• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The invention relates to a method for following a workpiece surface through a robot system, wherein the robot system includes an end effector, a positioning device, and at least one sensor configured so that the surface on which an action is to be exerted is first observed before being observed affected by the tool. The method includes locally detecting the surface to be followed by the at least one sensor data creation sensor, determining the position of the end effector for creating position data, combining the sensor data and the position data to form a local surface model creating a local part of the surface to be tracked, generating a robot trajectory for the end effector based on the local surface model, and driving the end effector according to the robot trajectory.
    公开号:BE1024133B1
    申请号:E2016/5281
    申请日:2016-04-22
    公开日:2017-11-20
    发明作者:Schutter Joris De；Erwin Aertbelien
    申请人:Flexible Robotic Solutions Bvba；
    IPC主号:
    专利说明:

    Method and device for following and treating a surface by a robot
    Technical field
    The present invention relates to a method for following and treating a surface, and in particular for following and treating a surface by a robot.
    The present invention further relates to a device for monitoring and treating a surface.
    State of the art
    The tracking or treatment of a surface by a robot when no precise geometric model of the workpiece is available or when the geometry of the workpiece changes with each individual workpiece is known. Examples of surface treatment include tasks that involve contact between the robot end effector and the workpiece, such as polishing, grinding, deburring, cleaning, removal of a coating, gluing, as well as tasks that do not involve contact, such as welding and spray painting. In the absence of any surface treatment, the robot task can be reduced to merely following a surface, for example during inspection and scanning or measuring tasks. Variation of the geometry of the workpiece occurs particularly during the treatment of natural products.
    Known solutions for following a surface by a robot rely on on-line generation and / or adaptation of the robot trajectory using direct feedback of sensor signals (vision, distance, force), or off-line trajectory using a geometrically obtained in advance fashion model.
    In the first known method, no or only an approximate geometric model of the workpiece is available in advance, while a sensor is mounted on the robot end effector. While tracking the surface, the sensor, for example a vision, distance or force sensor, provides information about the geometry of the workpiece and this information is fed back to a robot controller. The sensor measurements are used directly to generate the robot trajectory or to adjust a predefined desired robot trajectory. An intrinsic disadvantage of the direct feedback method is that the control signal generated by the robot controller is based on the existence of a tracking error and therefore always comes too late, because there is no control signal without the presence of a tracking error. The tracking error in this case corresponds to a geometric error between the desired position and orientation of the robot tool relative to the surface of the workpiece, and the actual position and orientation of the robot tool. In other words, the performance of the robot control is intrinsically limited and results in a limited accuracy of the surface tracking task. Furthermore, the tracking error will be greater, and consequently the accuracy will be lower if the feedback loop has a lower bandwidth (i.e., a lower dynamic response), or when the surface tracking task is performed at a higher speed. Therefore, this solution is typically used in applications that involve a relatively low speed between the end effector and the workpiece.
    In the second method, the off-line trajectory generation is based on a geometric model that is obtained in advance, either in the same arrangement as the surface tracking or surface treatment task, or in a separate arrangement. If the model is obtained in the same arrangement, this can be done on the basis of local observations using a sensor that is brought close to the surface of the work piece by attaching it to the robot end effector. A disadvantage of this method is that the resulting solution is not time efficient, because the robot must first perform one or more scanning movements before it can proceed to calculate and perform the movements for following or treating the surface. An alternative is to obtain the model in the same setup with the help of global observations, for example with the aid of a 3D vision camera, or an RGB-D camera, set up at a fixed location in the robot environment. However, although global observation is much faster, this alternative requires great absolute accuracy in a large operating range, both from the sensor system and from the robot. The same requirement occurs when the model is obtained in a separate arrangement; a high absolute accuracy is then required, both for the measurement setup and for the robot setup. In addition, both operating areas must then be well calibrated with respect to each other.
    Known surface treatment approaches involving material removal either combine trajectory generation and position control in some directions with force control in other directions, or use trajectory generation and position control in all directions.
