JP5433304B2 - Robot control apparatus and method - Google Patents

Description

  The present invention relates to a robot control apparatus and method.

  The robot control apparatus according to the prior art described in Patent Document 1 includes an input unit that inputs each axis value indicating the amount of rotation of each drive shaft of a robot arm, and each axis value input by the input unit is represented by orthogonal coordinates. Forward conversion means for forward conversion into position and orientation values indicating the position of the robot arm tip in the system and the posture of the robot, and correction means for correcting the position and orientation values forward-converted by the forward conversion means based on errors inherent in the robot And an inverse conversion means for inversely converting the position / orientation value corrected by the correction means into each axis value, and an instruction means for instructing the robot of each axis value inversely converted by the inverse conversion means. As a result, the positioning of the robot can be controlled with high accuracy even for teaching by each axis value of the robot.

JP 2006-159361 A.

Okada Takushi et al., “Multi-joint robot mechanism error correction method”, Transactions of the Japan Society of Mechanical Engineers (C), 51, 462, 324-331, February 1985.

  In the articulated robot according to the prior art, the position / orientation value indicating the position and orientation in the Cartesian coordinate system of the articulated robot is uniquely inverted to the rotation amount (hereinafter referred to as axis value) data of the drive shaft of each joint. It is known that there are positions called singular points that cannot be converted. However, Patent Document 1 does not describe an operation in a case where the position / orientation value corrected by the correcting unit cannot be converted back to each axis value. For this reason, at the singular point, it is necessary to perform control such as stopping the robot or operating the robot based on the position and orientation values before correction. For this reason, by performing continuous path control (hereinafter referred to as CP (Continuous Path) control) of the articulated robot, an operation that requires control of the trajectory such as spot welding, arc welding, or laser processing is performed. In some cases, the trajectory suddenly changed at a singular point. In addition, when switching from CP control to PTP (Point to Point) control in the vicinity of a singular point, there is a problem that the robot trajectory cannot be matched with a desired trajectory during PTP control.

  An object of the present invention is to solve the above problems and to provide a robot control device capable of improving the positioning accuracy of a robot compared to the prior art at a singular point where a corrected position and orientation value cannot be converted back into unique axis values. It is to provide a method.

According to a first aspect of the present invention, there is provided a robot control apparatus including a control unit that controls the position and orientation of a control point of a robot formed by connecting a plurality of links via at least one joint drive shaft at predetermined control timings. In the robot controller
The control means includes
Based on the position and orientation command value representing the position and orientation of the control point at each control timing, at each control timing,
(A) The position / orientation command value is inversely converted into each axis command value before correction representing the rotation amount of each joint drive shaft,
(B) correcting the position / orientation command value based on a robot-specific error in the actual position / orientation when a predetermined position / orientation is instructed, and calculating a corrected position / orientation command value;
(C) Determining whether or not the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft;
(C1) When reverse conversion is possible, each joint drive is performed by reversely converting the corrected position / orientation command value into the correction axis command values and subtracting the axis command values before correction from the correction axis command values. While calculating the axis correction value and storing it in the storage means,
(C2) When reverse conversion cannot be performed, an estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing; The corrected axis command values are calculated by adding the estimated correction values to the uncorrected axis command values, and the estimated correction values are stored in the storage means as correction values of the joint drive axes. ,
(D) Control is performed so that the joint drive shafts are rotated according to the correction axis command values.

In the robot control apparatus, when the control means cannot perform the reverse conversion,
An estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction value is calculated before the correction. In addition to calculating each correction axis command value by adding to each axis command value, and storing each estimated correction value as a correction value for each joint drive axis in the storage means,
An estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction value is calculated before the correction. The correction axis command values are calculated by adding to each axis command value.

