JP7433509B2 - Control device, robot system, learning device, trajectory model, control method, and program - Google Patents

Description

The present disclosure relates to a control device, a robot system, a learning device, a trajectory model, a control method, and a program.

Industrial robots include multiple links and multiple joints connecting the multiple links to enable various movements. A control device controls a plurality of servo motors that drive a plurality of joints of a robot, thereby enabling the robot to perform a desired operation. For example, if a difference occurs in the delay of each servo motor, the actual trajectory of the tip link, which is the link located at the tip of the plurality of links, may deviate from the target trajectory. An example of a technique for minimizing the error between the actual trajectory of a robot and a target trajectory is disclosed in Patent Document 1.

The robot position teaching device disclosed in Patent Document 1 measures the actual trajectory of the robot and performs prospective teaching to minimize the error based on the trajectory error, which is the distance between the actual trajectory and the target trajectory. Calculate points. Controlling the robot based on expected teaching points reduces trajectory errors.

Japanese Patent Application Publication No. 11-048176

Even if control is performed to bring the actual trajectory of the robot's tip closer to the target trajectory as described above, the tip link will receive force when attaching a processing tool to the tip link and machining the workpiece. trajectory deviates from the target trajectory. As a result, the control accuracy of the robot during machining is reduced.

The present disclosure has been made in view of the above circumstances, and aims to provide a control device, a robot system, a learning device, a trajectory model, a control method, and a program that can control a robot with high accuracy during machining.

In order to achieve the above object, a control device according to the present disclosure is a control device that controls a robot to which a processing tool for processing a workpiece can be attached, and includes a drive condition storage unit and a reference trajectory determination unit. , an actual trajectory acquisition section, a learning section, a target trajectory determination section, and a control section. The drive condition storage unit stores drive conditions that specify at least a motion start point and a motion end point of the robot in association with a reference intermediate point indicating a position that the robot passes from the motion start point to the motion end point. There is. The reference trajectory determining unit acquires a reference intermediate point from the driving condition storage unit, and determines a reference trajectory passing through the acquired reference intermediate point based on the driving condition. The actual trajectory acquisition unit acquires the actual trajectory of the robot. When machining the workpiece by controlling the robot according to the reference trajectory, the learning section is configured to detect the deviation of the actual trajectory from the ideal trajectory based on the driving conditions of the actual trajectory acquired by the actual trajectory acquisition section. The trajectory error is determined for the obtained reference intermediate point, the correspondence between the reference intermediate point and the machining trajectory error is learned, and a trajectory model is generated that indicates the target intermediate point that minimizes the machining trajectory error according to the driving conditions. do. The target trajectory determination unit determines a target trajectory passing through a target intermediate point obtained from the trajectory model. The control unit controls the robot according to the reference trajectory or the target trajectory.

A control device according to the present disclosure learns the correspondence between a reference intermediate point and a trajectory error during machining, generates a trajectory model that indicates a target intermediate point that minimizes the trajectory error during machining according to driving conditions, and generates a trajectory model that indicates a target intermediate point that minimizes the trajectory error during machining. Control the robot according to the target trajectory passing through the points. Therefore, the control accuracy of the robot during processing by the control device according to the present disclosure is high.

A perspective view of a robot according to Embodiment 1 Side view of the robot according to Embodiment 1 Block diagram of robot system according to Embodiment 1 Block diagram showing the hardware configuration of the control device according to Embodiment 1 Flowchart showing an example of learning processing performed by the control device according to Embodiment 1 A diagram showing an example of a driving condition table held by the driving condition storage unit according to Embodiment 1. A diagram showing an example of a reference trajectory and an actual trajectory in Embodiment 1 A diagram showing an example of an error table held by the learned data storage unit according to Embodiment 1. A diagram showing an example of a target intermediate point table held by the learned data storage unit according to Embodiment 1. Flowchart showing an example of operation processing performed by the control device according to Embodiment 1 A diagram showing an example of a target trajectory and an actual trajectory in Embodiment 1 Block diagram of robot system according to Embodiment 2 Flowchart showing an example of learning processing performed by the control device according to Embodiment 2 A diagram showing an example of an error table held by the learned data storage unit according to Embodiment 2. Flowchart showing an example of operation processing performed by the control device according to Embodiment 2 Block diagram of robot system according to Embodiment 3 Flowchart showing an example of learning processing performed by the control device according to Embodiment 3 A diagram showing an example of an error table held by the learned data storage unit according to Embodiment 3 Block diagram of robot system according to Embodiment 4 A diagram showing a modified example of the target intermediate point table held by the learned data storage unit according to the embodiment. A diagram showing a first modification of the error table held by the learned data storage unit according to the embodiment. A diagram showing a second modification of the error table held by the learned data storage unit according to the embodiment. A diagram showing a third modification of the error table held by the learned data storage unit according to the embodiment. Block diagram of a modified example of the robot system according to the embodiment A perspective view of a first modification of the robot according to the embodiment A perspective view of a second modification of the robot according to the embodiment

Hereinafter, a control device, a robot system, a learning device, a trajectory model, a control method, and a program according to embodiments of the present disclosure will be described in detail with reference to the drawings. In the figures, the same or equivalent parts are given the same reference numerals.

(Embodiment 1)
In Embodiment 1, a robot system including a robot and a control device that controls the robot will be described using a robot including a plurality of links and a plurality of joints as an example. The robot includes a plurality of links, a plurality of joints that connect the plurality of links to each other, and a plurality of motors that are associated with the plurality of joints and drive the corresponding joints.

As shown in FIGS. 1 and 2, the robot 10 is a vertically articulated robot. Specifically, the robot 10 includes a first arm 110, a second arm 120, a third arm 130, a fourth arm 140, a fifth arm 150, a flange 160, and a base 170 as a plurality of links. , is provided. The flange 160 corresponds to a tip link that is a link located at the tip of the plurality of links.

The robot 10 further includes a fixed part 180 that is fixed to a flat surface and supports the base 170, and a force sense that measures the reaction force that the flange 160 receives when processing the workpiece 300 with the processing tool 200 that is attached to the flange 160. A sensor 190 is provided. In the first embodiment, the fixing part 180 is a plate-like member whose main surface has a square shape.

1 and 2, the X-axis and Y-axis are set as axes that are included in a plane parallel to the main surface of the fixing part 180 and are orthogonal to each other, and the Z-axis is set as an axis that is orthogonal to each of the X-axis and Y-axis. Set. The X-axis and the Y-axis each extend along the end surface of the fixed part 180. If the fixed part 180 is fixed on a horizontal plane, the Z axis extends in the vertical direction. In the first embodiment, the trajectory of the tip link, specifically, the trajectory of the center of gravity of the tip surface 160a of the flange 160, is expressed in an XYZ orthogonal coordinate system. The origin of the XYZ orthogonal coordinate system is, for example, the position of the center of gravity of the tip surface 160a when the robot 10 is at the initial position. When the robot 10 is at the initial position, it means when each of the rotational positions of motors M1, M2, M3, M4, M5, and M6, which will be described later, is at the initial position.

The first arm 110 is attached to the base 170 in a rotatable state around a rotation axis AX1 parallel to the Z-axis. The second arm 120 is connected to the first arm 110 and is rotatable around the rotation axis AX2. The third arm 130 is connected to the second arm 120 and is rotatable around the rotation axis AX3. The fourth arm 140 is connected to the third arm 130 and is rotatable around a rotation axis AX4. The fifth arm 150 is connected to the fourth arm 140 and is rotatable around the rotation axis AX5. The flange 160 is connected to the fifth arm 150 and is rotatable around the rotation axis AX6. In other words, the robot 10 has six joints corresponding to the rotation axes AX1, AX2, AX3, AX4, AX5, and AX6.

The control device 1 controls the robot 10 to which the processing tool 200 for processing the workpiece 300 is attached, so that the processing tool 200 can process the workpiece 300. For example, as shown in FIG. 2, the control device 1 controls the robot 10 to move the flange 160 in the positive direction of the X-axis with the distal end surface of the flange 160 parallel to the plane on which the robot 10 is fixed. let As a result, the processing tool 200 attached to the flange 160 comes into contact with the side surface of the workpiece 300 that intersects with the X-axis direction and further moves in the positive direction of the X-axis, removing the burr 310 from the workpiece 300. can be removed. The burr 310 is a protrusion formed during molding of the workpiece 300, and is a portion that protrudes from the target outer shape of the workpiece 300.

As shown in FIG. 3, the robot 10 includes motors M1, M2, M3, M4, M5, and M6 corresponding to rotational axes AX1, AX2, AX3, AX4, AX5, and AX6. Motors M1, M2, M3, M4, M5, and M6 are servo motors. The control device 1 controls the motors M1, M2, M3, M4, M5, and M6 according to the driving conditions, so that the motors M1, M2, M3, M4, M5, and M6 rotate, and the first arm 110 and the second arm Arm 120, third arm 130, fourth arm 140, fifth arm 150, and flange 160 rotate. As a result, the trajectory of the flange 160, which is the tip link of the robot 10, becomes a trajectory that corresponds to the driving conditions. The robot 10 further includes encoders E1, E2, E3, E4, E5, and E6 that detect the rotational positions of the motors M1, M2, M3, M4, M5, and M6.