    In the method based on position control in some directions and force control in other directions, force control can be applied passively, i.e. by means of a passive impedance such as a spring, or in an active manner, i.e. by force control using a force sensor and force feedback. Passive force control is simple and easy, but not very accurate. On the other hand, active force control allows a more accurate check, but is more complex and more expensive. A known embodiment of force control is pressing the cutting tool against the surface with a known and controlled contact force. In such an embodiment, the speed at which material is removed depends on the tangential speed of the tool, the perpendicular contact force and the properties of the material of the workpiece. In this case, a model for the speed of material removal is required. An additional problem of force feedback in the absence of an accurate geometric model of the workpiece is that the force feedback requires a tracking error in terms of contact force to generate an adjustment of the robot trajectory relative to the workpiece, and therefore also an error on the actual material removal rate . A possible and known remedy for this is to press against the surface of a tool that is shaped in such a way that it removes a layer of material of constant thickness, for example a peeling tool. The limitation of such an approach is that the thickness of the layer that is removed is determined by the shape of the tool and cannot be adjusted on-line by the robot controller.
    In the method based on trajectory generation and position control in all directions, so also in the direction perpendicular to the surface, the surface treatment is based on position control in all directions. This method has the advantage that the thickness of the layer being removed can be adjusted on-line by the robot. However, the disadvantage of this method is that a very accurate model of the workpiece is required.
    Description of the invention
    It is an object of the present invention to provide a method for following a surface of a workpiece with a robot system in which one or more disadvantages are solved in the prior art.
    This object is achieved according to the invention with a method for following a workpiece surface with a robot system, wherein the robot system comprises a tool for exerting an action on the followed surface, a positioning device comprising an end effector for relative positioning of the tool relative to the surface, and at least one sensor configured so that the surface on which the action is to be exerted is first sensed by the at least one sensor before being affected by the robotic tool, and wherein the method includes detecting locally the surface to be followed with the at least one sensor for creating sensor data; determining the position of the end effector for creating position data; combining the sensor data and the position data to create a local surface model of a local part of the surface to be tracked; generating a desired robot trajectory for the end effector based on the local surface model; and controlling the end effector according to the desired robot trajectory.
    This method has the advantage that local sensor data is used to create a local surface model, which allows very fast and accurate tracking of a surface. The method combines observation and tracking of a surface in one set-up and in a single movement, resulting in a time, space and cost-efficient method. The method includes local observation by means of a sensor that is brought close to the surface of the workpiece to make its measurements of the surface. Consequently, the proposed method does not require high accuracy of the sensor and of the robot in a large operating range (the global dimensions of the robot), but rather relies on the accuracy of the robot in a limited operating range (local dimensions around the end of the robot tool) in the environment of its current position and on the accuracy of the geometric calibration between the sensor and the robot tool. The method further uses a local geometric model of the surface of the workpiece created from sensor data prior to actual surface tracking, i.e. the sensors are positioned in front of the robot tool. Furthermore, the control of the robot is not only based on direct feedback from the sensor signals, but involves accurate, model-based trajectory generation and control with trajectory feedback in advance. This avoids a geometric tracking error and allows a fast and accurate tracking of a surface.
    In some embodiments of the invention, the robot tool is a surface treatment tool and the method further includes performing surface treatment during relative movement of the robot tool relative to the surface to be tracked on which the action is to be applied.
    In such embodiments, the method has the additional advantage that control of the position of the end effector is possible in all directions. This is advantageous to check the thickness of the layer to be removed by the surface treatment. This method also makes it possible to use position control in some directions in combination with force control in some other directions.
    In some embodiments of the invention, the robot tool is a planer tool or cutter for removing the coating from a workpiece such as cheese, where the method includes the step of removing the coating from the workpiece such as cheese during relative movement of the robot tool relative to of the surface to be followed.
    In some embodiments of the invention, the robot system further comprises at least one second sensor that is configured such that the surface on which the action is to be applied is first touched by the cutting tool before it is detected by the at least one second sensor, the method further local detection includes the surface on which the action is applied with the at least one second sensor to create second sensor data, the action performed being based on the second sensor data to create data, and adjusting one or more parameters of the robot trajectory to basis of the data.
    The method of these embodiments has the advantage that the result or the quality of the surface treatment can be monitored. The second sensor data is then used to adjust parameters such as cutting depth, tangential speed, normal force, etc. to optimize the performance of the method.
    In some embodiments of the invention, the second sensor includes at least one color detection sensor and the second sensor data contains color data from the processed local surface. These embodiments are particularly advantageous for workpieces where a surface with a different color must be removed, such as the cheese coating. A very accurate result can be obtained by applying color detection.
    In some embodiments of the invention, the second sensor includes at least one sensor based on laser intensity measurement and the two sensor data contains data from laser intensity measurements of the treated local surface.