In the robot control apparatus, the control means includes
At the control timing at which the corrected position / orientation command value can be converted back to the respective corrected axis command values, the control timing is a predetermined number of times from the latest control timing at which the corrected position / attitude command value cannot be converted back to the corrected axis command values To determine whether it is within the control timing of
The control timing at which the corrected position / orientation command value can be converted back to the corrected axis command value is within a predetermined number of times from the latest control timing at which the corrected position / attitude command value cannot be converted back to the corrected axis command value. When the correction position / orientation command value is inversely converted, each axis command value is subtracted from each correction axis command value obtained by inversely converting the correction position / orientation command value, and at least one control immediately before the correction value of the subtraction result and the control timing. An estimated correction value for each joint drive axis is calculated based on each correction value stored in the storage means at the timing.

  Further, in the robot control device, when the control means cannot perform the reverse conversion, each joint is based on the correction value of each joint drive shaft stored in the storage means at a plurality of control timings before the control timing. When the estimated correction value of the drive shaft is calculated, an average value of the correction values stored in the storage unit is used as the estimated correction value.

  Still further, in the robot control device, when the reverse conversion cannot be performed, each of the control means is based on correction values of the joint drive axes stored in the storage means at a plurality of control timings before the control timing. When calculating the estimated correction value of the joint drive axis, the estimated correction value of each joint drive axis is calculated based on each difference value of each correction value stored in the storage means.

Further, in the robot control device, when the control unit cannot perform the reverse conversion, each joint is based on a correction value of each joint drive shaft stored in the storage unit at one control timing before the control timing. When the estimated correction value for the drive shaft is calculated, the correction value stored in the storage means is used as each estimated correction value.

  Further, in the robot control device, when the control means cannot perform the reverse conversion, each joint is based on the correction value of each joint drive shaft stored in the storage means at a plurality of control timings before the control timing. When calculating the estimated correction value of the drive axis, a predetermined function is applied to each correction value stored in the storage unit, and the estimated correction value of each joint drive axis is calculated using the applied function. It is characterized by doing.

Still further, in the robot control device, when the reverse conversion cannot be performed, each of the control means is based on correction values of the joint drive axes stored in the storage means at a plurality of control timings before the control timing. When calculating the estimated correction value of the joint drive axis, the weighted average value of each correction value is calculated using a predetermined weight set in advance for each correction value stored in the storage unit, The weighted average value is used as each of the estimated correction values.

  In the robot control device, the plurality of control timings before the control timing are a plurality of control timings immediately before the control timing.

According to a second aspect of the present invention, there is provided a robot control apparatus comprising a control means for controlling the position and posture of a control point of a robot formed by connecting a plurality of links via at least one joint drive shaft at predetermined control timings. In the robot controller
The control means includes
Based on the position and orientation command value representing the position and orientation of the control point at each control timing, at each control timing,
(A) correcting the position / orientation command value based on a robot-specific error in the actual position / orientation when a predetermined position / orientation is instructed, and calculating a corrected position / orientation command value;
( B ) determining whether the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft;
( B 1) When reverse conversion is possible, the correction position / orientation command value is reversely converted into the correction axis command values and the correction axis command values are stored in the storage means.
( B2 ) When the reverse conversion cannot be performed, each joint drive axis at the control timing is based on the corrected axis command value of each joint drive axis stored in the storage unit at at least one control timing before the control timing. And the correction axis command value is calculated and stored in the storage means,
( C ) Control is performed to rotate the joint drive shafts according to the correction axis command values.

In the robot control apparatus, when the control means cannot perform the reverse conversion,
Based on the correction axis command value of each joint drive axis stored in the storage means at at least one control timing before the control timing, the correction axis command value of each joint drive axis at the control timing is calculated and Instead of storing in storage means,
Calculating a correction axis command value for each joint drive axis at the control timing based on a correction axis command value for each joint drive axis stored in the storage means at at least one control timing before the control timing; Features.

In the robot control device, at least one control timing before the control timing is at least one control timing immediately before the control timing.