The robot system 100 includes a robot 10 having the above configuration and a control device 1 that controls the robot 10. The control device 1 performs control to bring the trajectory of the robot 10, specifically, the trajectory of the flange 160, closer to an ideal trajectory in accordance with drive conditions that specify at least an operation start point and an operation end point of the robot 10. The operation start point of the robot 10 means the operation start point of the movable part of the robot 10, specifically, the operation start point of the flange 160 included in the robot 10. The operation end point of the robot 10 means the operation end point of the movable part of the robot 10, specifically, the operation end point of the flange 160 included in the robot 10.

In the first embodiment, the control device 1 controls the motors M1, M2, M3, M4, M5, and Control M6. By controlling the robot 10 to which the processing tool 200 is attached as described above, the control device 1 performs deburring to remove the burr 310 of the workpiece 300 shown in FIG. 2, attachment of parts to the workpiece 300, etc. processing becomes possible.

When controlling the robot 10 according to drive conditions to process the workpiece 300, the actual trajectory of the tip link deviates from the ideal trajectory based on the drive conditions. In the first embodiment, the ideal trajectory is the shortest trajectory from the operation start point indicated by the drive conditions to the operation end point indicated by the drive conditions.

The deviation from the ideal trajectory includes deviations caused by mechanical factors and deviations caused by external factors. The deviation caused by a mechanical factor is, for example, a deviation caused by a difference in the delay in the operation of the motors M1, M2, M3, M4, M5, and M6 when the workpiece 300 is not processed. Misalignment caused by external factors is caused by, for example, gravity, centrifugal force, processing reaction force, etc. This is a deviation caused by mechanical deflection of at least one of the flange 160 and the flange 160.

In order to minimize the deviation caused by mechanical factors, the control device 1 sets a reference intermediate point corresponding to the driving conditions to minimize the trajectory error during non-machining, which indicates the deviation of the actual trajectory during non-machining from the ideal trajectory. I remember putting it on. The non-processing actual trajectory indicates the trajectory of the tip link of the robot 10 that is controlled according to the drive conditions when the processing object 300 is not processed. The reference intermediate point indicates the position that the tip link passes from the motion start point to the motion end point indicated by the drive conditions.

In order to minimize the deviation caused by machining reaction force, which is an example of an external factor, the control device 1 calculates the machining trajectory error that indicates the deviation of the actual machining trajectory from the ideal trajectory, and calculates the machining trajectory error between the reference intermediate point and the machining trajectory. Learn the correspondence with errors. The processing reaction force is a force that the flange 160 receives when processing the workpiece 300. The actual locus during machining indicates the locus of the tip link when machining the workpiece 300 by controlling the robot 10 according to the reference locus passing through the reference intermediate point. After learning the correspondence between the reference intermediate point and the trajectory error during machining as described above, the control device 1 generates a trajectory model that indicates the target intermediate point that minimizes the trajectory error during machining according to the driving conditions, and sets the target intermediate point. The robot 10 is controlled according to the target trajectory passing through the points. By controlling the robot 10 according to the target trajectory, it is possible to improve the control accuracy of the robot 10 during machining.

Each part of the control device 1 will be explained below.
The control device 1 includes a drive condition storage unit 11 that stores drive conditions and reference intermediate points in association with each other, a reference trajectory determination unit 12 that determines a reference trajectory passing through the reference intermediate points, and an actual trajectory of the robot 10. Specifically, it includes an actual trajectory acquisition unit 13 that acquires the actual movement trajectory of the flange 160, which is the tip link. Furthermore, when processing the workpiece 300 by controlling the robot 10 according to the reference trajectory, the control device 1 further selects the ideal trajectory from the actual trajectory during machining, which indicates the actual trajectory acquired by the actual trajectory acquisition unit 13. The machining trajectory error indicating the deviation is determined for each reference intermediate point. The control device 1 includes a learning unit 14 that learns the correspondence between the reference intermediate point and the machining trajectory error, and generates a trajectory model that indicates a target intermediate point that minimizes the machining trajectory error according to the driving conditions.

The control device 1 further determines a target trajectory passing through a target intermediate point according to the drive conditions based on a learned data storage unit 15 that stores the learning results of the learning unit 14 and a drive command including a trajectory model and drive conditions. A target trajectory determination unit 16 is provided. The control device 1 includes a control unit 17 that controls the robot 10, specifically, motors M1, M2, M3, M4, M5, and M6 according to a reference trajectory or a target trajectory, and a force sensor 190 that the robot 10 includes. A reaction force acquisition unit 18 is provided that acquires a sensor signal and calculates a reaction force, which is a force that the robot 10 receives during machining, from the sensor signal.

As shown in FIG. 4, the control device 1 includes a processor 31, a memory 32, and an interface 33 as a hardware configuration. Processor 31, memory 32 and interface 33 are connected to each other by bus 34. Each function of the control device 1 is realized by the processor 31 executing a program stored in the memory 32. The interface 33 connects the control device 1 and an external device and enables communication with the external device. In detail, the control device 1 is connected to the motors M1, M2, M3, M4, M5, M6, encoders E1, E2, E3, E4, E5, E6, and the force sensor 190 via the interface 33. . The interface 33 has multiple types of interface modules as necessary.

Although the control device 1 shown in FIG. 4 has one processor 31 and one memory 32, the control device 1 may have multiple processors 31 and multiple memories 32. In this case, each function of the control device 1 may be realized by the plurality of processors 31 and the plurality of memories 32 working together.

The control device 1 having the above configuration controls the robot 10 according to the reference trajectory passing through the reference intermediate point for each drive condition, and sets the target intermediate point that minimizes the trajectory error during machining according to the drive conditions. A learning process for generating a trajectory model shown in the figure and an operation process for controlling the robot 10 according to the target trajectory passing through the target intermediate point are performed. Minimizing the trajectory error means bringing the trajectory error close to the minimum value, a value near the minimum value, or a minimum value or a value near the minimum value.

An overview of the operation of the learning process performed by the control device 1 will be described using FIG. 5. For example, after the robot 10 is installed, the control device 1 starts the learning process shown in FIG. 5 when the control device 1 starts operating for the first time or when a new drive condition is added. The learning process in FIG. 5 is performed while the machining tool 200 is attached to the flange 160 and machining the workpiece 300. In other words, the workpiece 300 is machined in order for the control device 1 to obtain a machining trajectory error for each drive condition.

The reference trajectory determining unit 12 acquires the driving condition and the reference intermediate point from the driving condition storage unit 11 that stores the driving condition and the reference intermediate point in association with each other (step S11). Then, the reference trajectory determination unit 12 determines a reference trajectory passing through the reference intermediate point based on the driving conditions and the reference intermediate point acquired in step S11 (step S12). The control unit 17 controls the motors M1, M2, M3, M4, M5, and M6 of the robot 10 according to the reference trajectory determined in step S12, and the actual trajectory acquisition unit 13 acquires the actual trajectory of the robot 10. Then, the reaction force is calculated (step S13).

Specifically, the control unit 17 determines the target positions of the motors M1, M2, M3, M4, M5, and M6 in order to move the flange 160 according to the reference trajectory, and controls the motors M1, M2, M3, M4, M5, and M6. Performs control to move to the target position. The actual trajectory acquisition unit 13 acquires the actual trajectory of the robot 10 based on the rotational positions of the motors M1, M2, M3, M4, M5, and M6 acquired from encoders E1, E2, E3, E4, E5, and E6. . The reaction force acquisition unit 18 calculates the reaction force based on the sensor signal acquired from the force sensor 190. The learning unit 14 calculates a machining trajectory error indicating a deviation of the trajectory acquired in step S13 from the ideal trajectory based on the driving conditions acquired in step S11, and calculates the machining trajectory error and the machining trajectory error calculated in step S14. and the reaction force (step S14). As long as the association between machining trajectory error and reaction force is not completed for all drive conditions stored in the drive condition storage unit 11 (step S15; No), the processes of steps S11 to S14 described above are repeated. .

When the correspondence between the machining trajectory error and the reaction force is completed for all the drive conditions stored in the drive condition storage unit 11 (step S15; Yes), the learning unit 14 determines the target intermediate that minimizes the machining trajectory error. The points are determined, and a trajectory model indicating the target intermediate point is generated according to the driving conditions (step S16). When the process of step S16 is completed, the control device 1 ends the learning process.

Details of the learning process performed by the control device 1 will be described below.
As shown in FIG. 6, the driving condition storage unit 11 holds a driving condition table that associates driving conditions with reference midpoints. The driving conditions specify at least the operation start point and operation end point of the flange 160, which is the tip link. In the first embodiment, the driving conditions include the operation start point, operation end point, speed, and attitude of the tip link. The operation start point and the operation end point indicate positions in the XYZ orthogonal coordinate system shown in FIGS. 1 and 2. The speed indicates the target speed during operation of the tip link. The posture indicates the direction of the tip link, specifically, the angle formed between the tip surface 160a of the flange 160 and the plane on which the robot 10 is fixed.

For example, the drive conditions included in the first row record of the drive condition table shown in FIG. 0,0) to the operation end point (100,0,0) at the target speed V1. In this case, the ideal trajectory extends from the operation start point SP1 (0, 0, 0) to the operation end point EP1 (100, 0, 0), and is indicated by a line segment located on the X axis.