    In some embodiments of the invention, the step of combining the sensor data and the position data includes generating a cloud of 3D points of the surface that are representative of the local surface model.
    In some embodiments of the invention, the step of combining the sensor data and the position data further comprises generating an analytical model at the position of the end of the robot tool, or in the vicinity thereof, from the cloud of 3D points from the surface.
    In some embodiments of the invention, the step of generating an analytical model includes determining the parameters of the analytical model by applying a stochastic parameter estimation algorithm to the cloud of 3D points of the surface.
    In some embodiments of the invention, the step of determining the parameters of the analytical model includes estimating the parameters, taking into account initial values for the parameters that have been stored, were learned during previous embodiments of the surface following of the same or another work piece or otherwise available.
    The two latter embodiments have the advantage that the generation of the robot trajectory can already start during a start phase before the robot end effector or the robot tool reaches the surface to be followed or treated, or even before the first sensor starts to detect.
    In some embodiments of the invention, the step for controlling the end effector according to the desired robot path includes outputting successive control signals, and the control signals include at least one feedback-based component.
    In some embodiments of the invention, the at least one component includes control based on feedback position data or a component thereof.
    This is advantageous to compare the position data (or a component thereof) with the desired value.
    In some embodiments of the invention, the at least one control-based component includes data corresponding to contact force or moment of contact between the robot tool and the workpiece surface.
    In such embodiments, the method offers the advantage that it allows to control the position of the end effector in some directions based on sensor data and in other directions based on contact force or contact moment data.
    It is another object of the present invention to provide a device for quickly and accurately following a surface without the disadvantages of the prior art.
    This object is achieved according to the invention with a device for following a surface comprising an end effector configured to receive a robotic tool, a positioning device that holds the end effector for positioning the end effector relative to the surface and configured for making position data available, first sensor means provided on the end effector and configured so that, in use, the surface to be tracked is first sensed by the first sensor means to create sensor data before being affected by the robotic tool, and a control configured to a local create a surface model of a local part of the surface to be tracked based on the sensor data and on the received position data, and to generate a desired robot trajectory based on the local surface model.
    The advantages of this device are the same as the advantages of the embodiments of the method according to the invention described above.
    Brief description of the drawings
    The invention will be further elucidated with reference to the following description and the accompanying figures.
    Figure 1 illustrates a hardware arrangement and algorithmic flow chart according to an embodiment of the present invention.
    Figure 2 shows an image of a robot system at a certain point in time while following the surface of a workpiece.
    Embodiments of the invention
    The present invention will be described in the light of specific embodiments and with reference to certain drawings, but the invention is not limited thereto but only by the claims. The described drawings are only schematic and not restrictive. In the drawings, the size of some elements may be exaggerated and not drawn to scale for illustrative purposes. The dimensions and relative dimensions do not necessarily correspond to actual conversion to the practice of the invention.
    Furthermore, the terms first, second, third and the like are used in the description and in the claims to distinguish between similar elements and not necessarily for describing a sequence or chronological sequence. The terms are interchangeable under appropriate conditions and the embodiments of the invention may function in sequences other than described or illustrated herein.
    In addition, the terms top, bottom, top, bottom and the like in the description and claims are used for descriptive purposes and not necessarily for describing relative positions. The terms thus used are interchangeable under appropriate conditions and the embodiments of the invention described herein may function in orientations other than those described or illustrated herein.
    Furthermore, although the various embodiments are referred to as "preferred," the various embodiments should be construed as examples of ways in which the invention may be practiced rather than as limitations on the scope of the invention.
    The term "contains" used in the claims should not be interpreted as being limited to the elements or steps that are listed thereafter; it does not exclude other elements or steps. It must be interpreted as precisely stating the presence of the indicated characteristics, wholes, steps or components as referred to, but does not exclude the presence or addition of one or more other characteristics, wholes, steps or components, or groups thereof . Thus, the scope of the expression "a device containing A and B" should not be limited to devices that consist solely of parts A and B, but rather with regard to the present embodiment, the only listed parts of the device are A and B , and further, the conclusion must be interpreted as including equivalents of these parts.