A robot control method according to a third aspect of the present invention includes a control step of controlling the position and posture of a control point of a robot formed by connecting a plurality of links via at least one joint drive shaft at every predetermined control timing. In the robot control method,
The above control steps are:
Based on the position and orientation command value representing the position and orientation of the control point at each control timing, at each control timing,
(A) The position / orientation command value is inversely converted into each axis command value before correction representing the rotation amount of each joint drive shaft,
(B) correcting the position / orientation command value based on a robot-specific error in the actual position / orientation when a predetermined position / orientation is instructed, and calculating a corrected position / orientation command value;
(C) Determining whether or not the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft;
(C1) When reverse conversion is possible, each joint drive is performed by reversely converting the corrected position / orientation command value into the correction axis command values and subtracting the axis command values before correction from the correction axis command values. While calculating the axis correction value and storing it in the storage means,
(C2) When reverse conversion cannot be performed, an estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing; The corrected axis command values are calculated by adding the estimated correction values to the uncorrected axis command values, and the estimated correction values are stored in the storage means as correction values of the joint drive axes. ,
(D) Control is performed so that the joint drive shafts are rotated according to the correction axis command values.

In the robot control method, when the control step cannot perform the reverse conversion,
An estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction value is calculated before the correction. In addition to calculating each correction axis command value by adding to each axis command value, and storing each estimated correction value as a correction value for each joint drive axis in the storage means,
An estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction value is calculated before the correction. The correction axis command values are calculated by adding to each axis command value.

In the robot control method, the control step includes:
At a control timing at which the corrected position / orientation command value can be converted back to the corrected axis command value, the control timing can be a predetermined number of times from the latest control timing at which the corrected position / attitude command value cannot be converted back to the corrected axis command value. To determine whether it is within the control timing of
The control timing at which the corrected position / orientation command value can be converted back to the corrected axis command value is within a predetermined number of times from the latest control timing at which the corrected position / attitude command value cannot be converted back to the corrected axis command value. When the correction position / orientation command value is inversely converted, each axis command value is subtracted from each correction axis command value obtained by inversely converting the correction position / orientation command value, and at least one control immediately before the correction value of the subtraction result and the control timing. An estimated correction value for each joint drive axis is calculated based on each correction value stored in the storage means at the timing.

  Furthermore, in the robot control method, when the control step cannot be reversely converted, each joint is based on correction values of each joint drive shaft stored in the storage unit at a plurality of control timings before the control timing. When the estimated correction value of the drive shaft is calculated, an average value of the correction values stored in the storage unit is used as the estimated correction value.

  Still further, in the robot control method, when the reverse conversion cannot be performed, each of the control steps is performed based on correction values of the joint drive axes stored in the storage unit at a plurality of control timings before the control timing. When calculating the estimated correction value of the joint drive axis, the estimated correction value of each joint drive axis is calculated based on each difference value of each correction value stored in the storage means.

In the robot control method, when the control step cannot be reversely converted, each joint is based on a correction value of each joint drive axis stored in the storage unit at one control timing before the control timing. When the estimated correction value for the drive shaft is calculated, the correction value stored in the storage means is used as each estimated correction value.

  Furthermore, in the robot control method, when the control step cannot be reversely converted, each joint is based on correction values of each joint drive shaft stored in the storage unit at a plurality of control timings before the control timing. When calculating the estimated correction value of the drive axis, a predetermined function is applied to each correction value stored in the storage unit, and the estimated correction value of each joint drive axis is calculated using the applied function. It is characterized by doing.

Still further, in the robot control method, when the reverse conversion cannot be performed, each of the control steps is performed based on correction values of the joint drive axes stored in the storage unit at a plurality of control timings before the control timing. When calculating the estimated correction value of the joint drive axis, the weighted average value of each correction value is calculated using a predetermined weight set in advance for each correction value stored in the storage unit, The weighted average value is used as each of the estimated correction values.

