurality of stations generated for the first cleaning area, according to the pose of the mechanical arm at the current station and the target pose of the mechanical arm at each cleaning point in the first cleaning area and the overlap area, determine the coverage of the corresponding cleaning point under the first cleaning area at the current station;

[0030] For each of the plurality of stations generated for the second cleaning area, according to the pose of the mechanical arm at the current station and the target pose of the mechanical arm at each cleaning point in the second cleaning area and the overlap area, determine the coverage of the corresponding cleaning point under the second cleaning area at the current station.

[0031] According to the target station, the method for laser cleaning of the guide rail type mechanical arm provided by the present application controls the mechanical arm to clean the workpiece to be cleaned, which comprises:

[0032] Based on the travel constraint information of the guide rail and the distribution constraint information of the station, a target station sequence is generated according to the target stations;

[0033] Determine the cleaning action sequence of the mechanical arm at each target station;

[0034] According to the target station sequence, the mechanical arm is controlled to move on the guide rail, and according to the cleaning action sequence, the mechanical arm is controlled to clean the workpiece to be cleaned.

[0035] According to the guide rail type mechanical arm laser cleaning method provided by the present application, the at least one cleaning area is determined by the following steps:

[0036] Obtain the boundary constraint information of the cleaning area input by the user;

[0037] According to the boundary constraint information and the target pose of the mechanical arm at each cleaning point of the workpiece to be cleaned, determine the at least one cleaning area.

[0038] The present application also provides a guide rail type mechanical arm laser cleaning device, which comprises the following modules:

[0039] The obtaining module is used to obtain at least one cleaning area determined in advance for a workpiece to be cleaned, and the target pose of a mechanical arm at each cleaning point of the workpiece to be cleaned;

[0040] The determining module is used to generate a plurality of random stations for each cleaning area, and for each station, according to the pose of the mechanical arm at the current station and the target pose of the mechanical arm at each cleaning point in the cleaning area to which the current station belongs, determine the coverage of the corresponding cleaning point under the cleaning area to which the current station belongs, and the station is the position of the mechanical arm on the guide rail;

[0041] a screening module configured to screen target stations corresponding to each cleaning area from the plurality of stations according to coverage of the cleaning points;

[0042] a control module configured to control the robot arm to clean the workpiece to be cleaned according to the target stations.

[0043] The application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the laser cleaning method of the rail robot arm according to any one of the above.

[0044] The application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and the computer program is executable by a processor to implement the laser cleaning method of the rail robot arm according to any one of the above.

[0045] The application also provides a computer program product, including a computer program, and the computer program is executable by a processor to implement the laser cleaning method of the rail robot arm according to any one of the above.

[0046] The laser cleaning method of the rail robot arm provided by the application first acquires at least one cleaning area determined in advance for a workpiece to be cleaned, and target poses of a robot arm at cleaning points of the workpiece to be cleaned; then generates a plurality of random stations for each cleaning area, and for each station, determines coverage of a cleaning point corresponding to the current station in the cleaning area to which the current station belongs according to a pose of the robot arm at the current station and target poses of the robot arm at each cleaning point in the cleaning area to which the current station belongs; then screens target stations corresponding to each cleaning area from the plurality of stations according to the coverage of the cleaning points; and finally controls the robot arm to clean the workpiece to be cleaned according to the target stations. In the application, the cleaning of the workpiece is no longer performed by manually specifying fixed stations, which can significantly improve the efficiency of the cleaning operation. In addition, the same robot arm can move on the rail according to a plurality of target stations planned in advance and perform the cleaning operation, without the need to carry multiple lasers, that is, without the need to configure more cleaning equipment, which can significantly reduce the cost of the cleaning operation. Therefore, the method of the application can effectively solve the problems of low cleaning efficiency or high cleaning cost in the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0048] Figure 1 FIG. 1 is a flowchart of a guide rail type mechanical arm laser cleaning method according to an embodiment of the present application.

[0049] Figure 2 FIG. 2 is a cleaning area division result schematic diagram according to an embodiment of the present application.

[0050] Figure 3 FIG. 3 is a coordinate system change schematic diagram according to an embodiment of the present application.

[0051] Figure 4 FIG. 4 is a collision detection algorithm flowchart according to an embodiment of the present application.

[0052] Figure 5 FIG. 5 is a mechanical arm cleaning operation overall flowchart according to an embodiment of the present application.

[0053] Figure 6 FIG. 6 is a guide rail type mechanical arm laser cleaning device structure block diagram according to an embodiment of the present application.

[0054] Figure 7 FIG. 7 is an electronic device physical structure schematic diagram according to an embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, but not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0056] In order to solve the defects of low cleaning efficiency or high cleaning cost of the existing mechanical arm cleaning operation scheme, the present application provides a guide rail type mechanical arm laser cleaning method. The execution subject of the method can be any control device with data processing and communication functions. The control device first plans a guide rail station position suitable for the mechanical arm, and then controls the mechanical arm to move on the guide rail according to the planned guide rail station position and clean the workpiece to be cleaned.

[0057] Figure 1 FIG. 1 is a flowchart of a guide rail type mechanical arm laser cleaning method according to an embodiment of the present application. With reference to Figure 1 , the method of the present application can include the following steps:

[0058] Step 101, obtaining at least one cleaning area determined in advance for the workpiece to be cleaned, and the target pose of the mechanical arm at each cleaning point of the workpiece to be cleaned.

[0059] The type of the workpiece to be cleaned can be a small workpiece or a large workpiece.