    Figure 1 shows an embodiment of the present invention. The hardware arrangement in Figure 1 contains a positioning device with multiple degrees of freedom 1, which in the embodiment of Figure 1 is a robot arm, but which, depending on the application, may alternatively be a positioning device with fewer degrees of freedom, such as a 2D or 3D Cartesian positioning device . An end effector 2 is attached to or integrated with the positioning device 1. A control system 6, such as an industrial robot control or another multi-axis position control, controls the positioning device 1. A first sensor 3 is provided on the end effector 2, either attached to it or integrated part. In the embodiment of Figure 1, a second sensor 4 is also provided on the end effector 2, again either attached to it or as an integrated part. This second sensor 4 is optional and cannot be present in an alternative embodiment. The end effector 2 is configured to be connected to a tool 5. In an alternative embodiment, the tool 5 may be part of end effector 2. The tool 5 performs operations in contact or not in contact on workpiece 7 that has a product, a mechanical structure , or any other object with a surface to be followed or treated.
    The hardware arrangement of Figure 1 is controlled to follow a surface 9 of the workpiece 7 with an at least partially unknown or variable geometry or position, and to perform an operation or task on this surface 9. The operation or task may be a task without contact, such as inspection, geometric measurement, paint spraying, welding, or any other non-contact task known in the field. The operation or task may also be a task involving mechanical contact, such as deburring, polishing, grinding, removal of a coating, gluing, or any contact operation known in the art. To perform the operation or task, a suitable tool 5 is mounted on the end effector 2. The tool 5 may be a measuring system, cutting tool, welding tool, liquid distributor or any other tool known in the art. The hardware arrangement of Figure 1 can also be used to follow the surface 9 only, without performing an operation or task on the surface. In such use, the tool is not required and can be omitted.
    The first sensor 3 provided on the end effector 2 is used to measure the geometry of workpiece 7 for the purpose of controlling the robot during its surface tracking task. The first sensor 3 may include one or more sensors, such as, but not limited to, laser point sensors (1D laser displacement sensors), laser line sensors (2D laser displacement sensors), 3D laser scanners, 2D or 3D cameras, or any combination thereof. For example, in one embodiment, the first sensor 3 may be a combination of three laser point sensors that face the surface. The first sensor 3 is placed on the end effector 2 such that, in use, the first sensor 3 comes before the tool 5 with respect to the movement of the end effector 2 relative to workpiece 7. This is illustrated in Figure 1 in which arrow 10 indicates the direction of movement. Due to this placement of the first sensor 3 on the end effector 2, the first sensor 3 can take a close-up measurement on a limited area of the surface 9. This is called local observation. Thus, local observation is performing a close-up measurement on a limited area of the surface of the workpiece by means of a sensor that is brought close to the workpiece surface.
    The second sensor 4 is optional in the embodiment of Figure 1 as indicated above. The second sensor 4 can be added to the end effector 2 so that, during use, the second sensor 4 comes behind the tool with respect to the relative movement between the end effector 2 and the workpiece 7. This is illustrated in Figure 1. The second sensor 4 is used to monitor aspects related to the quality of the operation or task performed by tools 5 on the workpiece surface 9.
    The robot controller 6 comprises at least a low-level controller. The low level control controls the execution of the desired trajectory generated or adjusted based on the robot control input 11. The robot control input 11 is the result of a set of sensor processing and control algorithms represented by the algorithmic flow diagram in Figure 1. The robot control 6 is configured to control the robot control input 11 receive. The robot controller 6 sends actuation signals 12 to the actuators of the positioning device 1, receives measurement 27 of the positions of each of the actuators and / or robot joints, and, based on this measurement, sends information 26 about the end effector position and orientation in one or another representation (in terms of positions of the robot joints, or any other geometric representation, such as Cartesian position and orientation or pose) to the real-time sensor processing and control algorithms with a given sampling frequency. Additionally, the robot controller 6 may include an internal trajectory generation module. In such an embodiment, the robot controller 6 may consider the robot control input 11 as an adaptation signal to be combined with the trajectory generated by the internal trajectory generation module, resulting in an adjusted desired robot trajectory.
    In an alternative embodiment, the configuration of the arrangement is changed so that the first sensor 3, second sensor 4, and tool 5 are fixedly attached to the environment, while workpiece 7 is attached to the end effector, for example, using a gripper configured to close or opening by means of a controlled actuation signal to grasp or release the workpiece. For this purpose, the first sensor 3, second sensor 4 and tool 5 can be suitably brought together in a unit with fixed relative position between these parts 3, 4 and 5. In this embodiment the relative movement between parts 3, 4 and 5 is on the one hand, and workpiece 7 on the other, similar to the first embodiment. In yet other embodiments, sensor 3, sensor 4, and tool 5 can be manipulated by different positioning devices that move in an appropriately coordinated manner, each of these positioning devices making its measured position available.