  In the robot control method, the plurality of control timings before the control timing are a plurality of control timings immediately before the control timing.

A robot control method according to a fourth aspect of the present invention includes a control step of controlling the position and posture of a control point of a robot formed by connecting a plurality of links via at least one joint drive shaft at every predetermined control timing. In the robot control method,
The above control steps are:
Based on the position and orientation command value representing the position and orientation of the control point at each control timing, at each control timing,
(A) correcting the position / orientation command value based on a robot-specific error in the actual position / orientation when a predetermined position / orientation is instructed, and calculating a corrected position / orientation command value;
( B ) determining whether the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft;
( B 1) When reverse conversion is possible, the correction position / orientation command value is reversely converted into the correction axis command values and the correction axis command values are stored in the storage means.
( B2 ) When the reverse conversion cannot be performed, each joint drive axis at the control timing is based on the corrected axis command value of each joint drive axis stored in the storage unit at at least one control timing before the control timing. And the correction axis command value is calculated and stored in the storage means,
( C ) Control is performed to rotate the joint drive shafts according to the correction axis command values.

In the robot control method, when the control step cannot perform the reverse conversion,
Based on the correction axis command value of each joint drive axis stored in the storage means at at least one control timing before the control timing, the correction axis command value of each joint drive axis at the control timing is calculated and Instead of storing in storage means,
Calculating a correction axis command value for each joint drive axis at the control timing based on a correction axis command value for each joint drive axis stored in the storage means at at least one control timing before the control timing; Features.

In the robot control method, at least one control timing before the control timing is at least one control timing immediately before the control timing.

  According to the robot control device and method of the present invention, it is determined whether or not the corrected position / orientation command value can be inversely converted to each corrected axis command value representing the rotation amount of each joint drive shaft, and when the inverse conversion can be performed, The correction position / orientation command value is inversely converted to each correction axis command value, and each axis command value before correction is subtracted from each correction axis command value to calculate a correction value for each joint drive axis and store it in the storage means. On the other hand, when reverse conversion cannot be performed, an estimated correction value of each joint drive axis is calculated based on the correction value of each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction Each correction axis command value is calculated by adding the value to each axis command value before correction, and each estimated correction value is stored in the storage means as a correction value for each joint drive axis. Rotate the joint drive shaft Since the controlled so, the singularity that can not inversely converted correction position and orientation command values to each unique correction axis command value, the positioning accuracy of the robot can be improved as compared with the prior art.

1 is a block diagram illustrating a configuration of a robot control system including a robot control device 1, a robot 7, a computer 50, a teach pendant 51, and an external storage device 52 according to a first embodiment of the present invention. It is a block diagram which shows the structure of the robot 7 of FIG. It is a flowchart which shows the robot control process performed by the controller 2 of FIG. It is a flowchart which shows the estimation correction value calculation process performed in step S8 of FIG. 4 is a graph showing an example of an axis command value θnm, a corrected axis command value θcnm, and a correction value Dnm calculated by the robot control process of FIG. It is a flowchart which shows the estimation correction value calculation process which concerns on the 1st modification of the 1st Embodiment of this invention. It is a flowchart which shows the estimation correction value calculation process which concerns on the 2nd modification of the 1st Embodiment of this invention. It is a flowchart which shows the estimation correction value calculation process which concerns on the 3rd modification of the 1st Embodiment of this invention. It is a flowchart which shows the estimation correction value calculation process which concerns on the 4th modification of the 1st Embodiment of this invention. It is a flowchart which shows the estimation correction value calculation process which concerns on the 5th modification of the 1st Embodiment of this invention. It is a flowchart which shows the robot control processing which concerns on the 2nd Embodiment of this invention. It is a flowchart which shows the estimation correction value calculation process performed in step S86 of FIG. 12 is a graph showing an example of an axis command value θnm, a corrected axis command value θcnm, and a correction value Dnm calculated by the robot control process of FIG. It is a flowchart which shows the robot control processing which concerns on the 3rd Embodiment of this invention. It is a flowchart which shows the present position output process which concerns on the 4th Embodiment of this invention. It is a flowchart which shows the robot control processing which concerns on the 5th Embodiment of this invention.