In the first embodiment, a reference midpoint is determined for each drive condition. As described above, the reference midpoint minimizes the deviation caused by the difference in the delay in the operation of the motors M1, M2, M3, M4, M5, and M6 when the workpiece 300 is not processed. Therefore, the position that the tip link passes from the operation start point to the operation end point indicated by the drive conditions is shown. For example, when the robot 10 is controlled according to a trajectory passing through the initial intermediate points without processing the workpiece 300 for each of a plurality of initial intermediate points located between the operation start point and the operation end point. The deviation of the actual trajectory of the tip link from the ideal trajectory may be calculated, and the initial intermediate point that minimizes the deviation may be used as the reference intermediate point.

The reference trajectory determination unit 12 acquires the associated driving conditions and reference intermediate points from the driving condition table shown in FIG. Then, the reference trajectory determination unit 12 determines a reference trajectory passing through the acquired reference intermediate point based on the driving conditions. Specifically, the reference trajectory determination unit 12 performs spline interpolation based on the operation start point, reference intermediate point, and operation end point indicated by the drive conditions, and moves the reference trajectory from the operation start point indicated by the drive conditions to the reference intermediate point. A reference trajectory is calculated that passes through the point and reaches the operation end point indicated by the driving conditions. The reference trajectory indicates, for example, the position of the tip link in the XYZ orthogonal coordinate system for each control cycle. The control period is determined, for example, according to the arithmetic processing capacity of the control device 1.

For example, the reference trajectory determining unit 12 acquires the driving conditions and the reference intermediate point indicated by the record in the first row of the driving condition table shown in FIG. The operation start point (0, 0, 0) and operation end point (100, 0, 0) included in this drive condition are shown in FIG. 7 as the operation start point SP1 and the operation end point EP1, respectively, and the reference midpoint (50 , -2.5, 0) is shown in FIG. 7 as the reference midpoint RP1. The reference trajectory determination unit 12 determines a reference intermediate point RP1 (50, -2 .5, 0) is determined. The reference trajectory RT1 indicated by a dotted line in FIG. 7 is a smooth curve from the operation start point SP1 through the reference intermediate point RP1 to the operation end point EP1, and is indicated by a curve that protrudes in the negative direction of the Y-axis in the XY plane. . After calculating the reference trajectory as described above, the reference trajectory determination unit 12 sends the driving conditions and the determined reference trajectory to the learning unit 14 and the control unit 17.

The control unit 17 shown in FIG. 3 generates control commands for the motors M1, M2, M3, M4, M5, and M6 according to the drive conditions and the reference trajectory acquired from the reference trajectory determining unit 12, and sends the control commands to the motors M1, M2, M3, M4, M5, and M6. , M2, M3, M4, M5, and M6. Specifically, the control unit 17 generates control commands for the motors M1, M2, M3, M4, M5, and M6 according to a reference position indicating the position of the tip link indicated by the reference locus, and transmits the control commands to the motors M1, M2, M3, M4, M5, and M6. Transmission to M2, M3, M4, M5, and M6 is performed every control cycle. As a result, the motors M1, M2, M3, M4, M5, and M6 rotate according to the control command, and the first arm 110, the second arm 120, the third arm 130, the fourth arm 140, the fifth arm 150, and the flange 160 rotates. This causes the flange 160, which is the tip link, to move.

The control unit 17 acquires the rotational positions of the motors M1, M2, M3, M4, M5, and M6 from the encoders E1, E2, E3, E4, E5, and E6, and controls the motors M1, M2, M3, M4, M5, and M6. Feedback control is preferable.

While the motors M1, M2, M3, M4, M5, and M6 are being driven by the control unit 17 as described above, the force sensor 190 measures the reaction force that is the force that the flange 160 receives, and converts it into a measured value. Outputs the corresponding sensor signal.

The reaction force acquisition unit 18 detects the sensor signal acquired from the force sensor 190 included in the robot 10 in each control cycle, and calculates the reaction force, which is the force that the robot 10 receives during machining, from the value detected in each control cycle. Then, the calculated reaction force is sent to the control section 17.

The actual trajectory acquisition unit 13 acquires the rotational positions of the motors M1, M2, M3, M4, M5, M6 from the encoders E1, E2, E3, E4, E5, E6, and Obtain the actual position of the tip link from the rotational position of M6. The actual trajectory acquisition unit 13 performs the above-mentioned process for each control cycle, and determines an actual trajectory that indicates the actual position of the tip link in the XYZ orthogonal coordinate system for each control cycle.

As an example, when the motors M1, M2, M3, M4, M5, and M6 are controlled according to the driving conditions indicated by the first record of the driving condition table shown in FIG. 6 and the reference trajectory RT1 based on the reference intermediate point RP1. , the actual machining trajectory AT1 acquired by the actual trajectory acquisition unit 13 is shown by a solid line in FIG. The actual trajectory acquisition unit 13 sends the actual trajectory acquired as described above to the learning unit 14. When the workpiece 300 is not machined, when the motors M1, M2, M3, M4, M5, and M6 are controlled according to the reference trajectory RT1, the tip link moves along the actual trajectory AT0 shown by the broken line in FIG. , passes through a trajectory close to the ideal trajectory. However, due to the reaction force received during machining of the workpiece 300, the actual machining trajectory AT1 deviates from the ideal trajectory.

When the learning unit 14 acquires the actual trajectory from the actual trajectory acquisition unit 13, the learning unit 14 acquires the actual trajectory by the actual trajectory acquisition unit 13 when the workpiece 300 is machined by controlling the robot 10 according to the reference trajectory. The machining trajectory error, which indicates the deviation of the actual machining trajectory from the ideal trajectory, is calculated. Specifically, the learning unit 14 determines an ideal trajectory indicating the shortest trajectory from the operation start point to the operation end point based on the driving conditions, and determines the position of the tip link for each control cycle on the ideal trajectory. Then, the learning unit 14 determines the position of the robot 10 for each control cycle indicated by the ideal trajectory, specifically, the ideal position of the tip link of the robot 10 and the position of the robot 10 for each control cycle indicated by the actual trajectory during machining. The position, specifically, the distance from the actual position, which is the position of the tip link, is calculated. The learning unit 14 uses the deviation of the actual position from the ideal position when the distance reaches the maximum value as a machining trajectory error.

For example, when the learning unit 14 acquires the driving conditions and reference trajectory indicated by the first record of the driving condition table shown in FIG. An ideal trajectory that is the shortest trajectory to point EP1 (100, 0, 0) is calculated. In this case, the ideal trajectory is a line segment located on the X-axis that extends from the motion start point SP1 (0,0,0) to the motion end point EP1 (100,0,0) shown in FIG. Then, the learning unit 14 calculates the distance between the ideal position of the tip link for each control cycle indicated by the ideal trajectory and the actual position of the tip link for each control cycle indicated by the actual machining trajectory AT1. The learning unit 14 uses the deviation of the actual position from the ideal position when this distance reaches the maximum value as the machining trajectory error, as shown by the arrow ER1 in FIG. In the example of FIG. 7, the actual position is shifted from the ideal position in the positive direction of the Y-axis on the XY plane.

The learning unit 14 further associates the reaction force with the machining trajectory error. Specifically, the learning unit 14 associates the machining trajectory error with the reaction force when the distance between the ideal position and the actual position becomes the maximum value.

The learning unit 14 performs the above-described process for all drive conditions stored in the drive condition table shown in FIG. 6 . The learning unit 14 calculates the machining trajectory error for each drive condition, and learns the correspondence between the drive condition, the reference midpoint, the machining trajectory error, and the reaction force. The learning unit 14 then generates an error table shown in FIG. 8 and stores it in the learned data storage unit 15. In the example of FIG. 8, the machining trajectory error is stored as a vector component. Specifically, the machining trajectory error is expressed by a component of a vector starting from the ideal position and indicating a deviation from the actual position. In the example of FIG. 8, the reaction force represented by (Fx, Fy, Fz) is stored in the error table.

When the correspondence between the machining trajectory error and the reaction force is completed for all drive conditions, the learning unit 14 determines a target intermediate point that minimizes the machining trajectory error from the correspondence between the reference intermediate point, the machining trajectory error, and the reaction force. is determined, and a trajectory model indicating the target intermediate point is generated according to the driving conditions. Specifically, the learning unit 14 determines the target intermediate point at which the machining trajectory error is minimized from the ideal trajectory, the reference trajectory passing through the reference target point, the machining trajectory error, and the reaction force for each driving condition. For example, the learning unit 14 adjusts the position of the reference midpoint according to the reaction force, and sets the adjusted reference midpoint as the target midpoint. In the example of the record in the first row of the error table shown in FIG. 8, the magnitude of the Y-axis component of the reaction force is larger than the magnitudes of the X-axis component and the Z-axis component. Therefore, in order to minimize the influence of the reaction force applied in the positive direction of the Y-axis, the learning unit 14 adjusts the reference midpoint in the negative direction of the Y-axis, and sets the adjusted reference midpoint as the target midpoint. In the first embodiment, the learning unit 14 generates the target intermediate point table shown in FIG. 9 as a trajectory model, and stores it in the learned data storage unit 15.

An overview of operation processing, which is control processing for the robot 10, performed after the above-described learning processing is completed will be described using FIG. 10. For example, when the control device 1 receives a drive command including drive conditions for the robot 10 through an operation from an operation unit (not shown) after the learning process is completed, the control device 1 starts the operation process shown in FIG. 10 . The operation process in FIG. 10 is performed while the machining tool 200 is attached to the flange 160 and machining the workpiece 300.