    As illustrated by the algorithmic flow chart in Figure 1, the sensor processing and control algorithms can be conveniently brought together in three consecutive software modules: i) a geometric surface module 15 for creating a geometric model of the surface 21, ii) a robot trajectory module for generating or adjusting a robot trajectory 22 based on the created geometric model of the surface 21, and iii) a control module 17 for calculating the robot control input 11 based on the generated or modified desired robot trajectory 22. The sensor processing and control software runs with suitable sampling frequencies that may be the same or different for the different parts or modules.
    In the first software module 12, the local sensor information 25 of the first sensor 3 is combined with the end effector position and orientation 26 at corresponding times to calculate the position of measured points on the surface 9. As the end effector 2 moves relative to the workpiece surface 9, more and more points of the surface 9 are calculated, resulting in a cloud of 3D points on the surface. It is important, especially when the robot is moving fast, to be sure that the correspondence between the times of the measurement of the first sensor 3 on the one hand, and the measurement of the position and orientation of the end effector 26 on the other hand, is as accurate as possible. is. The correspondence between these two measurements can be improved by compensating in the software module for any time delay between the two measurements. This ensures that the geometric information of any sensor on the end effector 2 on the one hand, and any robot position sensor on the other hand, is combined (fused) at exactly the same time.
    The cloud of 3D points on the surface is a first possible representation of a geometric model of the surface. The cloud of 3D points on the surface could grow so that it covers an extended part of the workpiece and is then called a global model. However, only a subset of the cloud of 3D points on the surface near the end of the tool is considered for further processing. This subset of the cloud of 3D points on the surface near the end of the tool is called the local model. Other representations of the geometric model for the surface can be derived from the cloud of 3D points of the surface.
    A second type of representation of the surface model uses a set of local characteristics, for example: 3D position of a representative point, normal and tangential vectors (stamp / roll angle), information about curvature and torsion. The set of local characteristics is representative of the surface in the position of the end of the tool, or in the vicinity thereof. In an embodiment of the present invention, the local characteristics of the surface can be calculated directly from the cloud of 3D points of the surface.
    A third type of representation of the surface model contains the parameters of an analytical model, for example a local quadratic surface in the vicinity of the position of the end of the tool 5, whereby the local quadratic surface is compensated with the aid of the cloud of 3D points using a numerical procedure. This numerical procedure can involve a stochastic, for example Bayesian, approach to estimate the parameters of the surface model. In this way, a preliminary estimate of the model parameters is used to create a reliable model of the surface even if only a limited number of sensor measurements are available, for example at the start of the surface tracking task at, or even before, the edge of the workpiece. Furthermore, as an alternative embodiment, the local surface characteristics in the second type representation of the surface model can be calculated from the set of parameters of the analytical model instead of calculating them directly from the cloud of 3D points.
    In addition, in order to further improve the surface tracking accuracy, in particular at high relative speeds between robot and workpiece, the position of the end of the tool 5, in the vicinity of which the local model of the surface is created, does not correspond to the actual position of the end of the tool at the time the local model of the surface is created, but corresponds to the estimated future position of the end of the tool at the time the control input based on this model will have its effective effect. This can be accomplished by compensating for any check or input delay in the software module when the position of the end of the tool is to be selected.
    In the robot trajectory module 16, the geometric model of the surface 21 is used to either generate a desired robot trajectory or adjust a given robot trajectory, so that in both situations the output is a desired robot trajectory 22. The resulting robot trajectory 22 can be expressed in any suitable form. For example, the representation of the robot trajectory 22 may take the form of desired values for the instantaneous 3D positions and 3D orientations of a reference axis system connected to the end of the tool, and / or desired values for the instantaneous 3D translation speed and 3D rotation speed for the reference axis system connected to the end of the tool. The calculation of the desired robot trajectory can be based on any representation of the geometric model of the surface, for example the cloud of 3D points, the local characteristics of the surface or the parameters of an analytical model of the surface, as discussed earlier.
    Optionally, when the prior estimation of the parameters of the analytical model is used, the generation or modification of the robot trajectory 22 can already start during an approach phase before the end effector 2 or the tool 5 reaches the surface 9, and even before the first sensor 3 starts by measuring the first points on the surface. In this way the programming and control approach of the approach phase is facilitated, which leads to a reduced programming effort.