  Hereinafter, embodiments according to the present invention will be described with reference to the drawings. In addition, in each following embodiment, the same code | symbol is attached | subjected about the same component.

First embodiment.
FIG. 1 is a block diagram showing a configuration of a robot control system including a robot control device 1 according to a first embodiment of the present invention, a robot 7, a computer 50, a teach pendant 51, and an external storage device 52. FIG. 2 is a block diagram showing the configuration of the robot 7 of FIG. In FIG. 1, a robot control apparatus 1 according to the present embodiment includes a controller 2 composed of a CPU (Central Processing Unit), a memory 3, an interface circuit 4, a drive device 5, and a measurement device 6. Is done.

Here, as will be described in detail later, the robot control apparatus 1 is configured to control a control point A of an articulated robot (also referred to as a robot arm) 7 in which a plurality of links 11 to 16 are connected via a plurality of joint drive shafts. Based on the position and orientation command values P1m to P6m representing the position and orientation at the control timing Tm, at each control timing Tm,
(A) The position / orientation command values P1m to P6m are inversely converted into respective axis command values θnm before correction representing the rotation amount of each joint drive shaft,
(B) correcting the position / orientation command values P1m to P6m based on the robot-specific error in the actual position / orientation when the predetermined position / orientation is instructed, and calculating corrected position / orientation command values Pc1m to Pc6m;
(C) It is determined whether or not the corrected position / orientation command values Pc1m to Pc6m can be inversely converted into the corrected axis command values θcnm representing the rotation amounts of the joint drive shafts.
(C1) When reverse conversion is possible, each of the corrected position / orientation command values Pc1m to Pc6m is converted back to each corrected axis command value θcnm, and each axis command value θnm before correction is subtracted from each corrected axis command value θcnm. While calculating the correction value Dnm of the joint drive axis and storing it in the memory 3,
(C2) When reverse conversion cannot be performed, the estimated correction value DPnm of each joint drive axis is calculated based on the correction value Dnm of each joint drive axis stored in the memory 3 at the three control timings immediately before the control timing Tm. Then, each corrected axis command value θcnm is calculated by adding each estimated correction value DPnm to each axis command value θnm before correction, and each estimated correction value DPnm is stored in the memory 3 as the correction value Dnm of each joint drive axis. And
(D) It is characterized in that each joint drive shaft is controlled to rotate in accordance with each correction axis command value θcnm.

  In FIG. 2, the robot 7 is a six-axis articulated robot, and includes a base 10, links 11 to 16, rotary joints J1, J4, J6, swivel joints J2, J3, J5, and a hand 17. Configured. Here, one end of the base pillar 10 is fixed vertically to the base 9, and the other end is connected to the joint J1. Each end of the link 11 is connected to the joints J1 and J2, each end of the link 12 is connected to the joints J2 and J3, each end of the link 13 is connected to the joints J3 and J4, and Each end is connected to joints J4 and J5, and each end of the link 15 is connected to joints J5 and J6. Further, one end of the link 16 is connected to the joint J 6, and the other end A of the link 16 (hereinafter referred to as a control point A) is connected to the hand 17. For example, a predetermined tool such as a welding torch or a gripping hand is attached to the hand 17 provided at the control point A. Here, as shown in FIG. 2, the work origin O of the robot 7 (the position where the base 10 is fixed to the base 9) is the origin of the right-handed orthogonal coordinate system, and the direction orthogonal to the base 9. Is the Z direction.