[0060] The number of workpieces to be cleaned can be set according to actual needs. For example, when the workpieces to be cleaned are multiple small workpieces, all the small workpieces can be placed together for cleaning, so that multiple workpieces can be cleaned at the same time, thereby improving the cleaning efficiency.

[0061] In the present application, at least one cleaning area can be determined in advance by the following method:

[0062] obtaining boundary constraint information of a cleaning area input by a user;

[0063] determining at least one cleaning area according to the boundary constraint information and target poses of the robot arm at each cleaning point of the workpiece to be cleaned.

[0064] In the present application, a user can determine constraint information that needs to be met by the boundary of the entire cleaning work area according to various factors such as the range of the area available for movement of the robot arm and the range occupied by the workpiece to be cleaned, and input the boundary constraint information to the control device.

[0065] Secondly, the user also needs to design a cleaning process point set of the workpiece to be cleaned in advance, which records the positions of all cleaning points in the workpiece to be cleaned and the poses that the robot arm should assume when reaching each cleaning point. In the following text, the position and pose information will be collectively referred to as pose.

[0066] In the present application, a cleaning point refers to a point on the workpiece to be cleaned that needs to be cleaned. These point positions are determined in advance and recorded in the cleaning process point set. A cleaning area refers to an area formed by at least one cleaning point on the workpiece to be cleaned.

[0067] Next, the control device reads the cleaning process point set and the boundary constraint information, and automatically plans at least one cleaning area in the cleaning work area according to the boundary constraint information and the positions of all cleaning points recorded in the cleaning process point set.

[0068] In the present application, the control device divides the fixed guide rail length into regions according to the distribution of cleaning points and the boundary constraint of the cleaning area divided by humans, and the division effect can be as shown in Figure 2 . Figure 2 is a cleaning area division result schematic diagram according to an embodiment of the present application. In Figure 2 , the four irregular shaded parts represent the cleaning areas obtained by division, and the two vertical dashed lines on the left and right represent the boundary constraint information. The area between the two vertical dashed lines represents the entire cleaning work area. In Figure 2 , the arrow to the right represents the moving direction of the guide rail.

[0069] In the present application, the control device dynamically and adaptively divides the entire cleaning work area, which means that the cleaning of the workpiece is no longer dependent on the fixed number of stations, but is realized through an intelligent processing flow: first, the cleaning process point set and the user-set boundary constraint are received, then a small-to-large division strategy is adopted, and the cleaning work area is preliminarily divided into at least one cleaning region (each cleaning region corresponds to potentially multiple stations) according to the preset boundary. Then, it is checked whether each cleaning region obtained by division contains at least two cleaning points to verify its effectiveness, and the cleaning region containing no cleaning point is regarded as an invalid cleaning region and removed. Finally, the remaining cleaning regions are taken as the at least one cleaning region in step 101, and the stations will also be planned according to the remaining cleaning regions. The method of the present application is especially suitable for workpieces with large size differences (such as large single parts) or complex layouts (such as small multiple parts), and can significantly improve the efficiency and resource utilization of the robot cleaning operation.

[0070] Step 102, generate a plurality of random stations for each cleaning region, and for each station, determine the coverage rate of the corresponding cleaning point in the cleaning region to which the current station belongs according to the pose of the robot at the current station and the target pose of each cleaning point in the cleaning region to which the current station belongs. The station is the position of the robot on the guide rail.

[0071] In the present application, since there can be multiple stations when the robot cleans the cleaning points in a cleaning region, in order to find the best station, the present application first randomly generates a plurality of stations for each cleaning region. Then, according to whether the pose of the robot at a certain station can reach the target pose of the robot at each cleaning point in the cleaning region to which the station belongs, the coverage rate of the corresponding cleaning point in the cleaning region to which the station belongs is counted, i.e. the reachability verification of the cleaning point is performed for each generated random station, and then the coverage rate of the cleaning point is obtained.

[0072] For example, cleaning region one has four cleaning points, including cleaning points one to four, and two random candidate stations one and two are designed for cleaning region one. Then, for candidate station one, assuming that the robot is located at candidate station one on the guide rail, it is checked whether the robot at the pose of candidate station one can smoothly reach the target pose of each of cleaning points one to four, and assuming that it can reach the target pose of each of cleaning points one and two, then the coverage rate of the corresponding cleaning point in cleaning region one for candidate station one is 50%. Similarly, for candidate station two, assuming that the robot is located at candidate station two on the guide rail, it is checked whether the robot at the pose of candidate station two can smoothly reach the target pose of each of cleaning points one to four, and assuming that it can reach the target pose of each of cleaning points one to three, then the coverage rate of the corresponding cleaning point in cleaning region one for candidate station two is 75%.

[0073] Step 103, according to the coverage of the cleaning point, screening the target station corresponding to each cleaning area in multiple stations.

[0074] In the present application, the higher the coverage of the cleaning point corresponding to a station, the better the station. Therefore, for multiple stations corresponding to a cleaning area, the station with the highest coverage of the cleaning point can be screened from at least one station with a coverage of the cleaning point greater than a preset coverage threshold, as the best station, i.e. the target station.

[0075] Of course, in actual implementation, the target station corresponding to each cleaning area can also be determined according to the coverage of the cleaning point according to other strategies, which is not limited in the present application.

[0076] Step 104, according to the target station, controlling the cleaning of the workpiece to be cleaned by the mechanical arm.