    As an additional option, the generation or modification of the robot trajectory 22 may take into account measurement 28 of sensor 4 that monitors quality aspects with which the machining or task is performed on the workpiece surface 9 by the tool 5, and which is fed back to the control module 17. For example, the cutting depth of the tool, the speed of the robot, or another controllable parameter in the process can be adjusted based on measurement 28 to improve the quality of the process.
    The control module 17 (the third software module in the embodiment of Figure 1) has a robot trajectory controller that receives the output of the robot trajectory module 16 as well as the robot position measurement 26. The output of the robot trajectory module, the desired robot trajectory 22, corresponds to a reference or pre-coupling signal, or is used to create such a signal, while the robot position measurement 26 is used to close a feedback loop around robot 1 and its robot controller 6, ie a loop formed by connecting components and signals 17-11-6-12-1 -27-6-26-17. The feedback loop compares the actual robot position given by or determined based on the robot position measurement 26, for example expressed in terms of the 3D position and 3D orientation of the reference axis connected to the end of the tool, with the corresponding desired reference values determined by the robot path 22, and generates a feedback signal which is added to the forward feedback signal, resulting in the total robot control input 11 being sent to the robot controller 16.
    Optionally, in an alternative embodiment, the control module 17 can compensate for any time delay in the feedback loop, for example by replacing the measured robot position with an estimate thereof at a future time.
    In an alternative embodiment, the feedback control of the measured robot position or the measured position of the end of the tool can be replaced at least in some translational or rotational directions by the feedback of the contact force or the moment of contact between the tool 5 and the workpiece surface 9. The contact forces or moments can be measured by a force sensor attached, in one embodiment, to end-effector 2 or, in another embodiment attached to the environment, for example to a support of the workpiece.
    Furthermore, in an alternative embodiment, any kinematic redundancy in the system, for example due to the rotational symmetry of the tool, can be exploited to keep the position and / or orientation of at least one of the sensors relative to the surface within a certain range. .
    Furthermore, the software module 17 can use a constraint-based approach, ie the robot task is at least partially specified using a set of constraints expressing desired geometric relationships between the end of the tool, the sensors and end-effector 2, on the one hand, and workpiece 17, its surface 19 or other objects in the robot environment, on the other. To improve the application of such a constraint-based approach, task programming can be performed in a software framework or software package specially designed for constraint-based task specification and control, such as, but not limited to, iTaSC [1], SoT [2] , or eTaSL [3],
    In an embodiment of the invention, the robot is configured to remove the coating from cheese. The workpiece 7 in that case is cheese, and the tool 5 is a tool that removes surface material such as a planer tool, a milling cutter (e.g. end milling cutter, ordinary milling cutter, etc.) or any other cutting tool suitable for removing a surface layer. The optional second sensor 4 could be a sensor to measure color or laser intensity. Based on the color measurement or laser intensity measurement, the controller can check whether the coating has been completely removed or not.
    In Figure 2, a robot system 100 is illustrated as it performs surface machining on workpiece 7. The first sensor 3 includes three laser point sensors, i.e., one-dimensional laser displacement sensors. The use of three laser point sensors is advantageous because on the one hand it provides accurate sensor data and on the other hand allows to limit the amount of data that needs to be processed. The laser point sensors are characterized by a fixed and calibrated position relative to end effector 2 and robot tools 5.
    Known surface tracking and surface machining solutions include several topologically different configurations to generate the relative movement of the tool relative to the workpiece. The workpiece can be fixed in the environment, on a conveyor belt, positioned by a positioning device or manipulated by a robot, while, similarly, the tool can be attached to the environment, attached to a positioning device or manipulated by a robot. It is obvious to those skilled in the art that the methods of the present invention described above apply to any of these topological configurations.
    权利要求:
    Claims (30)
    [1]
    Conclusions
    A method for tracking a workpiece surface by a robotic system, the robotic system comprising a tool configured to perform an action on the tracked surface, a positioning device including an end effector for relative positioning of the tool with respect to the surface, and at least one sensor configured so that the surface on which the action is to be applied is first observed by the at least one sensor before being affected by the tool, the method comprising detecting the surface to be followed locally with the at least one sensor for creating sensor data; determining the position of the end effector for creating position data; combining the sensor data and the position data to create a local surface model of a local part of the surface to be tracked; generating a desired robot trajectory for the end effector based on the local surface model; and driving the end effector according to the robot trajectory.
