Robot control method and control device

Abstract

A control method and a control device are provided which suppress generation of vibration regardless of a change in the mass of a workpiece held by a robot, and perform stable control when controlling a multi-joint robot. A control device (10) which controls a multi-joint robot whose joint shaft is driven by a motor (50) has a state feedback control section (11) which performs state feedback control based on a motor position command, a single inertia system control section (12) which performs control based on a single inertia system model based on the motor position command, and a selection section (13) which selects the state feedback control section (11) when the mass of a workpiece held by the robot is below a threshold value, and selects the single inertia system control section (12) otherwise. The motor (50) is controlled by the output of either one of the control sections (11, 12) selected by the selection section (13) in accordance with the mass of the workpiece.

Description

Technical Field

[0001] This invention relates to a control method and control device for robots, and particularly to a control method and control device for controlling the position and speed of a multi-joint robot whose axes are driven by motors. Background Technology

[0002] Multi-joint robots (hereinafter referred to as robots) are horizontally articulated robots with multiple links or arms connected by joints. They are driven by motors located at the joints, for example, to hold or place a workpiece at their front end. The situation where a robot holds or places a workpiece is collectively referred to as "robot holding a workpiece." In large robots, the characteristics of the robot as the controlled object change depending on the mass of the workpiece held at its front end (i.e., fingertips) and the change in the inertial torque of the links about the joint axis. If the nominal control model is fixed without relying on the mass of the workpiece, model mismatch will occur due to changes in the mass of the workpiece, and feedback-based control will become unstable, potentially causing vibration. In particular, in robots, motors are usually connected to the joint axis via reducers, which also have the properties of spring elements. Therefore, under inappropriate control, the spring elements, including reducers, can easily induce vibrations in the robot due to their movement. There is a need for a control method and control device that can stably control the position and speed of the robot regardless of changes in the mass of the workpiece held by the robot.

[0003] Patent Document 1 discloses a positioning control device used in robot control, etc., which includes a classic control unit for positioning control based on P-PI (proportional-proportional-integral) control and a modern control unit for positioning control based on state feedback control using a two-inertial-system state observer. P-PI control (or PID (proportional-integral-derivative) control) is implemented when the inertial torque about the joint axis is less than a threshold, and state feedback control is implemented when it is above the threshold, thereby suppressing vibration and achieving high-speed and high-precision positioning control. Furthermore, Non-Patent Document 1 discloses an RPP-type robot that uses PID control under normal conditions and state feedback control to suppress fingertip vibration when the arm is extended. Patent Document 2 discloses a method for controlling a robot using state feedback control via a two-inertial-system state observer, where the physical model, i.e., the control parameters, of the state observer is switched according to the magnitude of the load on the robot's front end. Patent Document 3 discloses one specific example of state feedback control based on a two-inertial-system state observer. [Existing technical documents] [Patent Literature]

[0004] Patent Document 1: Japanese Patent Application Publication No. 7-319546 Patent Document 2: Japanese Patent No. 6097174 Patent Document 3: International Publication No. 2018 / 061097 [Non-patent literature]

[0005] Non-Patent Literature 1: Yoshiaki Ito, "Vibration Reduction Control for Large Robots", Toyota Central Research Institute R&D Review, March 1995, Vol. 30, No. 1, p. 57 Summary of the Invention

[0006] In the method described in Patent Document 1, P-PI control and state feedback control are switched based on the magnitude of the inertial torque around the joint axis. When this method is applied to robot control, if the mass of the workpiece held at the front end is large when viewed from the joint axis of the controlled object, state feedback control is selected as the robot control method. According to the inventor's research, when the method of Patent Document 1 is applied to multi-joint robots, especially horizontal multi-joint robots, sufficient vibration reduction cannot be obtained when the mass of the workpiece held by the robot is large.

[0007] The purpose of this invention is to provide a control method and control device for controlling a multi-joint robot, which can suppress the generation of vibration and achieve stable control regardless of changes in the mass of the workpiece held by the robot.