[0077] In the present application, after determining the target station of each cleaning area, the mechanical arm can be controlled to reach each target station to clean the cleaning points in the corresponding cleaning area. For example, the cleaning area includes cleaning areas one to three, cleaning area one corresponds to target station one, cleaning area two corresponds to target station two, and cleaning area three corresponds to target station three. The mechanical arm can be controlled to reach target station one by moving the guide rail, and the cleaning points in cleaning area one can be cleaned according to the poses of the cleaning points in the process point set corresponding to cleaning area one. The mechanical arm can be controlled to reach target station two by moving the guide rail, and the cleaning points in cleaning area two can be cleaned according to the poses of the cleaning points in the process point set corresponding to cleaning area two. The mechanical arm can be controlled to reach target station three by moving the guide rail, and the cleaning points in cleaning area three can be cleaned according to the poses of the cleaning points in the process point set corresponding to cleaning area three.

[0078] In the present application, if there are multiple target stations, the order in which the mechanical arm reaches each target station can be set according to actual needs, which is not limited in the present application.

[0079] The method of the present application is implemented, first, at least one cleaning area determined in advance for the workpiece to be cleaned and the target pose of the mechanical arm at each cleaning point of the workpiece to be cleaned are obtained; then, a plurality of random stations are generated for each cleaning area, and for each station, the coverage of the corresponding cleaning point of the cleaning area to which the current station belongs is determined according to the pose of the mechanical arm at the current station and the target pose of the mechanical arm at each cleaning point in the cleaning area to which the current station belongs; then, according to the coverage of the cleaning point, the target station corresponding to each cleaning area is selected from the plurality of stations; finally, according to the target station, the mechanical arm is controlled to clean the workpiece to be cleaned. In the present application, the fixed station for cleaning the workpiece is no longer manually specified, which can significantly improve the efficiency of the cleaning operation. In addition, the same mechanical arm can move on the guide rail according to the plurality of target stations planned in advance and implement the cleaning operation, without the need to carry multiple lasers, i.e. without the need to configure more cleaning equipment, which can significantly reduce the cost of the cleaning operation. Therefore, the method of the present application can effectively solve the problems of low cleaning efficiency or high cleaning cost in the prior art.

[0080] In combination with the above embodiments, in an implementation, step 104 can include:

[0081] Step 1041, generating a target station sequence according to each target station based on the travel constraint information of the guide rail and the distribution constraint information of the station.

[0082] In the present application, after determining the optimal station of all cleaning areas, the control device will comprehensively consider the travel constraint information of the guide rail (such as the limited length of the guide rail) and the distribution constraint information of the station (such as the constraint information set to realize the continuity between stations), to generate a complete and optimal guide rail station sequence, i.e. the target station sequence.

[0083] Step 1042, determining the cleaning action sequence of the mechanical arm at each target station.

[0084] In the present application, the control device will also generate a target sequence of the mechanical arm corresponding to each target station, i.e. the sequence of specific cleaning actions that the mechanical arm needs to perform at the target station, i.e. the cleaning action sequence.

[0085] Step 1043, controlling the mechanical arm to move on the guide rail according to the target station sequence, and controlling the mechanical arm to clean the workpiece to be cleaned according to the cleaning action sequence.

[0086] In the present application, each selected target station has undergone strict reachability verification to ensure that it can completely cover all cleaning points in the corresponding cleaning area.

[0087] In the present application, once the target station sequence is generated and verified, the control device will control the robot arm to move to each target station in sequence according to the planned target station sequence. At each target station, the robot arm will accurately complete the cleaning work of the corresponding cleaning area according to the preset laser cleaning program. The whole process is fully automated and does not require manual intervention.

[0088] In the present application, the fixed station for cleaning workpieces is no longer manually assigned, which can significantly improve the efficiency of cleaning work. In addition, the same robot arm can move on the guide rail according to a plurality of target stations planned in advance and implement cleaning work, without the need to configure more cleaning equipment, which can significantly reduce the cost of cleaning work.

[0089] In combination with the above embodiments, in an implementation, step 102 can include:

[0090] Step 1021, for each station, according to the pose of the robot arm at the current station and the target pose of the robot arm at each cleaning point in the cleaning area to which the current station belongs, determine the cleaning points in the cleaning area to which the current station belongs that can be reached by the end effector of the robot arm.

[0091] Step 1022, according to the ratio of the reachable cleaning points to the total number of cleaning points in the cleaning area to which the current station belongs, determine the coverage rate of the cleaning points in the cleaning area to which the current station belongs.

[0092] In actual implementation, if all cleaning points in the cleaning area to which the current station belongs include a target cleaning point, and the target cleaning point is any one of all cleaning points in the cleaning area to which the previous station belongs, then step 1021 can include:

[0093] determine the relative pose of the target pose of the robot arm at the target cleaning point in the cleaning area to which the current station belongs and the pose of the robot arm at the current station;

[0094] According to the relative pose, solve the joint angles of the robot arm through the inverse kinematics model of the robot arm;

[0095] If at least one set of joint angles is determined to be solved, it is determined that the end effector of the robot arm can reach the target cleaning point in the cleaning area to which the current station belongs.

[0096] In the present application, for each dynamically divided cleaning area, the present application uses the Monte Carlo algorithm to perform detailed reachability analysis on the station to determine the best station. The specific steps are as follows:

[0097] Step 1: Random station generation. A series of potential stations are randomly generated in each cleaning area.

[0098] Second step: coordinate system transformation and reachability evaluation. For each generated potential station, corresponding coordinate system transformation is performed to map the mechanical arm workspace under the station to the workpiece coordinate system.

[0099] Third step: using the inverse kinematics model of the mechanical arm, the end effector of the mechanical arm under the station is calculated one by one to determine whether it can reach all cleaning points in the cleaning area, that is, to determine whether there is at least one reachable joint configuration for each cleaning point. If it is determined that there is at least one reachable joint configuration, that is, at least one set of joint angles can be solved, it is determined that the end effector of the mechanical arm can reach the cleaning point in the cleaning area to which the current station belongs.