    [2]
    A method according to claim 1, wherein the tool is a surface inspection tool, and wherein the method further comprises performing surface inspection during relative movement of the tool relative to the surface on which the action is to be applied.
    [3]
    A method according to claim 1, wherein the tool is a surface treatment tool, and wherein the method further comprises performing surface treatment during relative movement of the tool relative to the surface on which the action is to be applied.
    [4]
    A method according to any of the preceding claims, wherein the at least one sensor is attached to the end effector.
    [5]
    A method according to any of claims 1 to 3, wherein the end effector contains the at least one sensor.
    [6]
    A method according to claim 3, wherein the tool is a planer tool or cutter for removing a coating from a workpiece such as cheese, wherein the method includes the step of removing the coating from a workpiece such as cheese during relative movement of the workpiece. tool relative to the surface on which the action is to be applied.
    [7]
    A method according to any of the preceding claims, wherein the robot system comprises at least one second sensor configured so that the surface on which the action is to be exerted is first affected by the tool before being detected by the at least one second sensor, the method further comprising locally sensing the surface on which the action is to be applied with the at least one second sensor to create second sensor data; verifying the action taken based on the second sensor data for creating verification data; and adjusting one or more parameters of the desired robot trajectory based on the verification data.
    [8]
    A method according to claim 7, wherein the second sensor comprises at least one color detection sensor and wherein the second sensor data contains color data from the local surface on which the action is applied.
    [9]
    A method according to any of claims 7 to 8, wherein the second sensor comprises at least one sensor based on laser intensity measurement and wherein the second sensor data contains laser intensity measurement data from the local surface on which the action is applied.
    [10]
    A method according to any one of the preceding claims, wherein the step of combining the sensor data and the position data comprises generating a cloud of 3D points on the surface representative of the local surface model.
    [11]
    A method according to any of the preceding claims, wherein the step of combining the sensor data and position data includes compensating any time delay between the sensor data and the position data.
    [12]
    A method according to any of claims 10 or 11 when dependent on 10, wherein the step of combining the sensor data and position data further comprises calculating a number of local surface characteristics at a position of the end of the tool or in the vicinity thereof, from the cloud of 3D points of the surface.
    [13]
    A method according to any of claims 10 or 11 when dependent on 10, wherein the step of combining the sensor data and position data further comprises generating an analytical model at a position of the end of the tool or in the vicinity thereof from the cloud of 3D points of the surface.
    [14]
    A method according to claim 13, wherein the step of generating an analytical model comprises determining the parameters of the analytical model by applying a stochastic parameter estimation algorithm to the cloud of 3D points of the surface.
    [15]
    A method according to any of claims 13 or 14, wherein the step of combining the sensor data and position data further comprises deriving a number of local surface characteristics at a position of the end of the tool or in the vicinity thereof from the analytical model.
    [16]
    A method according to any of claims 12 to 15, wherein the desired robot trajectory is generated based on the local surface model and the step of controlling the end effector according to the desired robot trajectory includes sending successive control signals.
    [17]
    A method according to claim 16, wherein the position of the end of the tool is the position of the end of the tool corresponding to the time at which the control signal will have its effective effect.
    [18]
    A method according to claim 14, wherein the step of determining parameters of the analytical model comprises estimating the parameters taking into account initial parameter values that are stored, learned from previous execution of the surface tracking task, or available on any other which other way.
    [19]
    A method according to any one of the preceding claims, wherein the step of controlling the end effector according to the desired robot trajectory comprises sending successive control signals and wherein the control signals comprise a pre-coupling component based on the desired robot trajectory.
    [20]
    A method according to any of the preceding claims, wherein the step of controlling the end effector according to the desired robot trajectory comprises sending successive control signals and wherein the control signals contain at least one feedback component.
    [21]
    A method according to claim 20, wherein the at least one feedback component contains the position data or a component thereof.
    [22]
    A method according to claim 21, wherein the feedback position data or component thereof corresponds in time to a predicted position of the tool relative to the workpiece to compensate for any delay in the feedback control.
    [23]
    A method according to any of claims 20 to 22, wherein the at least one feedback component contains data corresponding to contact force or moment of contact between the tool and the surface of the workpiece.