[0008] The inventors' research has shown that in multi-joint robots, when the mass of the workpiece held by the robot's fingertips is large, control based on a single inertial system model can suppress robot vibrations and achieve stable robot control compared to controlling the robot's joint axes using motors via state feedback control. Conversely, when the workpiece mass is small, greater vibration reduction performance can be achieved by implementing state feedback control. Therefore, the control method of this invention is a control method for multi-joint robots whose joint axes are driven by motors. For at least one motor, control is performed by switching based on the mass according to instructions, so that state feedback control is implemented when the mass of the workpiece held by the multi-joint robot is below a threshold, and control based on a single inertial system model is implemented when the mass exceeds the threshold.

[0009] The state feedback control that can be used in the control device of the present invention is, for example, state feedback control based on a two-inertial system state observer, as exemplified by the state feedback control described in Patent Document 3. Furthermore, the control method of the present invention that can be used is, for example, P-PI control, or PID control. The command is, for example, a motor position command.

[0010] In the control method of this invention, the motion of a multi-joint robot can be controlled according to a motion program that at least specifies a loading action for holding a workpiece and an unloading action for removing the held workpiece from the multi-joint robot. Switching between state feedback control and control based on a single inertial system model is performed based on the motion program. Robots used for workpiece transport, etc., typically control their motion based on motion programs, which at least specify loading and unloading actions. If the motion program specifies how to transport the workpiece, it can be assumed that the mass of the workpiece is also known when the motion program is executed. Therefore, by switching between state feedback control and control based on a single inertial system model according to the motion program, the control switching can be performed with less processing power.

[0011] In the control method of the present invention, it is preferable to implement state feedback control when the multi-joint robot, which is not holding the workpiece, moves to the receiving position of the workpiece according to the motion program, and then, if the mass exceeds a threshold, switch to control based on a single inertial system model before the workpiece contacts the multi-joint robot. By controlling in this way, vibrations caused by holding a heavy workpiece while in state feedback control can be prevented.

[0012] In the control method of this invention, when controlling the robot through an action program, the multi-joint robot can be moved to the workpiece loading / unloading position while holding the workpiece, according to the action program. After performing the unloading action, it switches to state feedback control. By implementing state feedback control when not holding the workpiece, the robot can be moved at high speed and stably.

[0013] In the control method of this invention, state feedback control and control based on a single inertial system model can be switched by measuring or estimating the mass of the workpiece. For example, in the case of a multi-joint robot equipped with a motor on a vertical axis, the mass of the workpiece can be determined based on the torque command value for the motor on the vertical axis. Since the mass of each part of the robot's mechanism is known, the mass corresponding to the torque command value for the motor on the vertical axis can be calculated, and the mass of the workpiece can be obtained by subtracting the mass of each part on the robot side from this mass. Alternatively, in the case of a multi-joint robot equipped with a load sensor that detects whether a workpiece meeting specified conditions is being held, control based on a single inertial system model can be switched when the load sensor detects that a workpiece meeting specified conditions is being held. The load sensor can also be a sensor that detects mass. In addition, since a workpiece with large thickness and planar dimensions can be presumed to be a workpiece with large mass, the load sensor can also be a sensor that detects thickness and planar dimensions. By determining or estimating the mass of the workpiece based on the torque command value of the motor on the vertical axis or the load sensor, even if a program error occurs when switching control according to the motion program, it is possible to prevent the robot from being controlled by state feedback control when holding a heavy workpiece.

[0014] In the control method of the present invention, the multi-jointed robot is, for example, a horizontal multi-jointed robot. Horizontal multi-jointed robots are often used to hold a workpiece on their fingertips and transport the workpiece. By applying the control method of the present invention to such a horizontal multi-jointed robot, the effects of the present invention, such as suppressing the generation of vibrations in the robot and performing stable control, can be further realized.

[0015] The control device of the present invention is a control device for controlling a multi-joint robot whose joint axes are driven by motors, comprising: a first control unit that performs state feedback control based on instructions; a second control unit that performs control based on a single inertial system model based on instructions; and a selection unit that selects the first control unit when the mass of the workpiece held by the multi-joint robot is below a threshold, and selects the second control unit when the mass exceeds the threshold, and controls the motor by outputting the control unit selected by the selection unit according to the mass.

[0016] According to the control device of the present invention, since the robot is moved by state feedback control when the workpiece mass is small, and by control based on a single inertial system model when the workpiece mass is large, the generation of vibration can be suppressed and the robot can be stably controlled regardless of the workpiece mass. The state feedback control performed by the first control unit is, for example, state feedback control based on a two-inertial system state observer; as an example, it is the state feedback control described in Patent Document 3. The control based on a single inertial model system performed by the second control unit is, for example, P-PI control or PID control. The command is, for example, a motor position command.