Upon acquiring the drive command including the drive conditions, the target trajectory determination unit 16 acquires a target intermediate point according to the drive conditions indicated by the drive command from the learned data storage unit 15 (step S21). Then, the target trajectory determination unit 16 determines a target trajectory passing through the target intermediate point acquired in step S21 based on the driving conditions (step S22). The control unit 17 controls the motors M1, M2, M3, M4, M5, and M6 of the robot 10 according to the target trajectory determined in step S22 (step S23).

Details of the operational processing performed by the control device 1 will be described below.
Upon acquiring the drive command, the target trajectory determination unit 16 acquires a target waypoint according to the drive condition indicated by the drive command from the target waypoint table stored in the learned data storage unit 15. Then, the target trajectory determination unit 16 calculates a target trajectory representing a trajectory from the operation start point indicated by the drive conditions, passing through the target intermediate point, and ending at the operation end point indicated by the drive conditions. Specifically, the target trajectory determination unit 16 performs spline interpolation based on the motion start point, the target intermediate point, and the motion end point indicated by the drive conditions, and moves the target midway point from the motion start point indicated by the drive conditions. A target trajectory passing through the point and reaching the operation end point indicated by the driving conditions is determined. The target trajectory indicates the target position of the tip link for each control cycle.

For example, the target trajectory determination unit 16 moves the flange 160 from a motion start point (0,0,0) to a motion end point (100 , 0, 0) at the target speed V1. The target trajectory determination unit 16 acquires the target intermediate point (50, -3.5, 0) indicated by the record in the first row of the target intermediate point table shown in FIG. 9 that corresponds to the acquired driving condition. This target halfway point (50, -3.5, 0) is shown in FIG. 11 as the target halfway point TP1. Then, the target trajectory determining unit 16 moves from the motion start point SP1 (0,0,0) to the motion end point EP1 (100,0,0) through the target intermediate point TP1 (50, -3.5,0). A target trajectory TT1 to be reached is determined. As shown by the dotted line in FIG. 11, the target trajectory TT1 is a smooth curve from the movement start point SP1 to the movement end point EP1 through the target intermediate point TP1, and is a curve that protrudes in the Y-axis negative direction in the XY plane. It is indicated by. After determining the target trajectory as described above, the target trajectory determination unit 16 sends the drive conditions and the calculated target trajectory to the control unit 17.

The control unit 17 shown in FIG. 3 generates control commands to the motors M1, M2, M3, M4, M5, and M6 according to the target trajectory acquired from the target trajectory determination unit 16, and transmits the control commands to the motors M1, M2, Send to M3, M4, M5, M6. As a result, the motors M1, M2, M3, M4, M5, and M6 rotate according to the control command, and the first arm 110, second arm 120, third arm 130, fourth arm 140, fifth arm 150, and flange 160 rotate. Rotate. This causes the flange 160, which is the tip link, to move. By moving the flange 160 to which the processing tool 200 is attached, the workpiece 300 is processed by the processing tool 200. By controlling the motors M1, M2, M3, M4, M5, and M6 according to the target trajectory that minimizes the trajectory error during machining, the actual trajectory AT1' of the tip link during machining is ideal, as shown in Fig. 11. Approaching the trajectory.

As explained above, the control device 1 according to the first embodiment learns the correspondence between the reference intermediate point and the trajectory error during machining, and further determines the target intermediate point that minimizes the trajectory error during machining based on the reaction force. A trajectory model is generated depending on the driving conditions. Processing of the workpiece 300 is performed by the control device 1 controlling the robot 10 according to the target trajectory passing through the target intermediate point. Since the target trajectory is adjusted according to the reaction force, the actual trajectory during machining approaches the ideal trajectory, and the control accuracy of the robot 10 by the control device 1 during machining increases.

(Embodiment 2)
The method of controlling the robot 10 is not limited to the example of the first embodiment, and may be any method that takes into account the reaction force during machining. Regarding the control device 2 that controls the motors M1, M2, M3, M4, M5, and M6 according to information about the workpiece 300 that causes fluctuations in reaction force in order to improve control accuracy of the robot 10 during processing. , will be described in Embodiment 2 focusing on the points that are different from the control device 1 according to Embodiment 1. The information about the workpiece 300 includes the weight, shape, material, etc. of the workpiece 300.

A robot system 101 according to the second embodiment shown in FIG. 12 includes a control device 2 and a robot 10. The control device 2 learns the workpiece identification unit 19 that acquires information about the workpiece 300 and the information about the workpiece 300 acquired by the workpiece identification unit 19, and determines the target intermediate point based on the driving conditions and the workpiece identification unit 19. A learning unit 20 that generates a trajectory model shown in accordance with information about the workpiece 300, and a target trajectory that determines a target trajectory passing through a target intermediate point based on drive commands including drive conditions and information about the workpiece 300. A determination unit 21 is provided. Unlike the first embodiment, the robot 10 does not include a force sensor 190.

As the control device 2 controls the robot 10, the processing tool 200 performs deburring to remove the burr 310 from the workpiece 300, as in the first embodiment. The workpiece 300 is, for example, a urethane foam insulation material. During molding of the workpiece 300, burrs 310 are generated. When the workpiece 300 is formed of a urethane foam insulation material, the thickness, length, and other shapes of the burr 310 vary. The thickness of the burr 310 is the thickness of the burr 310 in the extending direction of the rod-shaped processing tool 200 shown in the example of FIG. 2, that is, in the Z-axis direction. The length of the burr 310 is the length of the burr 310 in the moving direction of the processing tool 200 shown in the example of FIG. 2, that is, in the X-axis direction. As the shape of the burr 310 changes, the reaction force that is the force that the flange 160 receives during processing changes, and the trajectory error during processing changes. Therefore, the control device 2 controls the robot 10 according to the target trajectory passing through the target intermediate point according to the driving conditions and the shape of the burr 310. As a result, even if the shape of the burr 310 changes, the control accuracy of the robot 10 during processing can be improved.

The workpiece specifying unit 19 shown in FIG. 12 includes a photographing device 191 that generates image data of the workpiece 300, and a shape determining unit 192 that determines the shape of the workpiece 300 from the image data generated by the photographing device 191. , has.

In the second embodiment, the workpiece specifying unit 19 acquires the shape of the burr 310 on the workpiece 300 and sends the acquired shape of the burr 310 to the learning unit 20 . The shape of the burr 310 includes at least one of a minimum value, a maximum value, and an average value of the thickness of the burr 310 in the extending direction of the processing tool 200, that is, the Z-axis direction. In the following description, it is assumed that the workpiece identification unit 19 acquires the maximum thickness of the burr 310 in the Z-axis direction.

When the learning unit 20 acquires the shape of the burr 310 from the workpiece specifying unit 19, it associates it with the machining trajectory error. Then, the learning unit 20 associates the machining trajectory error with the maximum thickness of the burr 310 in the Z-axis direction, and determines a target intermediate point that minimizes the machining trajectory error according to the driving conditions and the thickness of the burr 310. .

Upon acquiring the drive command including the drive conditions, the target trajectory determination unit 21 acquires the shape of the burr 310 from the workpiece identification unit 19 . Then, the target trajectory determining unit 21 acquires a target intermediate point according to the drive condition indicated by the drive command and the shape of the burr 310 from the learned data storage unit 15. Once the target intermediate point is acquired, the target trajectory determination unit 21 calculates a target trajectory passing through the target intermediate point based on the driving conditions.

The hardware configuration of the control device 2 is similar to the hardware configuration of the control device 1 shown in FIG. The control device 2 is connected via an interface 33 to motors M1, M2, M3, M4, M5, M6 and encoders E1, E2, E3, E4, E5, E6.

The learning process performed by the control device 2 having the above configuration will be explained using FIG. 13. Similar to the control device 1, for example, when the control device 2 starts operating for the first time after the robot 10 is installed, the control device 2 starts the learning process shown in FIG. 13. The learning process in FIG. 13 is performed while the machining tool 200 is attached to the flange 160 and machining the workpiece 300. In other words, the workpiece 300 is machined in order for the control device 2 to obtain the machining trajectory error for each drive condition.

The workpiece identification unit 19 acquires information about the workpiece 300 (step S31). For example, the workpiece identification unit 19 acquires the shape of the burr 310 on the workpiece 300. Specifically, the photographing device 191 includes a plurality of cameras provided around the workpiece 300, and generates image data of the workpiece 300 by photographing the workpiece 300 with the plurality of cameras. The shape determining unit 192 determines the shape of the burr 310 on the workpiece 300 from the image data. For example, a portion where the position of the outer surface changes rapidly can be regarded as the burr 310. The shape of the burr 310 on the workpiece 300 calculated by the shape determining section 192 is sent to the learning section 20. The processes from subsequent steps S11 to S12 are similar to the processes performed by the control device 1 shown in FIG. 5.