    [24]
    A device for tracking the surface of a workpiece, which includes a tool for performing an action on the tracked surface, a positioning device for relatively positioning the tool configured to perform an action relative to the surface, and wherein the positioning device is configured to provide position data, first sensor means configured such that, in use, the surface on which the action is to be exerted is first sensed by the first sensor means to create sensor data before being affected by the tools and a controller configured to create a local surface model of a local part of the surface to be tracked based on sensor data and received position data, and to generate a desired trajectory for the positioning device based on the local surface model.
    [25]
    A device according to claim 24, wherein the first sensor means comprises at least one one-dimensional laser displacement sensor.
    [26]
    A device according to any of claims 24 to 25, wherein the first sensor means comprises three one-dimensional laser displacement sensors.
    [27]
    A device according to any of claims 24 to 26, wherein the first sensor means comprises at least one two-dimensional laser displacement sensor.
    [28]
    A device according to any of claims 24 to 27, wherein the device further comprises second sensor means and wherein the second sensor means are configured such that, during use, the surface on which the action is to be applied is first affected by the tool before being perceived by the second sensor means to create second sensor data.
    [29]
    A device according to any of claims 24 to 28, wherein the end effector is configured to hold the tool.
    [30]
    A device according to any of claims 24 to 28, wherein the end effector is configured to hold the workpiece.
    类似技术:
    公开号 | 公开日 | 专利标题
    US11110611B2|2021-09-07|Automatic detection and robot-assisted machining of surface defects
    US11202682B2|2021-12-21|Techniques for modifying tool operation in a surgical robotic system based on comparing actual and commanded states of the tool relative to a surgical site
    JP2015212012A|2015-11-26|Method of controlling robot tool
    US20200038954A1|2020-02-06|Apparatus and method for additive manufacturing
    JP6445070B2|2018-12-26|Machine tool control system
    US20170357239A1|2017-12-14|Method for compensating for a deviation in an operating point
    CN103764339A|2014-04-30|Laser cutting machine
    US11179793B2|2021-11-23|Automated edge welding based on edge recognition using separate positioning and welding robots
    BE1024133B1|2017-11-20|Method and device for monitoring and treating a surface by a robot
    CN111014879B|2020-12-08|Automatic welding method for corrugated plate of robot based on laser weld seam tracking
    WO2017103489A1|2017-06-22|System and method for correcting a trajectory of an effector carried by a robot
    JP2022504092A|2022-01-13|Methods and equipment for monitoring the machining process of workpieces using a laser beam
    JP2019217563A|2019-12-26|Robot system and robot
    Dhanaraj et al.2022|A mobile manipulator system for accurate and efficient spraying on large surfaces
    Nguyen et al.2011|High precision laser tracker system for contactless position measurement
    Chalus et al.2018|Calibration and using a laser profile scanner for 3D robotic welding
    WO2021235331A1|2021-11-25|Following robot
    Bestard2020|Online Measurements in Welding Processes
    CN108700400B|2021-08-24|Measuring method and measuring device for detecting surface topological structure of workpiece
    US20220016776A1|2022-01-20|Real time feedback and dynamic adjustment for welding robots
    Ryll et al.2019|Accelerator for Ultrafast Laser Serial Production: Interaction of important components opens up new possibilities
    JP7000361B2|2022-01-19|Follow-up robot and work robot system
    JPWO2020175425A1|2021-11-18|Robot systems, robot controls, and robot control programs
    JP6816060B2|2021-01-20|Work robot system and work robot
    Mewes et al.2021|The Correction of the Nozzle-Bed-Distance in Robotic Fused Deposition Modeling
    同族专利:
    公开号 | 公开日
    BE1024133B9|2017-12-19|
    BE1024133A9|2017-12-19|
    BE1024133A1|2017-11-17|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US5300868A|1991-01-28|1994-04-05|Fanuc Ltd.|Robot teaching method|
    EP2258513A1|2008-03-31|2010-12-08|Honda Motor Co., Ltd.|Work mounting system and method of mounting work|
    DE102013110667A1|2013-09-26|2015-03-26|Deutsches Zentrum für Luft- und Raumfahrt e.V.|Method for the non-destructive inspection of three-dimensional workpieces and apparatus for carrying out such a method|
    法律状态:
    2018-02-12| FG| Patent granted|Effective date: 20171120 |
    优先权:
    申请号 | 申请日 | 专利标题
    BE20165281A|BE1024133B9|2016-04-22|2016-04-22|Method and device for following and treating a surface by a robot|BE20165281A| BE1024133B9|2016-04-22|2016-04-22|Method and device for following and treating a surface by a robot|
    [返回顶部]