[0017] Multi-joint robots with workpiece holding capabilities are typically controlled according to motion programs. These motion programs at least specify loading actions to hold the workpiece and unloading actions to remove the held workpiece from the multi-joint robot. Therefore, in the control device of the present invention, it is preferable to provide a computation unit that executes such a motion program, and a selection unit selects either a first control unit or a second control unit based on a switching instruction from the computation unit. With this configuration, the switching between state feedback control based on the motion program and control based on a single inertial system model can be performed with less processing power.

[0018] In the control device of the present invention, it is preferable that, when an arithmetic unit is provided, the arithmetic unit selects a first control unit via a selection unit when moving the joint robot that is not holding the workpiece to the receiving position of the workpiece, and then, if the mass exceeds a threshold, selects a second control unit via the selection unit before the workpiece contacts the joint robot. By selecting the control unit in this way, vibration caused by holding a heavy workpiece during state feedback control can be prevented.

[0019] In the control device of the present invention, if a calculation unit is provided, the calculation unit can move the multi-joint robot to the workpiece loading / unloading position while holding the workpiece. After the unloading action is performed, a first control unit is selected by a selection unit. By selecting the first control unit in this way, state feedback control is implemented when the workpiece is not held, enabling the robot to move at high speed and stably.

[0020] In the control device of the present invention, the switching between the first control unit and the second control unit can be performed by measuring or estimating the mass of the workpiece. For example, when the multi-joint robot is equipped with a motor for the vertical axis, the arithmetic unit can also determine the mass of the workpiece based on the torque command value for the motor for the vertical axis. Alternatively, if the multi-joint robot is equipped with a load sensor that detects whether a workpiece meeting the specified conditions is being held, the second control unit can be selected when the load sensor detects that a workpiece meeting the specified conditions is being held. By measuring or estimating the mass of the workpiece based on the torque command value of the motor for the vertical axis or the load sensor, even if a program error occurs during the switching of control according to the motion program, it is possible to prevent the robot from being controlled by state feedback control when holding a heavy workpiece.

[0021] In the control device of the present invention, the multi-jointed robot that is controlled is, for example, a horizontal multi-jointed robot. Horizontal multi-jointed robots are often used to hold a workpiece on their fingertips and transport the workpiece. By using the control device of the present invention in the control of a horizontal multi-jointed robot, the effects of the present invention, such as suppressing the generation of vibrations in the robot and performing stable control, can be further realized.

[0022] According to the present invention, when controlling a multi-joint robot, it is possible to suppress the generation of vibrations regardless of changes in the mass of the workpiece held by the robot, thereby achieving stable control. Attached Figure Description

[0023] Figure 1 This is a block diagram illustrating the structure of a control device according to one embodiment of the present invention. Figure 2 The figure shows an example of the structure of a horizontal multi-joint robot, (a) is a top view, and (b) is a front view. Figure 3 It is a waveform diagram illustrating the vibration of the robot's front end when state feedback control and P-PI control are implemented. Figure 4 (a) and (b) are pole and zero configuration diagrams when state feedback control is implemented. Figure 5 (a) and (b) are pole-zero configuration diagrams when P-PI control is implemented. Figure 6 This is a block diagram illustrating the structure of a control device according to another embodiment. Detailed Implementation

[0024] The embodiments of the present invention will now be described with reference to the accompanying drawings. Figure 1 A control device according to one embodiment of the present invention is shown.

[0025] Figure 1 The control device 10 shown is used for drive control of a multi-joint robot, which has a motor 50 for each joint axis. A link 60 is mechanically connected to the motor 50, which is driven by the motor 50 and applies a reaction force or interference to the motor 50. Figure 1 The illustrated link 60 not only shows the link or arm directly connected to the motor 50 of interest, but also abstractly represents all parts of the robot that have a mechanical impact on the motor 50 of interest. Therefore, Figure 1 The linkage 60 shown includes, in addition to the arm and other components in the robot, a reducer connected to the motor 50 and a workpiece 70 held on the robot (see reference). Figure 2 Motor 50 is a motor with an encoder, and its position is fed back to the control device 10. Figure 1 The diagram only shows the portion used in the control of one joint axis in a multi-joint robot, but in reality, mechanisms are set up for controlling the motors 50 of each joint axis.