  In FIG. 1, the computer 50 performs CP control based on predetermined teaching data representing the position of the control point A at each teaching point of the robot 7 and the posture of the hand 17 and the moving speed of the hand 17. The position of the control point A and the posture of the hand 17 are calculated at each control timing Tm (m = 1, 2,..., M) of the robot 7 so that the control point A draws a straight line between the points. Then, the computer 50 outputs control data including position / orientation command values P1m to P6m representing the position of the control point A and the attitude of the hand 17 at each control timing Tm to the robot control device 1. Here, the position / orientation command values P1m, P2m and P3m at the control timing Tm are the X coordinate, Y coordinate and Z coordinate in the XYZ coordinate system of the control point A, respectively, and the position / orientation command values P4m, P5m and P6m are respectively These are the roll angle, pitch angle, and yaw angle of the hand 17. Further, the display 50 d displays the operation state of the computer 50 and displays each data received from the robot control device 1. Furthermore, the input unit 50i is, for example, a keyboard, and is used to input an instruction command for controlling the operation of the robot control apparatus 1. Further, the teach pendant 51 holds the hand 17 by a human and moves it to a target position and posture to cause the robot 7 to perform a predetermined operation pattern, whereby the position of the control point A at each teaching point of the robot 7 and This is an operation terminal for acquiring predetermined teaching data representing the posture of the hand 17 and generating control data based on the acquired teaching data in the same manner as the computer 50.

  The interface circuit 4 executes predetermined interface processing such as signal conversion with the robot control device 1 on the control data input from the computer 50, the teach pendant 51, and the external storage device 52, and outputs the control data to the controller 2. Predetermined interface processing such as signal conversion with the computer 50, the teach pendant 51, and the external storage device 52 is executed on the data input from the controller 2 and output to the computer 50. Further, the controller 2 controls the overall operation of the robot control apparatus 1 and stores position / orientation command values P1m to P6m in the memory 3. Based on the stored position / orientation command values P1m to P6m, details will be described later. The corrected axis command value θcnm, which is the rotation amount of the drive shaft of each joint Jn (n = 1, 2,..., 6) at each control timing Tm, is calculated by the robot control processing to be output to the drive device 5. The drive device 5 is an actuator for rotating the drive shaft of each joint Jn, and rotates the drive shaft of each joint Jn according to the correction axis command value θcnm from the controller 2. The memory 3 stores various software programs executed by the controller 2 in advance. The measuring device 6 is an angle sensor provided for each joint Jn, measures the actual axis value of each joint Jn at each control timing Tm, and outputs the measured axis value to the controller 2. .

  Next, robot control processing executed by the controller 2 will be described. FIG. 3 is a flowchart showing the robot control process executed by the controller 2 of FIG. 1, and FIG. 4 is a flowchart showing the estimated correction value calculation process executed in step S8 of FIG. In FIG. 3, first, in step S1, the position / orientation command values P1m to P6m at each control timing Tm are read from the memory 3, and in step S2, the parameter m is set to an initial value 1. In step S3, the controller 2 reversely converts the position / orientation command values P1m to P6m at the control timing Tm into an axis command value θnm which is the rotation amount of the drive shaft of each joint Jn. When the position / orientation command values P1m to P6m cannot be converted back into the axis command value θnm, the controller 2 controls the driving device 5 so as to stop the operation of the robot 7 by performing a predetermined error process. The control process ends.

  Subsequently, in step S4, the controller 2 corrects each of the position / orientation command values P1m to P6m using the “multi-joint robot mechanism error correction method” described in Non-Patent Document 1, and then corrects the corrected position / orientation command. Values Pc1m to Pc6m are calculated. Here, in the articulated robot mechanism error correction method, the actual position of the control point A and the hand 17 when the position of the control point A is moved to a predetermined position and the hand 17 is controlled to assume a predetermined posture. This corrects the deviation from the actual posture. Such a deviation is caused by a mechanism error based on deflection, processing error, assembly error, and the like inherent to the robot 7, an installation error, and an error of the machine origin of each axis. Hereinafter, such correction using the error correction method is referred to as correction using a high-precision robot model.