[0100] In the method, the reachability of the cleaning point is taken as the core index of the station optimization and is evaluated in a quantitative way. The pros and cons of a potential station are evaluated by generating a reachability score through the key index of the cleaning point coverage rate. The cleaning point coverage rate of a certain station represents the number of cleaning points in the cleaning area to which the station belongs that can be successfully reached by the mechanical arm, and the proportion of the total number of cleaning points in the cleaning area to which the station belongs. The higher, the stronger the cleaning point coverage ability of the station, and the higher the reachability score. F is as follows:

[0101]

[0102] wherein, represents the number of cleaning points in a single cleaning area, represents the number of cleaning points in the current cleaning area that can be successfully reached by the mechanical arm.

[0103] In solving the optimal station, by repeating the above process of randomly generating a station, evaluating the reachability and calculating the score, the Monte Carlo algorithm iteratively finds the station with a reachability score greater than the preset score and the highest score, which is taken as the best station.

[0104] In actual implementation, the target station can be directly screened according to the height of the cleaning point coverage rate, or the reachability score can be obtained according to the cleaning point coverage rate, and the target station can be screened according to the height of the reachability score. The height of the cleaning point coverage rate is proportional to the height of the reachability score.

[0105] In the present application, by analyzing the reachability of the cleaning point for each randomly generated station, the best station suitable for each cleaning area can be screened, and the mechanical arm is controlled to move on the guide rail and perform cleaning operation according to the best station, which can significantly improve the efficiency of the cleaning operation.

[0106] In combination with the above embodiments, in one implementation, when performing the reachability analysis of the cleaning points for each randomly generated station, the collision risk also needs to be considered. Specifically, for each station, according to the pose of the robot arm at the current station and the target poses of the robot arm at each cleaning point in the cleaning area to which the current station belongs, the cleaning points that can be reached by the end effector of the robot arm in the cleaning area to which the current station belongs are determined, including:

[0107] obtaining an initial pose of the robot arm;

[0108] determining an interpolation trajectory of the robot arm from the initial pose to each target pose through each station, and eliminating the stations with collision risks on the interpolation trajectory among the plurality of stations to obtain remaining stations;

[0109] for each station in the remaining stations, according to the pose of the robot arm at the current station and the target poses of the robot arm at each cleaning point in the cleaning area to which the current station belongs, the cleaning points that can be reached by the end effector of the robot arm in the cleaning area to which the current station belongs are determined.

[0110] wherein the initial pose of the robot arm refers to the pose of the robot arm before it reaches any target station, i.e., the pose of the robot arm when it is waiting on the guide rail.

[0111] For example, before performing the reachability analysis of the cleaning points for a certain station P, it is necessary to determine whether there is a collision risk on the interpolation trajectory of the robot arm from the initial pose to the target pose of each cleaning point in the cleaning area to which the station P belongs through the station P. If there is a collision risk on the interpolation trajectory of the robot arm from the initial pose to the target pose of a certain cleaning point M in the cleaning area to which the station P belongs through the station P, the station P is directly excluded, and other stations are evaluated to finally obtain the remaining stations without collision risks.

[0112] Then, the reachability analysis of the cleaning points is performed again for each station in the remaining stations, as described above.

[0113] The interpolation trajectory refers to a continuous, smooth and time-parameterized motion path that satisfies certain kinematic constraints, which is generated by a specific mathematical interpolation algorithm (such as spherical linear interpolation and polynomial interpolation) according to a series of given discrete path points (usually a sequence of poses). The interpolation trajectory precisely defines the pose, velocity and acceleration state of the robot end effector or each joint thereof at each instantaneous moment during task execution from the starting point to the target point.

[0114] Therefore, in the present application, when evaluating each station, collision detection is performed in real time, that is, the interpolation trajectory of the robot arm from the initial pose to the target pose of each cleaning point is detected, and the station with a collision risk in the interpolation trajectory is automatically excluded, which can further ensure the safety and feasibility of the selected station.

[0115] In combination with the above embodiments, in an implementation, before performing step 101, the control device of the present application also needs to perform the following work:

[0116] Obtain a Unified Robot Description Format (URDF) file of the robot arm;

[0117] According to the Unified Robot Description Format file, obtain the geometric parameters and kinematic parameters of the robot arm;

[0118] According to the geometric parameters and kinematic parameters, construct an inverse kinematics model of the robot arm.

[0119] In the URDF file, all physical properties of the robot arm are described, such as the length, weight, motion limit, coordinate system relationship, etc. of each joint.

[0120] In the present application, the control device first obtains the URDF file of the robot arm, then parses the file, extracts the joint information, link length, joint limit, etc. of the robot arm, and then automatically constructs the forward kinematics model and inverse kinematics model of the robot arm based on these parameters. Among them, the forward kinematics model is used to calculate the pose of the end effector under the given joint angle, and the inverse kinematics model is used to calculate the joint angle required for the end effector to reach the specified pose.

[0121] In the present application, when obtaining the interpolation trajectory of the robot arm from the initial pose to the target pose of a certain cleaning point, the forward kinematics model can be used.

[0122] In addition, the control device can also use a special program library KDL (Kinematics and Dynamics Library) to parse the robot arm URDF file and establish the forward and inverse kinematics model, to ensure the accuracy and universality of the model, and the coordinate system transformation relationship of the whole system is as shown in Figure 3 . Figure 3 is a coordinate system change diagram according to an embodiment of the present application.

[0123] In Figure 3 , a plurality of coordinate systems are mentioned, and each coordinate system will be briefly introduced below.