[0026] The control device 10 controls the position and speed of the motor 50 according to instructions, such as motor position instructions. To control the position and speed of the motor 50, the control device 10 includes: a state feedback control unit 11, which performs state feedback control of the motor 50 according to the motor position instructions; a single inertial system control unit 12, which performs control based on a single inertial system model of the motor 50 according to the motor position instructions; and a selection unit 13, which selects either the output of the state feedback control unit 11 or the output of the single inertial system control unit 12. The state feedback control unit 11 receives feedback on the motor position and performs state feedback control based on a two-inertial system state observer, such as the state feedback control described in Patent Document 3. The single inertial system control unit 12 also receives feedback on the motor position, for example, performing P-PI control or PID control. The selection unit 13 selects the state feedback control unit 11 when the mass of the workpiece 70 held by the multi-joint robot is below a threshold, and selects the single inertial system control unit 12 when the mass of the workpiece 70 exceeds the threshold. The motor 50 is driven by a motor current based on the output selected by the selection unit 13.

[0027] Figure 2 Shown by Figure 1 This is an example of a multi-joint robot controlled by the control device 10 shown. The multi-joint robot shown here is a horizontal multi-joint robot used for transporting workpiece 70. Figure 2 In the image, (a) is a top view of the horizontal multi-joint robot, and (b) is a front view of the horizontal multi-joint robot. Figure 2 (b) also depicts a control device 10 connected to the robot. The illustrated robot includes: a first link 60A, one end of which is mounted on a base 55 and extends horizontally; a second link 60B, one end of which is mounted on the other end of the first link 60A and extends horizontally; and a hand 61, mounted on the other end of the second link 60B and serving as the robot's end effector. A workpiece 70, which is the object to be transported, is placed on the hand 61. In the hand 61, the front end opposite to the connection position of the second link 60B becomes the fingertip position of the robot.

[0028] The base 55 is connected to the first link 60A at joint A. A motor 50A and a reducer 51A are installed within the base 55 to drive joint A. The first link 60A is driven by the motor 50A via the reducer 51A, causing it to rotate in the horizontal plane around joint A. Similarly, the first link 60A is connected to the second link 60B at joint B. A motor 50B and a reducer 51B are installed within the first link 60A to drive joint B. The second link 60B is driven by the motor 50B via the reducer 51B, causing it to rotate in the horizontal plane around joint B. The second link 60B is connected to the hand 61 at joint C. A motor 50C and a reducer 51C are installed within the second link 60B to drive joint C. The hand 61 is driven by the motor 50C via the reducer 51C, causing it to rotate in the horizontal plane around joint C. As a result, in this robot, by driving motors 50A, 50B, and 50C, the hand 61 can be moved horizontally in the direction of travel, as shown by the arrow in the figure, while maintaining its orientation, to move it closer to or away from the base 55. The direction in the horizontal plane, i.e., the direction orthogonal to the direction of travel, is defined as the lateral direction. In the robot, the portion from joint A to the hand 61 is a horizontal multi-joint mechanism. Furthermore, to enable the robot to move up and down in the vertical direction of the hand 61, a lifting mechanism 56 is provided inside the base 55. The lifting mechanism 56 includes a ball screw, etc., and is driven by a vertical shaft motor 50Z via a reducer 51Z, causing the robot's horizontal multi-joint mechanism to move up and down.

[0029] Next, use Figure 3 The switching between state feedback control and control based on a single inertial system model in this embodiment is explained. Figure 2 In the robot shown, consider driving motors 50A, 50B, and 50C to move the hand 61 between a state where the first link 60A and the second link 60B are fully folded and a state where the first link 60A and the second link 60B are extended to their maximum length in a straight line. Figure 2 The indicated direction of travel is repeated. Figure 3 In the diagram, waveform (a) shows the time variation of position commands related to the position of the robot's fingertips. Furthermore, the lateral vibration velocity at the fingertips was investigated when the hand 61 was reciprocated. Workpieces 70 held on the hand 61 were used, with a mass of 2 kg (corresponding to "workpiece mass: small") and a mass of 80 kg (corresponding to "workpiece mass: large"). Additionally, the lateral vibration velocity was investigated in both cases: using state feedback control and using P-PI control in the control of the position command-based motors 50A, 50B, and 50C. The results are shown in… Figure 3 (b)~(e).