The control unit 17 controls the motors M1, M2, M3, M4, M5, and M6 of the robot 10 according to the reference trajectory determined in step S12, and the actual trajectory acquisition unit 13 controls the encoders E1, E2, E3, and E4. , E5, and E6, the trajectory of the robot 10 controlled according to the reference trajectory is obtained based on the rotational positions of the motors M1, M2, M3, M4, M5, and M6 obtained from the motors M1, M2, M3, M4, M5, and M6 (step S32). The learning unit 20 calculates a machining trajectory error indicating the deviation of the trajectory acquired in step S32 from the ideal trajectory based on the driving conditions acquired in step S11, and calculates a machining trajectory error and a machining trajectory error obtained in step S31. Information about the target object 300 is associated (step S33). The learning unit 20 generates an error table shown in FIG. 14 and stores it in the learned data storage unit 15. In the example of FIG. 14, the maximum thickness of the burr 310 is stored.

As long as the predetermined number of associations for each drive condition stored in the drive condition storage unit 11 has not been completed (step S34; No), the above-described process is repeated. The predetermined number of times is the number of times that data for generating a trajectory model can be acquired. Since the shape of the burr 310 on the workpiece 300 changes, by repeating the above process for each drive condition, data on the trajectory error during machining that changes according to the change in the burr 310 as shown in FIG. 14 is obtained. can do. For example, as a result of performing the above-described processing, as shown in FIG. 14, an error table showing the machining trajectory error for each thickness of the burr 310 is obtained for each drive condition.

When the predetermined number of correspondences are completed for each drive condition stored in the drive condition storage unit 11 (step S34; Yes), the learning unit 20 determines a target intermediate point that minimizes the machining trajectory error, and sets the drive condition A trajectory model indicating the target intermediate point is generated according to information about the workpiece 300 (step S35).

Specifically, the learning section 20 generates a trajectory model indicating a target midpoint during machining according to the thickness of the burr 310, and stores it in the learned data storage section 15. For example, the learning unit 20 performs supervised learning according to multidimensional function fitting or a neural network model to generate a trajectory model. Supervised learning means that by giving a large amount of input and result datasets to a learning device, the learning device learns the features in the large dataset and generates a model that estimates the result from the input. . For example, the learning unit 20 uses the driving conditions, the reference intermediate point, and the thickness of the burr 310 as input data, performs learning using the machining trajectory error as result data, and generates a trajectory model. Using the generated trajectory model, it is possible to obtain a target intermediate point that minimizes the trajectory error during machining.

In machine learning, it is known that estimation accuracy decreases due to overfitting, so it is preferable to suppress overfitting no matter which learning model is used. Examples of methods for suppressing overfitting include a method of normalizing learning data and a method of reducing the number of parameters in a learning model. For example, when performing multidimensional function fitting, it is preferable to limit the number of dimensions of the function, for example, to the fifth order or less. As another example, in a neural network model, it is preferable to limit the number of intermediate layers, for example, to 10 layers or less. Furthermore, in a neural network model, the output of a specific layer of the learning model may be randomly set to 0 during learning. Furthermore, overfitting may be suppressed by combining the above-described methods.

When the process of step S35 is completed, the control device 2 ends the learning process. The operation of the operation process performed after the above-described learning process is completed will be explained using FIG. 15. For example, when the control device 2 receives a drive command including drive conditions for the robot 10 through an operation from an operation unit (not shown) after the learning process is completed, the control device 2 starts the operation process shown in FIG. 15 .

Upon acquiring the drive command including the drive conditions, the target trajectory determination unit 16 acquires information about the workpiece 300, for example, the thickness of the burr 310 of the workpiece 300, from the workpiece identification unit 19 (step S24). . The target trajectory determination unit 16 acquires from the learned data storage unit 15 a target intermediate point according to the drive condition indicated by the drive command acquired in step S24 and the thickness of the burr 310 acquired in step S24 (step S25). If there is no burr 310 with the same thickness, interpolation processing may be performed based on data with similar burr 310 thickness values to obtain the target intermediate point.

The processing in steps S22 and S23 is similar to the processing performed by the control device 1 according to the first embodiment shown in FIG.

As explained above, the control device 2 according to the second embodiment learns the correspondence between the reference intermediate point, the trajectory error during machining, and the information about the workpiece 300, and the target intermediate point that minimizes the trajectory error during machining. is determined, and a trajectory model indicating the target intermediate point is generated according to the driving conditions and information about the workpiece 300. Processing of the workpiece 300 is performed by the control device 2 controlling the robot 10 according to the target trajectory passing through the target intermediate point. Since the target trajectory is adjusted according to information about the workpiece 300, for example, the thickness of the burr 310 on the workpiece 300, the actual trajectory during processing approaches the ideal trajectory, and the control device 2 controls the robot during processing. 10, the control accuracy becomes higher.

(Embodiment 3)
The control of the robot 10 in consideration of the reaction force during machining is not limited to the above-described embodiments, and the robot 10 may be controlled by a method that combines the above-described embodiments. A control device 3 that controls the robot 10 according to the reaction force during processing and information about the workpiece 300 will be described in a third embodiment, focusing on the differences from the first and second embodiments.

A robot system 102 according to the third embodiment shown in FIG. 16 includes a control device 3 and a robot 10. In addition to the configuration of the control device 1 according to the first embodiment, the control device 3 includes a workpiece identification unit 19 that acquires information about the workpiece 300. The control device 3 learns the correspondence between the machining trajectory error, the reaction force, and the information about the workpiece 300 acquired by the workpiece identification unit 19, and determines a target intermediate point that minimizes the machining trajectory error. , a learning unit 22 that generates a trajectory model indicating a target intermediate point according to the driving conditions and information about the workpiece 300. In the control device 3, the control section 17 acquires the reaction force from the reaction force acquisition section 18.

The hardware configuration of the control device 3 is similar to the hardware configuration of the control device 1 shown in FIG. Control device 3 is connected to motors M1, M2, M3, M4, M5, M6, encoders E1, E2, E3, E4, E5, E6, and force sensor 190 via interface 33.

The learning process performed by the control device 3 having the above configuration will be explained using FIG. 17. Similar to the control device 1, for example, when the control device 3 starts operating for the first time after the robot 10 is installed, the control device 3 starts the learning process shown in FIG. The process in step S31 is similar to the process performed by the control device 2 shown in FIG. 13. The processing from steps S11 to S13 is similar to the processing performed by the control device 1 shown in FIG.

However, in step S13, when the reaction force increases, the control unit 17 adjusts the position indicated by the reference trajectory to a position farther from the workpiece 300, and controls the motors M1, M2, M3, M4 according to the adjusted reference trajectory. , M5, and M6. This prevents the trajectory of the tip link from deviating significantly from the ideal trajectory due to a large reaction force.

As in the first embodiment, the learning unit 22 calculates a machining trajectory error indicating the difference between the ideal trajectory based on the driving conditions acquired in step S11 and the actual trajectory measured in step S14, and calculates the machining trajectory error. The error, the reaction force calculated in step S14, and the information about the workpiece 300 acquired in step S31 are associated (step S36). Specifically, the learning unit 22 performs machining when the distance between the ideal position of the tip link for each control cycle on the ideal trajectory and the actual position of the tip link for each control cycle indicated by the actual trajectory during machining is the maximum value. Correlate the time trajectory error and reaction force. The learning unit 22 generates an error table shown in FIG. 18, and stores it in the learned data storage unit 15. In the example of FIG. 18, the reaction force represented by (Fx, Fy, Fz) is stored in the error table. In the example of FIG. 18, the maximum thickness of the burr 310 is stored.

As shown in FIG. 17, the above-described process is repeated until the predetermined number of associations for each drive condition stored in the drive condition storage unit 11 has not been completed (step S34; No). The predetermined number of times is the number of times that data for generating a trajectory model can be acquired.

When the predetermined number of associations for each drive condition stored in the drive condition storage unit 11 is completed (step S34; Yes), the learning unit 20 sets the target intermediate point that minimizes the trajectory error during machining to the drive conditions and machining. A trajectory model is generated according to information about the target object 300 (step S37).

Specifically, the learning unit 22 generates a trajectory model indicating a target midpoint during machining according to the thickness of the burr 310, and stores it in the learned data storage unit 15. For example, the learning unit 22 performs supervised learning according to a neural network model to generate a trajectory model. Specifically, the learning unit 22 uses the driving conditions, reference midpoint, reaction force, and thickness of the burr 310 as input data, performs learning using the machining trajectory error as result data, and generates a trajectory model. Using the generated trajectory model, it is possible to determine the target intermediate point that minimizes the trajectory error during machining. To suppress overfitting, the trajectory model preferably has less than 10 hidden layers.

When the process of step S37 is completed, the control device 3 ends the learning process. The operation process performed after the above-described learning process is completed is the same as the process performed by the control device 2 shown in FIG. 15.

As explained above, the control device 3 according to the third embodiment learns the correspondence between the reference intermediate point, the machining trajectory error, the reaction force, and the information about the workpiece 300, and sets the goal of minimizing the machining trajectory error. The intermediate point is determined, and a trajectory model is generated that indicates a target intermediate point that minimizes the trajectory error during machining according to the driving conditions and information about the workpiece 300. Processing of the workpiece 300 is performed by the control device 3 controlling the robot 10 according to the target trajectory passing through the target intermediate point. Since the target trajectory is adjusted according to information about the workpiece 300, for example, the thickness of the burr 310 on the workpiece 300, the actual trajectory during processing approaches the ideal trajectory, and the control device 3 controls the robot during processing. 10, the control accuracy becomes higher.