[0124] World coordinate system: an absolute, fixed global reference frame. Its origin and coordinate axis directions are unique and always constant throughout the entire work station.

[0125] Rail coordinate system: a coordinate system attached to the rail. The rail coordinate system itself does not move. When the base of the robot arm moves along the rail, the robot arm coordinate system moves with the rail. One of its axes usually aligns with the direction of movement of the rail, as shown in Figure 2 . This coordinate system is used to describe the position of the robot arm on the rail.

[0126] Robot arm base coordinate system: this coordinate system is fixed to the base of the robot arm and moves with the rail. For the robot arm itself, this is its own origin. The robot arm base coordinate system is the reference for all the robot arm's own motion calculations. The end tool position calculated by the forward and inverse kinematics model of the robot arm is relative to this base coordinate system.

[0127] Robot arm tool coordinate system: this coordinate system is fixed to the end of the robot arm wrist, i.e. where the laser cleaning head is installed. When the joints of the robot arm move, the position and orientation of this coordinate system change accordingly, so it is also called the end effector coordinate system (or robot arm end coordinate system). The robot arm tool coordinate system directly defines the working point and orientation of the tool. For example, for laser cleaning, it is necessary to accurately control the distance between the origin of this coordinate system (laser exit point) and the workpiece surface, and the Z-axis (laser exit direction) perpendicular to the workpiece surface. Its pose is determined relative to the robot arm base coordinate system.

[0128] Camera coordinate system: a coordinate system attached to the camera lens. If a vision camera is installed, this coordinate system can be used to describe the world as seen by the camera.

[0129] Robot arm joint coordinate system, used to describe the final solution. It does not represent a spatial coordinate, but a set of angle values. Through inverse kinematics operation, according to the set end target, the specific angle of each joint is calculated. This set of angle values will eventually be sent as instructions to the servo motors of the robot arm to drive the robot arm to complete the action.

[0130] Wherein, the conversion process of the cleaning point in each coordinate system can refer to the prior art, which is not limited in the present application.

[0131] In the present application, the forward kinematics model and the inverse kinematics model of the robot arm are established according to the URDF file of the robot arm, which provides a technical basis for subsequent determination of interpolation trajectories and implementation of collision detection of the robot arm.

[0132] In combination with the above embodiments, in an implementation, among the plurality of stations, the stations with collision risks on the interpolation trajectory are removed to obtain remaining stations, which can include:

[0133] An oriented bounding box (OBB) hierarchy tree corresponding to the robot arm is acquired, and the OBB hierarchy tree stores a plurality of directional OBB models, which are generated in advance for each joint, link, end effector of the robot arm and fixed obstacles in a working area where the robot arm is located;

[0134] Based on the OBB hierarchy tree, a flexible collision detection library is used to determine the stations with collision risks on the interpolation trajectory.

[0135] Among the plurality of stations, the stations with collision risks are removed to obtain remaining stations.

[0136] In the present application, the OBB technology can be used for geometric fitting in advance for each joint, link, end effector of the robot arm and obstacles that can exist in the working environment. The OBB is a compact and accurate bounding box representation, which can effectively describe the shape of a complex three-dimensional object.

[0137] In the present application, the control device generates OBB models for each link, joint, end effector of the robot arm and fixed obstacles in the working area, and these OBB models will be stored in the collision detection tree structure, i.e., the OBB hierarchy tree.

[0138] When performing collision detection, the flexible collision detection library (FCL) is used to achieve efficient real-time collision detection.

[0139] The real-time collision detection process based on the OBB model can include:

[0140] Input joint state: the algorithm proposes a set of possible joint angles.

[0141] Update OBB pose: according to the set of joint angles and the forward kinematics model, the latest pose of the OBB box on each link of the robot arm is calculated. The OBB of the environment remains unchanged.

[0142] Perform collision detection: query the FCL engine: “Please check whether there is any overlap between the set of OBBs representing the robot arm and the set of OBBs representing the environment in the current state?”

[0143] Return result: FCL uses its tree structure to make a judgment and returns a simple Boolean value: true (collision occurs) or false (no collision).

[0144] Decision: the algorithm decides according to the returned result, if there is no collision, the joint state (the set of joint angles) is available; if collision occurs, the joint state is discarded immediately.

[0145] Figure 4 is a flowchart of a collision detection algorithm according to an embodiment of the present application.

[0146] As shown in Figure 4 , the control device, on one hand, acquires the STL files of the links of each joint of the robot arm, and accurately fits them into OBBs (Oriented Bounding Boxes), and on the other hand, generates the corresponding OBBs for the obstacles according to the known pose information of the environment obstacles. After the collision models of both sides are completed, starting from a known initial pose of the robot arm, the trajectory interpolation is performed for the motion task to be executed, so as to acquire a series of continuous poses in the motion process from the starting point to the ending point. Finally, the system performs continuous collision detection between the OBB model of the robot arm in motion and the static OBB model of the environment throughout the entire interpolation trajectory, so as to determine whether any instantaneous state in the motion process exists interference, thereby ensuring the absolute safety of the motion trajectory.

[0147] In the present application, the collision detection is realized based on the oriented bounding box technology, which helps to improve the quality of the target station positions screened, and further improve the efficiency of the robot cleaning operation. In addition, compared with the simple axis-aligned bounding box (AABB) technology, the use of the oriented bounding box technology to realize the collision detection can more accurately reflect the actual shape of the object, reduce false positives, and improve the accuracy and efficiency of the collision detection.