[0030] When using state feedback control in the control of motors 50A, 50B, and 50C, if the mass of workpiece 70 is small, then as follows: Figure 3 As shown in (b), a good vibration reduction effect can be obtained, but if the mass of workpiece 70 is large, then as... Figure 3 As shown in (c), the vibration does not converge but diverges. That is, it can be seen that when the mass of workpiece 70 is large, state feedback control will make the control unstable. On the other hand, when P-PI control is used in the control of motors 50A, 50B, and 50C, when the mass of workpiece 70 is small ( Figure 3 (d) and cases with large mass ( Figure 3 Vibration reduction was achieved in any of the cases (e). However, when the mass of the workpiece was small (70%), the state feedback control showed a better vibration reduction effect than the P-PI control.

[0031] In this way, using Figure 4 and Figure 5 This describes how the control behavior changes depending on the quality of the workpiece. Figure 4 and Figure 5 Show Figure 3 The pole-zero configurations for each case shown in (b) to (e) are from the position command of motor 50A of joint A to the angle of joint A. In the figures, the poles enclosed by the quadrilateral with a single-dot dash are representative poles. Figure 4 The diagram shows the use of state feedback control in motor control. (a) is the pole and zero point configuration diagram when the workpiece mass is small, and (b) is the pole and zero point configuration diagram when the workpiece mass is large. Figure 5 The diagram shows the use of P-PI control in motor control. (a) is the pole and zero point configuration diagram when the workpiece mass is small, and (b) is the pole and zero point configuration diagram when the workpiece mass is large.

[0032] In the case of multi-joint robots, inter-axis disturbance forces are generated in the torque between joints. With a large workpiece mass, the amount of inter-axis disturbance increases, making it impossible to approximate each axis using a two-inertial system. Single-inertial system control is more robust than state feedback control. For representative values ​​from the motor position commands of each axis to the joint angles of each axis, while keeping the workpiece light, a comparison is made... Figure 4 (a) and Figure 5 (a) It can be seen that, compared with control based on a single inertial system model (P-PI control), state feedback control moves further away from the imaginary axis in the negative direction. Therefore, state feedback control has higher vibration reduction performance. Conversely, when maintaining a heavy workpiece, the comparison... Figure 4 (b) and Figure 5 (b) It can be seen that the real part representing the pole is negative in control based on the single-inertial system model, but positive in state feedback control, which is in the so-called unstable region. Therefore, control based on the single-inertial system model is more stable.

[0033] Therefore, in this embodiment, by switching the motor control mode to state feedback control when the workpiece 70 has a small mass and to control based on a single inertial system model when the mass is large, vibration reduction performance when handling light workpieces and stability under heavy workpiece loads are ensured. The instability of control when holding a heavy workpiece in state feedback control becomes significant when the spring component (such as the reducer) in the system increases and the distance from the joint of interest to the workpiece increases. Therefore, in Figure 2 In the robot shown, state feedback control corresponding to the mass of the workpiece 70 and control based on a single inertial system model can be switched only in the motors that are far from the holding position of the workpiece 70, such as motor 50A of joint A, or only in motors 50A and 50B of joints A and B.

[0034] In the embodiments described above, by switching between state feedback control and control based on a single inertial system model according to the mass of the workpiece held in the robot, the generation of vibration can be suppressed regardless of changes in the mass of the workpiece held by the robot, thus achieving stable control.

[0035] Figure 6 A control device 10 according to another embodiment of the present invention is shown. Generally, a multi-joint robot controls its actions by executing an action program. The action program of the robot that transports or holds the workpiece 70 at least specifies a loading action to hold the workpiece 70 on, for example, the hand 61 of the multi-joint robot and an unloading action to unload the held workpiece from the multi-joint robot, and further specifies the robot's movement path, etc. When transporting the workpiece 70 between user devices, in the loading action, the robot 10, which is not holding the workpiece 70, moves to the receiving position of the workpiece 70 in the user device of the transport source. When the workpiece 70 is actually held on the hand 61 of the robot 10 in the loading action, the workpiece 70 can be placed on the hand 61 by, for example, by the movement of the robot 10 in the vertical axis direction, or by the movement on the user device side. When using a gripping hand, the workpiece can be held by hand. Similarly, in the unloading action, the robot 10 holding the workpiece 70 moves to the loading / unloading position of the workpiece 70 in the user device of the transport destination. During the unloading action, when the workpiece 70 is actually unloaded from the hand 61 of the robot 10, the workpiece 70 can be unloaded to the user device side by moving along the vertical axis of the robot 10, or the workpiece 70 can be unloaded by moving along the user device side. When using a gripping hand, the hand can also release the workpiece it is gripping.