The information about the workpiece 300 is not limited to the above-mentioned example, and may be any information about the workpiece 300 that causes fluctuations in the reaction force. As an example, the workpiece identification unit 19 may determine the minimum value, maximum value, or average value of the length of the burr 310 in the moving direction of the processing tool 200, in the above example, the X-axis direction.

(Embodiment 4)
The method of acquiring the trajectory of the robot 10 is not limited to the above example. A robot system 103 that includes a robot 10 and a control device 4 that acquires the trajectory of the robot 10 using a method different from that in the first embodiment will be described with a focus on the points that are different from the robot system 100 according to the first embodiment.

The control device 4 shown in FIG. 19 includes an actual trajectory acquisition unit 23 that acquires the trajectory of the robot 10 regardless of the rotational positions of the motors M1, M2, M3, M4, M5, and M6. The actual trajectory acquisition unit 23 has a measuring device, for example, a three-dimensional measuring device, for measuring the trajectory of the robot 10, and acquires the trajectory of the robot 10 according to the measurement value of the measuring device. The three-dimensional measuring device irradiates a laser beam onto the moving range of the tip link of the robot 10 and receives the laser beam reflected by the tip link of the robot 10, thereby obtaining the position of the tip link of the robot 10.

In the fourth embodiment, the actual trajectory acquisition unit 23 performs the above-mentioned processing for each control period, and calculates an actual trajectory that indicates the actual position of the tip link in the XYZ orthogonal coordinate system for each control period. . The actual trajectory acquisition unit 23 then sends the actual trajectory to the learning unit 14.

The hardware configuration of the control device 4 is similar to that of the control device 1 shown in FIG.
The learning process performed by the control device 4 having the above configuration is the same as that of the control device 1 except for the method of acquiring the trajectory of the tip link. The operation processing performed by the control device 4 is similar to that of the control device 1.

As explained above, the control device 4 according to the fourth embodiment acquires the locus of the robot 10 regardless of the rotational positions of the motors M1, M2, M3, M4, M5, and M6. Errors due to mechanical rigidity are not included, and the accuracy of the actual trajectory is improved.

The present disclosure is not limited to the embodiments described above.
The driving conditions are not limited to the above example. As an example, the driving conditions may further include information about the workpiece 300. Specifically, the driving conditions may include the operation start point, operation end point, speed, and posture of the tip link, and the weight of the workpiece 300. The learning unit 14 may acquire a measurement value from an external device, for example, a weight sensor that measures the weight of the workpiece 300. Specifically, the control device 1-4 acquires the measured value from the weight sensor, controls the robot 10 according to the drive conditions stored in the drive condition storage unit 11, and controls the robot 10 in accordance with the drive conditions and the measurement value of the weight sensor. Calculate the machining trajectory error, learn the correspondence between the reference intermediate point, the measured value of the weight sensor, and the machining trajectory error, find the target intermediate point that minimizes the machining trajectory error, and determine the driving conditions and workpiece. A trajectory model indicating the target intermediate point is generated according to the weight of the vehicle. As a result, as shown in FIG. 20, the learning unit 14 has a target that associates the drive conditions including the operation start point, operation end point, speed, and posture of the tip link, and the weight of the workpiece 300 with the target intermediate point. Generate waypoint table.

During operation, upon acquiring a drive command, the target trajectory determining unit 16 retrieves from the learned data storage unit 15 the operation start point, operation end point, speed, and posture of the tip link indicated by the drive command, as well as the measured values of the weight sensor. Obtain the target midpoint according to. Then, the target trajectory determination unit 16 determines a target trajectory passing through the target intermediate point based on the operation start point, operation end point, speed, and attitude of the tip link indicated by the drive command. The control accuracy of the robot 10 is improved by controlling the robot 10 based on the target trajectory according to the weight of the workpiece 300.

As another example, the driving conditions may include an operation start point, an operation end point, a time required from the operation start point to the operation end point, and a posture of the tip link.
In the embodiments described above, the driving conditions instruct linear movement on a plane, but the driving conditions may instruct three-dimensional movement, curved movement, etc.

In the embodiment described above, one reference midpoint is stored in one record of the driving condition table stored in the driving condition storage unit 11, but a plurality of standards are stored in one record of the driving condition table. Intermediate points may be stored. In this case, the reference trajectory determination unit 12 may determine a reference trajectory that passes through a plurality of reference intermediate points stored in the same record.

The driving condition storage section 11 and the learned data storage section 15 may be provided outside the control device 1-4. As an example, the driving condition storage section 11 and the learned data storage section 15 may be realized as functions of a storage device on a network.

The method of determining the reference trajectory, ideal trajectory, and target trajectory is not limited to the above-mentioned example. Specifically, the reference trajectory RT1 determined by the reference trajectory determination unit 12 in the first embodiment is a smooth curve from the operation start point SP1 to the operation end point EP1 through the reference intermediate point RP1, and is a smooth curve in the XY plane. Although shown as a curve that projects in the negative direction of the Y-axis, the reference trajectory RT1 is not limited to such a curve, and may be shown as a curve that projects in a plurality of directions. As an example, the reference trajectory RT1 protrudes in the Y-axis negative direction from the operation start point SP1 to the operation end point EP1, and protrudes in the Y-axis positive direction on the XY plane near the operation start point SP1 and the operation end point EP1. It may also be shown as a protruding curve.

As another example, the reference trajectory determination unit 12 calculates the reference trajectory by linearly interpolating the operation start point and reference intermediate point indicated by the driving condition, and by linearly interpolating the reference intermediate point and the operation end point indicated by the driving condition. You may. Even if the reference trajectory is calculated by linear interpolation, the control unit 17 controls the motors M1, M2, M3, M4, M5, and M6 to rotate smoothly, so that the tip link moves smoothly and the actual trajectory is smooth. It becomes a curve.

As another example, the reference trajectory determination unit 12 uses a table (not shown) in which the operation start point, reference intermediate point, and operation end point are associated with the reference position of the tip link for each control cycle to determine the reference trajectory. may be determined.

The ideal trajectory and the target trajectory can also be determined using a method similar to the method for determining the reference trajectory described above.

In the first embodiment, the ideal trajectory is a straight line because the robot 10 is moved linearly, but the ideal trajectory may be determined according to the way the robot 10 is moved and is not limited to a straight line. Specifically, the ideal trajectory may be a circular arc, a curved line, or a combination of a straight line, a circular arc, and a curved line.

In the first embodiment, the target trajectory TT1 determined by the target trajectory determination unit 16 is a smooth curve from the motion start point SP1 through the target intermediate point TP1 to the motion end point EP1, and is a smooth curve extending in the Y-axis negative direction in the XY plane. However, the target trajectory TT1 is not limited to such a curve. As an example, the target trajectory TT1 protrudes in the Y-axis negative direction from the motion start point SP1 to the motion end point EP1, and in the Y-axis positive direction on the XY plane near the motion start point SP1 and the motion end point EP1. It may also be indicated by a protruding curve.

As another example, the target trajectory determination unit 16 determines the target trajectory by linearly interpolating the operation start point and the target intermediate point indicated by the driving conditions, and by linearly interpolating the target intermediate point and the operation end point indicated by the driving conditions. You may. Even if the target trajectory is determined by linear interpolation, the control unit 17 controls the motors M1, M2, M3, M4, M5, and M6 to rotate smoothly, so that the tip link moves smoothly and the actual trajectory is smooth. It becomes a curve.

The method for determining the target midpoint is not limited to the above example. As an example, the learning unit 14 may determine the target intermediate point using linear regression, curve regression, polynomial regression, multidimensional function fitting, or the like. For example, the learning unit 14 may determine the target intermediate point by using the XYZ coordinate components of the reference intermediate point as independent variables and performing multidimensional function fitting using the machining trajectory error as an objective function.

The learning unit 14 calculates the distance between the ideal position, which is the position of the tip link in each control cycle indicated by the ideal trajectory, and the actual position, which is the position of the tip link in each control cycle, indicated by the actual trajectory, and determines whether the distance is a maximum value. The deviation of the actual position from the ideal position when the actual position becomes the ideal position may be used as the trajectory error. In this case, the learning unit 14 may store a plurality of machining trajectory errors in an error table, as shown in FIG. 21. Then, the learning unit 14 may calculate the target intermediate point based on the maximum value stored as the machining trajectory error. The same applies to the learning units 20 and 22.

The learning unit 14 may store the actual position of the tip link corresponding to the trajectory error in an error table. For example, when the learning unit 14 uses the maximum value of the distance between the ideal position of the tip link for each control cycle and the actual position of the tip link for each control cycle indicated by the actual trajectory as the machining trajectory error, As shown in , the actual position of the tip link when the distance reaches its maximum value may be stored in the error table. Then, the learning unit 14 may calculate the target intermediate point based on the maximum value stored as the machining trajectory error and the actual position of the tip link corresponding to the maximum value. The same applies to the learning units 20 and 22.

The learning unit 14 may store the reaction force when the distance reaches a maximum value in an error table, as shown in FIG. 23. Then, the learning unit 14 may calculate the target intermediate point based on the maximum value stored as the machining trajectory error and the reaction force corresponding to the maximum value.

The learning units 14, 20, and 22 may perform the learning process again in parallel with the operational process performed after the learning process is completed. Specifically, during the operation process, the learning units 14, 20, and 22 distinguish between an ideal trajectory corresponding to the drive conditions indicated by the drive command acquired by the target trajectory determination unit 16 and an actual trajectory acquired by the actual trajectory acquisition unit 13. The trajectory error indicating the difference between is calculated, and the trajectory error during machining is obtained. By performing the learning process in parallel with the operation process, the accuracy of the target intermediate point that minimizes the trajectory error is improved, and as a result, the control accuracy of the robot 10 is improved.