[0148] In combination with the above embodiments, in an implementation, the cleaning area includes adjacent first and second cleaning areas, and an overlap area is arranged between the first and second cleaning areas, and the overlap area includes at least one cleaning point in the first cleaning area and at least one cleaning point in the second cleaning area.

[0149] In the present application, the overlap area can be arranged between part of the adjacent cleaning areas according to actual needs. The following description of the arrangement of the overlap area between the first and second cleaning areas is only exemplary.

[0150] Correspondingly, for each station position, according to the pose of the robot arm at the current station position and the target poses of the robot arm at each cleaning point in the cleaning area to which the current station position belongs, the coverage rate of the corresponding cleaning point of the current station position in the cleaning area is determined, including:

[0151] For each of the plurality of stations generated for the first cleaning area, according to the pose of the robot arm at the current station and the target poses of the robot arm at the cleaning points in the first cleaning area and the overlapping area, determine the coverage of the corresponding cleaning point under the current station in the first cleaning area.

[0152] For each of the plurality of stations generated for the second cleaning area, according to the pose of the robot arm at the current station and the target poses of the robot arm at the cleaning points in the second cleaning area and the overlapping area, determine the coverage of the corresponding cleaning point under the current station in the second cleaning area.

[0153] In the present application, since the laser cleaning points are usually discrete, and the cleaning trajectory may span different cleaning areas. In order to ensure the continuity of the cleaning trajectory, when a cleaning point group spans two or more cleaning areas, the present application sets a repeated coverage area, i.e. the overlapping area. The area is the overlapping part of two adjacent cleaning areas, and its size can be set as k times of the interval between each cleaning point in the cleaning process point set, where k is in the range of (1.5-2.0). The length of the overlapping area is designed to ensure that the key cleaning points can be completely covered by at least two adjacent stations between adjacent cleaning areas, so as to seamlessly connect when the station switches, and ensure the integrity of the cleaning trajectory, without involving the modification and interpolation of the original cleaning process point set.

[0154] In the present application, if at least two cleaning points contained in a cleaning area A are all located in the overlapping area between the adjacent another cleaning area B, then the originally planned station corresponding to the cleaning area A will be excluded. During cleaning, the robot arm is directly controlled to the station corresponding to the cleaning area B through the moving guide rail, and cleaning is realized.

[0155] Figure 5 is a whole flowchart of a robot arm cleaning operation according to an embodiment of the present application.

[0156] As Figure 5As shown, the control device receives the URDF file input and establishes the forward and inverse kinematics model of the robot arm based on it, and then uses the OBB technique to accurately geometrically fit each joint, link, end effector and working environment of the robot arm to build a collision detection model. At the same time, the control device receives the cleaning process point set input and the user-defined region boundary input, performs cleaning region division according to the two pieces of information, and performs continuity guarantee processing on the cleaning points at the junction of the cleaning regions in the division result. After completing the division of the cleaning region and the boundary point processing, the optimal station calculation link is entered, and a best station is solved for each cleaning region. This calculation process is closely coupled with a core feedback loop: the control device will make a real-time judgment on whether the robot arm collides with the environment, and if the judgment result is "yes", it means that the currently calculated station has a collision risk, and the process will return to the optimal station calculation step to re-solve or select other safe stations. Only when a station passes the collision detection, that is, the judgment result is "no", the station is confirmed as valid. After all the cleaning regions have found verified collision-free optimal stations, the control device finally generates a complete target station sequence and a cleaning action sequence.

[0157] The present application realizes the full-process automation of station planning in the guide rail type robot arm laser cleaning operation, effectively solves the problems of relying on manual experience, low planning efficiency, easy collision and unstable operation quality in the prior art. Through dynamic partitioning, overlapping region processing and deep fusion of real-time collision detection Monte Carlo optimization algorithm, the present application can quickly and accurately generate an optimal station sequence that is continuous in path and safe without collision.

[0158] In summary, the method proposed by the present application has the following technical effects:

[0159] First, high automation and flexible adaptability: only a small number of key parameters need to be input, and the entire station planning process can be automatically completed without manual intervention. The dynamic region division mechanism can flexibly adapt to various types of robot arms and workpieces of different sizes and quantities, effectively dealing with scenarios where small-sized parts do not require fixed station numbers.

[0160] Second, strong process adaptability and planning independence: by directly processing discrete cleaning points without modifying the original trajectory, the method of the present application exhibits strong adaptability, can meet the cleaning needs of workpieces of different sizes and quantities, and maintains the independence of station planning and cleaning point generation.

[0161] Third, high efficiency and high safety: special optimization is performed for the constraint conditions of the limited length guide rail, which can quickly generate a safe and reliable station sequence, and its calculation efficiency fully meets the requirements of online planning, ensuring collision-free operation during the entire cleaning operation process and significantly improving production safety and efficiency.

[0162] A rail type mechanical arm laser cleaning device provided by the present application is described below, and the rail type mechanical arm laser cleaning device described below can be correspondingly referred to the rail type mechanical arm laser cleaning method described above.

[0163] Figure 6 is a structural block diagram of a rail type mechanical arm laser cleaning device according to an embodiment of the present application. Referring to Figure 6 , a rail type mechanical arm laser cleaning device 600 according to the present application can include:

[0164] The acquisition module 601 is configured to acquire at least one cleaning area determined in advance for a workpiece to be cleaned, and a target pose of a mechanical arm at each cleaning point of the workpiece to be cleaned.

[0165] The determination module 602 is configured to generate a plurality of random stations for each cleaning area, and for each station, determine a coverage rate of a cleaning point corresponding to the station in a cleaning area to which the station belongs according to a pose of the mechanical arm at the current station and target poses of the mechanical arm at each cleaning point in the cleaning area to which the station belongs, the station being a position of the mechanical arm on a rail.