[0036] Figure 6 The control device 10 shown is Figure 1The control device 10 shown is identical, but it also includes an arithmetic unit 14 that generates motor position commands by executing an action program, and a storage unit 15 that stores the action program. The action program is generally generated as a program that knows how to process a workpiece, such as moving the workpiece from one location to another, so the quality of the workpiece can be assumed to be known during the execution of the action program. Therefore, if the action program specifies whether to control the motor through state feedback control or control based on a single inertial system model, the arithmetic unit 14 can output a switching command to the selection unit 13 when the action program is executed. The selection unit 13 can then switch between the output of the state feedback control unit 11 and the output of the single inertial system control unit 12 according to the switching command.

[0037] When controlling the robot via the motion program executed by the arithmetic unit 14, the state feedback control unit 11 is selected when the robot moves to the receiving position of the workpiece 70 without holding it. Then, when holding a heavy workpiece (a workpiece exceeding a threshold mass), the output of the single inertial system control unit 12 can be selected via the selection unit 13 before the workpiece 70 contacts the hand 61. This control prevents vibration caused by holding a heavy workpiece 60 while in state feedback control. Furthermore, when controlling the robot via the motion program executed by the arithmetic unit 14, after moving the robot holding the workpiece 70 to the loading / unloading position of the workpiece 70 and performing the unloading action, the output of the state feedback control unit 11 can be selected via the selection unit 13. This control, implementing state feedback control when not holding the workpiece 70, enables high-speed and stable robot movement.

[0038] When the calculation unit 14 outputs a switching command to the selection unit 13 based on the action program, due to a program error in the action program, it is possible that state feedback control will be executed while holding a heavy workpiece 70. Furthermore, in manual operation of the robot that does not depend on the action program, it is preferable to automatically switch between state feedback control and control based on a single inertial system model. Therefore, in the control device 10 of this embodiment, it is also possible to configure it to switch between state feedback control and control based on a single inertial system model by measuring or estimating the mass of the workpiece.

[0039] like Figure 2As shown, in the control device 10 that controls the robot with a vertical axis motor 50Z, a vertical axis control unit 16 is provided to control the vertical axis motor 50Z. Even when the robot is not moving up or down, the vertical axis motor 50Z generates a holding torque to maintain the mass of the robot and the workpiece 70. The vertical axis control unit 16 generates a torque command value corresponding to the holding torque and drives the motor 50Z with a motor current corresponding to the torque command value. Since the mass of each part of the robot's mechanism is known, the calculation unit 14 calculates the mass corresponding to the torque command value for the vertical axis motor. By subtracting the mass of each part on the robot side from this mass, the mass of the workpiece 70 can be calculated, and a switching command can be output based on the mass of the workpiece 70.

[0040] In robots, a load sensor 62 is sometimes installed on the handpiece 61 to detect whether a workpiece 70 meeting specified conditions is being held, and the detection result of the load sensor 62 is sent to the calculation unit 14. When such a load sensor 62 is installed, when a workpiece meeting specified conditions is detected by the load sensor 62, a switching command can be used to cause the selection unit 13 to select the output of the single inertial system control unit 12. The load sensor can be a sensor that detects mass, or it can be a sensor that detects workpieces 70 with a thickness or planar dimension greater than a specified value. Since it is possible to infer that workpieces with large thickness and planar dimensions are heavy workpieces, by switching between state feedback control and control based on the single inertial system model according to the detection result of the load sensor 62, it is possible to prevent the robot from being controlled by state feedback control when holding heavy workpieces.