The learning units 14, 20, and 22 may determine the target intermediate point using the absolute value, average value, and median value of the deviation of the actual position of the tip link from the ideal position as the trajectory error.

The workpiece identification unit 19 may be any unit as long as it acquires information about the workpiece 300. As an example, the photographing device 191 may be a three-dimensional camera or a two-dimensional camera. As another example, the workpiece identification unit 19 may include a laser scanner that is attached to the robot 10 and measures the shape of the workpiece 300.

The method of determining the shape of the burr 310 by the workpiece specifying unit 19 is not limited to the above-mentioned example. As an example, the shape determining unit 192 included in the workpiece specifying unit 19 compares the image data generated by the photographing device 191 and the design data, and determines that a portion where the image data and the design data do not match is a burr 310. You can.

The learning process performed by the control device 1-4 may be a function of another device. As an example, FIG. 24 shows an example in which the functions of the control device 1 according to the first embodiment are realized by a control device 5 that controls the robot 10 and a learning device 6 that generates a trajectory model. The robot system 104 shown in FIG. 24 includes a robot 10, a control device 5 that controls the robot 10, and a learning device 6 that generates a trajectory model for controlling the robot 10.

The learning device 6 includes a drive condition storage section 11, a reference trajectory determination section 12, an actual trajectory acquisition section 13, a learning section 14, a learned data storage section 15, and a reaction force acquisition section 18. The functions of each part of the learning device 6 are similar to the functions of the corresponding parts of the control device 1. The control device 5 includes a target trajectory determining section 16 and a control section 17. The functions of each part of the control device 5 are similar to the functions of the corresponding parts of the control device 1. The target trajectory determining unit 16 included in the control device 5 may acquire the trajectory model from the learned data storage unit 15 included in the learning device 6 . The control unit 17 included in the control device 5 controls the robot 10 according to the reference trajectory acquired from the reference trajectory determination unit 12 included in the learning device 6 or the target trajectory calculated by the target trajectory determination unit 16. The learned data storage section 15 may be an independent storage device that can be accessed from the control device 5 and the learning device 6. Similarly, the functions of the control devices 2 and 3 can be realized by the control device 5 and the learning device 6.

As another example, in the robot system 104 shown in FIG. 24, among the functions of the learning unit 14 included in the learning device 6, the control device 5 may be provided with processing for generating a trajectory model. In this case, the learning device 6 may calculate the machining trajectory error for each reference intermediate point according to the driving conditions and store the machining trajectory error in the learned data storage unit 15. The control device 5 may acquire and hold the data learned in advance by the learning device 6, specifically, the machining trajectory error stored in the learned data storage section 15. Then, the control device 5 may generate a trajectory model that indicates a target intermediate point that minimizes the trajectory error during machining for each drive condition.

The trajectory of the robot 10 that is controlled by the control device 1-5 is not limited to the trajectory of the tip link, but is the trajectory of any part of the robot 10 or the processing tool 200 attached to the robot 10.

The method of acquiring the trajectory of the tip link by the actual trajectory acquisition unit 23 is not limited to the above-mentioned example. As an example, the actual trajectory acquisition unit 23 includes a plurality of cameras as measuring instruments. The actual trajectory acquisition unit 23 may photograph the moving range of the tip link of the robot 10 using a plurality of cameras, and determine the position of the tip link of the robot 10 from the images captured by the plurality of cameras.

As another example, the actual trajectory acquisition unit 23 includes an acceleration sensor attached to the tip link of the robot 10 as a measuring device. The actual trajectory acquisition unit 23 may determine the position of the tip link of the robot 10 based on the measured value of the acceleration sensor.

The actual trajectory acquisition unit 23 may measure the position of the tip link of the robot 10 at every acquisition period independent of the control period. In this case, the learning unit 14 may interpolate the position of the tip link of the robot 10 for each acquisition cycle, determine the position of the tip link of the robot 10 for each control cycle, and compare it with the ideal position. Specifically, the learning unit 14 performs interpolation processing using linear interpolation, polynomial interpolation, or the like.

The control unit 17 may perform feedback control of the motors M1, M2, M3, M4, M5, and M6 according to the position of the tip link of the robot 10 acquired by the actual trajectory acquisition unit 23 described above.

In addition, the above-mentioned hardware configuration and flowchart are merely examples, and can be changed and modified as desired. As an example, the control device 1-4 and the learning device 6 may start learning processing in response to an operation on an operation unit (not shown).

In the embodiment described above, the trajectory error is determined by actually operating the robot 10, but the learning process may also be performed by simulation.

A computer program for performing the above-mentioned operations is stored on a computer-readable recording medium such as a flexible disk, a CD-ROM (Compact Disc - Read Only Memory), or a DVD-ROM (Digital Versatile Disc - Read Only Memory). The control device 1-5 and learning device 6 that execute the above operations may be realized by distributing the computer program and installing the computer program on a computer. Alternatively, the control device 1-5 and the learning device 6 that execute the above operations may be realized by a dedicated system. The computer program may be superimposed on a carrier wave and provided via a communication network.

The robot 10 is not limited to the above-mentioned example, but is arbitrary. As an example, the control device 1-5 may control a robot 40 that is a horizontal articulated robot shown in FIG. The robot 40 includes a first arm 410, a second arm 420, a third arm 430, and a base 440. The first arm 410 is attached to the base 440 in a rotatable state around a rotation axis AX1 parallel to the Z-axis. The second arm 420 is connected to the first arm 410 and is rotatable around a rotation axis AX2 parallel to the Z-axis. The third arm 430 is connected to the second arm 420 and is rotatable around a rotation axis AX3 parallel to the Z-axis. Further, the third arm 430 is extendable and retractable in the direction along the rotation axis AX3. The base 440 is fixed to a plane and supports the first arm 410. It is possible to attach the processing tool 200 to the third arm 430. The control device 1-5 may control the position of the tip link, which is the tip of the third arm 430, by controlling a motor (not shown) included in the robot 40.

As another example, the control device 1-5 may control a robot 50, which is a robot having a plurality of axes as shown in FIG. 26. The robot 50 includes a first arm 510, a second arm 520, and a base 530. The first arm 510 is attached to the base 530 so as to be movable in the X-axis direction. The second arm 520 is connected to the first arm 510 and is movable in the Y-axis direction. The base 530 is fixed to a plane and supports the first arm 510. The processing tool 200 is attached to the second arm 520. In the example of FIG. 26, the second arm 520 is a print head, and forms electrodes on a substrate on a stage 540 placed on a base 530 using a processing tool 200 having a liquid ejection part and a ejection nozzle. Specifically, the tip of the processing tool 200 is movable on the stage 540 in the X-axis direction and the Y-axis direction. The control device 1-5 may control the position of the tip link, which is the tip of the second arm 520, by controlling a motor (not shown) included in the robot 50.

In the above example, the positions of the tip links of the robots 10, 40, and 50 were defined in the XYZ orthogonal coordinate system, but may be defined in the cylindrical coordinate system.

The present disclosure is capable of various embodiments and modifications without departing from the broad spirit and scope of the present disclosure. Moreover, the embodiments described above are for explaining this disclosure, and do not limit the scope of this disclosure. That is, the scope of the present disclosure is indicated by the claims rather than the embodiments. Various modifications made within the scope of the claims and the meaning of the disclosure equivalent thereto are considered to be within the scope of this disclosure.

This application is based on Japanese Patent Application No. 2021-23848, filed on February 18, 2021. The entire specification, claims, and drawings of Japanese Patent Application No. 2021-23848 are incorporated herein by reference.

1, 2, 3, 4, 5 control device, 6 learning device, 10, 40, 50 robot, 11 driving condition storage section, 12 reference trajectory determining section, 13, 23 actual trajectory acquisition section, 14, 20, 22 learning section , 15 learned data storage unit, 16, 21 target trajectory determination unit, 17 control unit, 18 reaction force acquisition unit, 19 workpiece identification unit, 31 processor, 32 memory, 33 interface, 34 bus, 100, 101, 102 , 103, 104 Robot system, 110, 410, 510 First arm, 120, 420, 520 Second arm, 130, 430 Third arm, 140 Fourth arm, 150 Fifth arm, 160 Flange, 160a Tip surface, 170 , 440, 530 base, 180 fixed part, 190 force sensor, 191 photographing device, 192 shape determining part, 200 processing tool, 300 workpiece, 310 burr, 540 stage, AT0, AT1, AT1' actual trajectory, AX1, AX2, AX3, AX4, AX5, AX6 Rotating axis, E1, E2, E3, E4, E5, E6 Encoder, EP1 Operation end point, ER1, ER2 Trajectory error, M1, M2, M3, M4, M5, M6 Motor, RP1 Reference intermediate point, RT1 reference trajectory, SP1 operation start point, TP1 target intermediate point, TT1 target trajectory.