[0166] The screening module 603 is configured to screen a target station corresponding to each cleaning area from the plurality of stations according to the coverage rate of the cleaning point.

[0167] The control module 604 is configured to control the mechanical arm to clean the workpiece to be cleaned according to the target station.

[0168] According to the rail type mechanical arm laser cleaning device 600 of the present application, the determination module 602 is specifically configured to:

[0169] For each station, determine a cleaning point in a cleaning area to which the station belongs that can be reached by an end effector of the mechanical arm according to a pose of the mechanical arm at the current station and target poses of the mechanical arm at each cleaning point in the cleaning area to which the station belongs.

[0170] Determine a coverage rate of a cleaning point corresponding to the station in the cleaning area to which the station belongs according to a ratio of the reachable cleaning point to a total amount of cleaning points in the cleaning area to which the station belongs.

[0171] According to the rail type mechanical arm laser cleaning device 600 of the present application, the determination module 602 is specifically configured to:

[0172] Acquire an initial pose of the mechanical arm.

[0173] determine an interpolation trajectory of the robot arm from the initial pose to each target pose of each of the stations, and remove a station with a collision risk on the interpolation trajectory from the stations to obtain remaining stations;

[0174] For each of the remaining stations, determine a cleaning point in a cleaning area to which the current station belongs and which can be reached by an end effector of the robot arm according to a pose of the robot arm at the current station and target poses of the robot arm at each cleaning point in the cleaning area to which the current station belongs.

[0175] According to the rail type robot arm laser cleaning device 600, each cleaning point in the cleaning area to which the current station belongs includes a target cleaning point, and the determination module 602 is specifically configured to:

[0176] determine a target pose of the target cleaning point in the cleaning area to which the current station belongs and a relative pose of the pose of the robot arm at the current station;

[0177] According to the relative pose, solve joint angles of the robot arm through an inverse kinematics model of the robot arm;

[0178] If at least one set of joint angles is determined to be solved, it is determined that the end effector of the robot arm can reach the target cleaning point in the cleaning area to which the current station belongs.

[0179] According to the rail type robot arm laser cleaning device 600, the determination module 602 is specifically configured to:

[0180] obtain an oriented bounding box hierarchy tree corresponding to the robot arm, and the oriented bounding box hierarchy tree stores a plurality of oriented bounding box models, and the plurality of oriented bounding box models are generated in advance for each joint, connecting rod, end effector of the robot arm and fixed obstacles in a working area in which the robot arm is located;

[0181] Based on the oriented bounding box hierarchy tree, a flexible collision detection library is used to determine a station with a collision risk on the interpolation trajectory;

[0182] From the stations, remove the station with the collision risk to obtain the remaining stations.

[0183] According to the rail type robot arm laser cleaning device 600, the cleaning area includes adjacent first and second cleaning areas, and an overlap area is arranged between the first and second cleaning areas, and the overlap area includes at least one cleaning point in the first cleaning area and at least one cleaning point in the second cleaning area, and the determination module 602 is specifically configured to:

[0184] For each of the multiple stations generated for the first cleaning area, the coverage rate of the cleaning points corresponding to the current station in the first cleaning area is determined based on the pose of the robotic arm at the current station and the target pose of the robotic arm at each cleaning point in the first cleaning area and the overlapping area.

[0185] For each of the multiple stations generated for the second cleaning area, the coverage rate of the cleaning points corresponding to the current station in the second cleaning area is determined based on the pose of the robotic arm at the current station and the target pose of each cleaning point of the robotic arm in the second cleaning area and the overlapping area.

[0186] According to the guide rail type robotic arm laser cleaning device 600 of this application, the control module 604 is specifically used for:

[0187] Based on the travel constraint information of the guide rail and the distribution constraint information of the station positions, a target station position sequence is generated according to each target station position;

[0188] Determine the cleaning action sequence of the robotic arm at each of the target stations;

[0189] The robotic arm is controlled to move on the guide rail according to the target station sequence, and the robotic arm is controlled to clean the workpiece to be cleaned according to the cleaning action sequence.

[0190] Figure 7 This is a schematic diagram of the physical structure of an electronic device according to an embodiment of this application. Figure 7 As shown, the electronic device may include a processor 710, a communications interface 720, a memory 730, and a communication bus 740, wherein the processor 710, communications interface 720, and memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions in the memory 730 to execute a laser cleaning method for a rail-mounted robotic arm.

[0191] In addition, the logic instructions in the memory 730 described above can be implemented in the form of a software function unit and sold or used as an independent product, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that make contributions to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0192] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program is executed by a processor, so that the computer can execute a guide rail type mechanical arm laser cleaning method provided by the above-mentioned methods.

[0193] In another aspect, the present application also provides a non-transitory computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize a guide rail type mechanical arm laser cleaning method provided by the above-mentioned methods.

[0194] The device embodiments described above are only schematic, wherein the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment. Those skilled in the art can understand and implement without creative labor.