[0041] By using Figure 6 The control device 10 shown can switch between state feedback control and control based on a single inertial system model through an action program, thus enabling control switching with less processing power. Symbol Explanation

[0042] 10…Control device; 11…Status feedback control unit; 12…Single inertial system control unit; 13…Selection unit; 14…Calculation unit; 15…Storage unit; 50, 50A, 50B, 50C, 50Z…Motors; 51A, 51B, 51C, 51Z…Reducers; 55…Base; 56…Lifting mechanism; 60, 60A, 60B…Linking rod; 61…Hand; 62…Load sensor; 70…Workpiece.

Claims

1. A control method for controlling a horizontal multi-joint robot whose joint axes are driven by motors, characterized in that, For at least one of the motors, control is performed based on motor position commands and the mass of the workpiece held by the end effector of the horizontal joint robot, such that a first control is implemented when the mass of the workpiece is below a threshold, and a second control is implemented when the mass of the workpiece exceeds the threshold. The first control is a state feedback control based on a 2-inertial system state observer, using the motor position command and the feedback motor position. The second control is P-PI control or PID control using the motor position command and the feedback motor position.

2. The control method as described in claim 1, characterized in that, The movement of the horizontal joint robot is controlled based on a motion procedure that at least specifies the loading action of holding the workpiece by the horizontal joint robot and the unloading action of unloading the workpiece held by the horizontal joint robot. The switching between the first control and the second control is performed according to the action procedure.

3. The control method as described in claim 2, characterized in that, When the horizontal joint robot, which is not holding the workpiece, is moved to the receiving position of the workpiece based on the motion program, the first control is implemented. Then, when the mass of the workpiece exceeds the threshold, the second control is switched before the workpiece and the horizontal joint robot come into contact.

4. The control method as described in claim 2 or 3, characterized in that, Based on the action program, while holding the workpiece, the horizontal multi-joint robot is moved to the loading and unloading position of the workpiece. After performing the unloading action, the first control is switched.

5. The control method according to any one of claims 1 to 3, characterized in that, The horizontal multi-joint robot is equipped with a motor with a vertical axis. The quality of the workpiece is determined based on the torque command value of the motor for the vertical shaft.

6. The control method according to any one of claims 1 to 3, characterized in that, The horizontal multi-joint robot is equipped with a load sensor to detect whether the workpiece has maintained a state that meets specified conditions. When the load sensor detects that the workpiece meets the specified conditions, the system switches to the second control.

7. A control device for controlling a horizontal multi-joint robot whose joint axes are driven by motors, characterized in that, include: A first control unit, which performs state feedback control based on a 2-inertial system state observer using the motor position command and the feedback motor position, based on the motor position command. The second control unit performs P-PI control or PID control using the motor position command and the feedback motor position based on the motor position command. as well as A selection unit selects the first control unit when the mass of the workpiece held by the end effector of the horizontal multi-joint robot is below a threshold, and selects the second control unit when the mass of the workpiece exceeds the threshold. The motor is controlled by the output of the control unit selected by the selection unit based on the mass of the workpiece.

8. The control device as described in claim 7, characterized in that, It also includes a computing unit that executes an action program, which at least specifies actions to control the movement of the horizontal joint robot, including loading actions to hold the workpiece and unloading actions to remove the workpiece from the horizontal joint robot. According to the switching instruction from the arithmetic unit, the selection unit selects either the first control unit or the second control unit.

9. The control device as described in claim 8, characterized in that, When the computation unit moves the horizontal joint robot, which is not holding the workpiece, to the receiving position of the workpiece, it selects the first control unit through the selection unit. Then, if the mass of the workpiece exceeds the threshold, it selects the second control unit through the selection unit before the workpiece contacts the horizontal joint robot.

10. The control device as described in claim 8 or 9, characterized in that, The computing unit moves the horizontal multi-joint robot to the loading / unloading position of the workpiece while holding the workpiece. After performing the unloading action, the selection unit selects the first control unit.

11. The control device as described in claim 8 or 9, characterized in that, The horizontal multi-joint robot is equipped with a motor with a vertical axis. The calculation unit calculates the mass of the workpiece based on the torque command value of the motor for the vertical shaft, and outputs the switching command to the selection unit according to the calculated mass of the workpiece.

12. The control device as described in claim 8 or 9, characterized in that, The horizontal multi-joint robot is equipped with a load sensor to detect whether the workpiece is maintained in a manner that meets specified conditions. When the load sensor detects that the workpiece meets the specified conditions, the second control unit is selected.

Images