Claims (21)

A control device that controls a robot to which a processing tool for processing a workpiece can be attached,
A drive in which a reference intermediate point indicating a position that the robot passes from the motion start point to the motion end point is stored in association with a drive condition that specifies at least a motion start point and a motion end point of the robot. a condition storage unit;
a reference trajectory determining unit that acquires the reference intermediate point from the driving condition storage unit and determines a reference trajectory passing through the reference intermediate point based on the driving condition for the acquired reference intermediate point;
an actual trajectory acquisition unit that acquires an actual trajectory of the robot;
When processing the workpiece by controlling the robot according to the reference trajectory, processing indicates a deviation of the actual trajectory during machining acquired by the actual trajectory acquisition unit from the ideal trajectory based on the driving conditions. The time trajectory error is determined for the acquired reference intermediate point, the correspondence between the reference intermediate point and the machining trajectory error is learned, and a target intermediate point that minimizes the machining trajectory error is determined according to the driving conditions. a learning unit that generates a trajectory model shown in
a target trajectory determining unit that determines a target trajectory passing through the target intermediate point obtained from the trajectory model;
a control unit that controls the robot according to the reference trajectory or the target trajectory;
A control device comprising: The actual trajectory acquisition unit includes a measuring device that measures the trajectory of the robot, and acquires the trajectory of the robot according to a measurement value of the measuring device.
The control device according to claim 1. A control device that controls a robot to which a processing tool for processing a workpiece can be attached,
Using a reference intermediate point that is associated with a drive condition that specifies at least a motion start point and a motion end point of the robot and indicates a position that the robot passes from the motion start point to the motion end point, Indicates the deviation of the actual trajectory during machining from the ideal trajectory based on the driving conditions, which indicates the actual trajectory of the robot controlled according to the reference trajectory passing through the reference intermediate point when processing the workpiece. A trajectory error during machining is determined, a correspondence between the reference intermediate point and the trajectory error during machining is learned, and a trajectory model is generated that indicates a target intermediate point that minimizes the trajectory error during machining according to the driving conditions. Learning department and
a target trajectory determining unit that determines a target trajectory passing through the target intermediate point obtained from the trajectory model;
a control unit that controls the robot according to the target trajectory;
A control device comprising: further comprising a reaction force acquisition unit that acquires a reaction force that is a force that the robot receives from the workpiece during processing with the processing tool,
The learning unit generates the trajectory model by learning the reference intermediate point, the machining trajectory error, and the response force when the machining trajectory error occurs.
The control device according to any one of claims 1 to 3. The control unit controls the robot according to the reference trajectory or the target trajectory and the reaction force acquired by the reaction force acquisition unit.
The control device according to claim 4. The control unit adjusts the position of the robot indicated by the reference trajectory or the target trajectory to a position farther from the workpiece according to the increase in the reaction force, and moves the robot according to the adjusted position of the robot. Control,
The control device according to claim 5. further comprising a workpiece identification unit that acquires information about the workpiece,
The learning unit learns the correspondence between the reference intermediate point, the machining trajectory error, and information about the workpiece, and determines the target intermediate point according to the driving condition and the information about the workpiece. generating the trajectory model shown in FIG.
The control device according to any one of claims 1 to 6. The workpiece identification unit is
an imaging device that generates image data of the processing object;
a shape determining unit that determines the shape of the workpiece from the image data generated by the photographing device;
acquiring the shape of the workpiece as information about the workpiece;
The control device according to claim 7. The workpiece identification unit includes a laser scanner that is attached to the robot and measures the shape of the workpiece,
acquiring the shape of the workpiece as information about the workpiece;
The control device according to claim 7. The processing by the processing tool is deburring the workpiece,
The workpiece identification unit acquires the shape of a burr included in the workpiece,
The control device according to any one of claims 7 to 9. The processing tool having a rod-like shape is attached to the robot,
The shape of the burr includes at least one of a minimum value, a maximum value, and an average value of the thickness of the burr in the stretching direction of the processing tool.
The control device according to claim 10. The shape of the burr includes at least one of a minimum value, a maximum value, and an average value of the length of the burr in the moving direction of the processing tool.
The control device according to claim 10. The machining trajectory error is defined as the ideal position of the actual position when the distance between the ideal position of the robot indicated by the ideal trajectory and the actual position of the robot indicated by the actual trajectory reaches a maximum value. indicates a deviation from the position,
A control device according to any one of claims 1 to 12. The local maximum value is the maximum value of the distance between the ideal position and the actual position,
The control device according to claim 13. A control device that controls a robot to which a processing tool for processing a workpiece can be attached,
an ideal trajectory based on the driving conditions of the actual trajectory of the robot when processing the workpiece by controlling the robot according to driving conditions that specify at least an operation start point and an operation end point of the robot; a target trajectory determining unit that obtains a trajectory model indicating a target intermediate point that minimizes a machining trajectory error indicating a deviation from the trajectory model, and determines a target trajectory passing through the target intermediate point obtained from the trajectory model;
a control unit that controls the robot according to the target trajectory;
A control device comprising: A non-machining trajectory that indicates a deviation of the actual non-machining trajectory of the tip link from the ideal trajectory, which is associated with the driving conditions of a robot equipped with a tip link to which a processing tool for machining the workpiece is attached. Regarding the reference intermediate point for minimizing the error, the machining trajectory error indicating the deviation of the actual machining trajectory from the ideal trajectory when machining the workpiece and the reaction force during machining are determined, and the When the correspondence between the intermediate point, the trajectory error during machining, and the reaction force during machining is learned, and a drive command including the drive conditions for the robot is input, the robot performs the machining process based on the reaction force. Generate a trajectory model that outputs a target waypoint that minimizes trajectory error,
Control device. A robot to which a processing tool can be attached for processing the workpiece,
The control device according to any one of claims 1 to 16 , which controls the robot according to drive conditions that specify at least an operation start point and an operation end point of the robot;
A robot system equipped with A learning device for learning a trajectory model for controlling a robot to which a processing tool for processing a workpiece can be attached,
A drive in which a reference intermediate point indicating a position that the robot passes from the motion start point to the motion end point is stored in association with a drive condition that specifies at least a motion start point and a motion end point of the robot. a condition storage unit;
a reference trajectory determining unit that acquires the reference intermediate point from the driving condition storage unit and determines a reference trajectory passing through the reference intermediate point based on the driving condition for the acquired reference intermediate point;
an actual trajectory acquisition unit that acquires an actual trajectory of the robot;
When processing the workpiece by controlling the robot according to the reference trajectory, processing indicates a deviation of the actual trajectory during machining acquired by the actual trajectory acquisition unit from the ideal trajectory based on the driving conditions. The time trajectory error is determined for the acquired reference intermediate point, the correspondence between the reference intermediate point and the machining trajectory error is learned, and a target intermediate point that minimizes the machining trajectory error is determined according to the driving conditions. a learning unit that generates a trajectory model shown in
A learning device equipped with. A non-machining trajectory that indicates the deviation of the actual non-machining trajectory of the tip link from the ideal trajectory, which is associated with the driving conditions of a robot equipped with a tip link to which a processing tool for machining the workpiece is attached. Regarding the reference intermediate point for minimizing the error, the machining trajectory error indicating the deviation of the actual machining trajectory from the ideal trajectory when machining the workpiece and the reaction force during machining are determined, and the machining reaction force is determined. Obtained by learning the correspondence between the intermediate point, the trajectory error during machining, and the reaction force during machining,
When a drive command including the drive conditions of the robot is input, a target intermediate point that minimizes the machining trajectory error is output based on the reaction force;
A trajectory model for making computers work. A control method for controlling a robot to which a processing tool for processing a workpiece can be attached, the method comprising:
Based on drive conditions that specify at least a motion start point and a motion end point of the robot, a reference that is associated with the drive condition and indicates a position that the robot passes from the motion start point to the motion end point. Determine the reference trajectory passing through the intermediate point,
When machining the workpiece by controlling the robot according to the reference trajectory, a machining trajectory error indicating a deviation of the actual machining trajectory of the robot from the ideal trajectory based on the driving conditions is set as the reference. each of the intermediate points, learns the correspondence between the reference intermediate point and the machining trajectory error, and generates a trajectory model that indicates a target intermediate point that minimizes the machining trajectory error according to the driving condition. ,
determining a target trajectory passing through the target intermediate point obtained from the trajectory model;
controlling the robot according to the reference trajectory or the target trajectory;
Control method. A computer that controls a robot to which a processing tool can be attached to process the workpiece.
A drive in which a reference intermediate point indicating a position that the robot passes from the motion start point to the motion end point is stored in association with a drive condition that specifies at least a motion start point and a motion end point of the robot. condition storage unit,
a reference trajectory determining unit that acquires the reference intermediate point from the driving condition storage unit and determines a reference trajectory passing through the reference intermediate point based on the driving condition for the acquired reference intermediate point;
an actual trajectory acquisition unit that acquires the actual trajectory of the robot;
When processing the workpiece by controlling the robot according to the reference trajectory, processing indicates a deviation of the actual trajectory during machining acquired by the actual trajectory acquisition unit from the ideal trajectory based on the driving conditions. The time trajectory error is determined for the acquired reference intermediate point, the correspondence between the reference intermediate point and the machining trajectory error is learned, and a target intermediate point that minimizes the machining trajectory error is determined according to the driving conditions. a learning unit that generates a trajectory model shown in
a target trajectory determination unit that determines a target trajectory passing through the target intermediate point obtained from the trajectory model; and a control unit that controls the robot according to the reference trajectory or the target trajectory.
A program to function as

Images