[0195] From the above description of the embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and necessary universal hardware platforms, and of course, can also be realized by hardware. Based on such understanding, the above technical solutions essentially or the parts that make contributions to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0196] It should be noted that the above examples are only used to illustrate the technical solutions of the present application, and are not intended to limit the same. Although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can be modified, or some of the technical features can be replaced by equivalent features. Such modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A rail-type robot laser cleaning method, characterized by, The method comprises the following steps: acquiring at least one cleaning area determined in advance for a workpiece to be cleaned, and a target pose of a mechanical arm at each cleaning point of the workpiece to be cleaned; generating a plurality of random stations for each cleaning area, and for each station, determining a coverage rate of a corresponding cleaning point under a cleaning area to which the station belongs according to a pose of the mechanical arm at the current station and target poses of the mechanical arm at each cleaning point in the cleaning area to which the station belongs, comprising: for each station, determining a cleaning point that can be reached by an end effector of the mechanical arm in the cleaning area to which the station belongs according to a pose of the mechanical arm at the current station and target poses of the mechanical arm at each cleaning point in the cleaning area to which the station belongs, comprising: acquiring an initial pose of the mechanical arm, determining an interpolation trajectory of the mechanical arm from the initial pose to each target pose through each station, and removing a station with a collision risk on the interpolation trajectory from the plurality of stations to obtain a remaining station, comprising: acquiring a bounding box hierarchy tree corresponding to the mechanical arm, the bounding box hierarchy tree storing a plurality of oriented bounding box models, the plurality of oriented bounding box models being generated in advance for each joint, connecting rod and end effector of the mechanical arm and a fixed obstacle in a working area in which the mechanical arm is located, determining a station with a collision risk on the interpolation trajectory based on the bounding box hierarchy tree through a flexible collision detection library, removing the station with the collision risk from the plurality of stations to obtain the remaining station; for each station in the remaining station, determining a cleaning point that can be reached by the end effector of the mechanical arm in the cleaning area to which the station belongs according to a pose of the mechanical arm at the current station and target poses of the mechanical arm at each cleaning point in the cleaning area to which the station belongs; determining a coverage rate of a corresponding cleaning point under a cleaning area to which the station belongs according to a ratio of the reachable cleaning point to a total amount of cleaning points in the cleaning area to which the station belongs, the station being a position of the mechanical arm on a guide rail; screening target stations corresponding to each cleaning area from the plurality of stations according to the coverage rates of the cleaning points; controlling the mechanical arm to clean the workpiece to be cleaned according to the target stations.

2. The rail-type robot laser cleaning method of claim 1, wherein, Each cleaning point in the cleaning area to which the current station belongs comprises a target cleaning point, and the determination of the cleaning point that can be reached by the end effector of the mechanical arm in the cleaning area to which the current station belongs according to the pose of the mechanical arm at the current station and the target poses of the mechanical arm at each cleaning point in the cleaning area to which the current station belongs comprises: determining a relative pose of the target pose of the target cleaning point in the cleaning area to which the current station belongs and the pose of the mechanical arm at the current station; solving joint angles of the mechanical arm according to the relative pose through an inverse kinematics model of the mechanical arm; If it is determined that at least a set of joint angles can be solved, it is determined that the end effector of the robot arm can reach the target cleaning point in the cleaning area to which the current station belongs.

3. The rail-type robot laser cleaning method of claim 2, wherein, The inverse kinematics model is determined by the following steps: Obtain a unified robot description format file of the robot arm; According to the unified robot description format file, obtain the geometric parameters and kinematics parameters of the robot arm; According to the geometric parameters and the kinematics parameters, construct the inverse kinematics model of the robot arm.

4. The rail-type robot laser cleaning method of claim 1, wherein, The cleaning area includes adjacent first and second cleaning areas, and an overlap area is provided between the first and second cleaning areas. The overlap area includes at least one cleaning point in the first cleaning area and at least one cleaning point in the second cleaning area. For each station, according to the pose of the robot arm at the current station and the target pose of each cleaning point in the cleaning area to which the current station belongs, the coverage rate of the corresponding cleaning point in the cleaning area to which the current station belongs is determined, which includes: For each station in the plurality of stations generated for the first cleaning area, according to the pose of the robot arm at the current station and the target pose of each cleaning point in the first cleaning area and the overlap area, the coverage rate of the corresponding cleaning point in the first cleaning area is determined. For each station in the plurality of stations generated for the second cleaning area, according to the pose of the robot arm at the current station and the target pose of each cleaning point in the second cleaning area and the overlap area, the coverage rate of the corresponding cleaning point in the second cleaning area is determined.

5. The rail-type robot laser cleaning method of claim 1, wherein, According to the target station, the robot arm cleans the workpiece to be cleaned, which includes: Based on the stroke constraint information of the guide rail and the distribution constraint information of the station, a target station sequence is generated according to the target station; Determine the cleaning action sequence of the robot arm at each target station; According to the target station sequence, control the robot arm to move on the guide rail, and according to the cleaning action sequence, control the robot arm to clean the workpiece to be cleaned.

6. The rail-type robot laser cleaning method of claim 1, wherein, The at least one cleaning area is determined by the following steps: Obtain the boundary constraint information of the cleaning area input by the user; According to the boundary constraint information and the target pose of each cleaning point of the workpiece to be cleaned, the at least one cleaning area is determined.

7. A rail-type robot laser cleaning apparatus for performing the rail-type robot laser cleaning method of any one of claims 1-6, characterized by, It includes: The acquisition module is used for acquiring at least one cleaning area determined in advance for the workpiece to be cleaned, and the target pose of each cleaning point of the workpiece to be cleaned; The determination module is used for generating a plurality of random stations for each cleaning area, and for each station, according to the pose of the robot arm at the current station and the target pose of each cleaning point in the cleaning area to which the current station belongs, the coverage rate of the corresponding cleaning point in the cleaning area to which the current station belongs is determined. The station is the position of the robot arm on the guide rail. The screening module is configured to screen target stations corresponding to the cleaning areas from the plurality of stations according to coverage of the cleaning points. The control module is configured to control the mechanical arm to clean the workpiece to be cleaned according to the target stations.

Images