  Next, in step S5, it is determined whether or not the corrected position / orientation command values Pc1m to Pc6m can be inversely converted into the unique correction axis command values θc1m to θc6m. If YES, the process proceeds to step S6. Proceed to step S8. Here, the controller 2 performs the determination in step S5 using a simultaneous equation discriminant for inversely converting the corrected position / orientation command values Pc1m to Pc6m into the corrected axis command values θc1m to θc6m.

  In step S6, the corrected position / orientation command values Pc1m to Pc6m are inversely converted into corrected axis command values θc1m to θc6m, and in step S7, the axis command value θnm before correction is subtracted from the corrected axis command value θcnm. Each correction value Dnm at the control timing Tm for Jn is calculated. Then, the process proceeds to step S11.

  On the other hand, in step S8, the estimated correction value DPnm is estimated for each joint Jn by the “estimated correction value calculation process” of FIG. In the estimated correction value calculation process of FIG. 4, in step S21, the correction value Dnm-1, the correction value Dnm-2, and the correction value Dnm- stored in the memory 3 at the control timings Tm-1, Tm-2, and Tm-3. 3 is calculated, and in step S22, the estimated correction value DPnm is set to the average value Dave, and the process returns to the robot control process. In step S9 in FIG. 3, the correction value Dnm of each joint Jn is set to the estimated correction value DPnm. In step S10, the correction value Dnm is added to the uncorrected axis command value θnm, thereby correcting the corrected axis command value. θcnm is calculated, and the process proceeds to step S11.

  In step S11, the correction value Dnm of each joint Jn is stored in the memory 3, and in step S12, the controller 2 controls the robot drive device 5 to operate the robot 7 based on the correction axis command value θcnm. Next, in step S13, it is determined whether or not the control timing Tm is a control timing before the last control timing TM. If YES, the robot control process is terminated. If NO, the parameter m is determined in step S14. 1 is added to and the process returns to step S3.

  If the corrected position / orientation command values Pc1m to Pc6m cannot be converted back into the unique corrected axis command values θc1m to θc6m at the control timing T2 or T3, they are stored in the memory 3 at the immediately preceding one or two control timings. The estimated correction value calculation process shown in FIG. If the corrected position / orientation command values Pc1m to Pc6m cannot be converted back into the unique corrected axis command values θc1m to θc6m at the control timing T1, the robot control process is terminated.

  FIG. 5 is a graph showing an example of the axis command value θnm, the corrected axis command value θcnm, and the correction value Dnm calculated by the robot control process of FIG. In FIG. 5, at the control timings T1 to T4, the corrected position / orientation command values Pc1m to Pc6m can be inversely converted into the unique corrected axis command values θc1m to θc6m, and it is determined as YES in step S5 of FIG. . However, at the control timings T5 and T6, the corrected position / orientation command values Pc1m to Pc6m cannot be converted into the unique corrected axis command values θc1m to θc6m, and it is determined as NO in step S5 of FIG. Therefore, for example, at the control timing T5, DPn5 that is an average value of the correction values Dn2, Dn3, and Dn4 stored in the memory 3 at the control timings T2, T3, and T4 is used as the correction value Dn5.

  As described above in detail, according to the present embodiment, even when the corrected position / orientation command values Pc1m to Pc6m cannot be reversely converted into the unique corrected axis command values θc1m to θc6m at the control timing Tm, immediately before the control timing Tm. Based on the correction value stored in the memory 3 at three control timings, the correction value Dnm at the control timing Tm is estimated to calculate the correction axis command value θcnm. Therefore, even at a singular point where the corrected position / orientation command values Pc1m to Pc6m cannot be converted back to the unique correction axis command values θc1m to θc6m, the trajectory of the robot 7 does not change abruptly as compared with the prior art, and positioning accuracy is improved. It can be improved.

  In the present embodiment, when the corrected position / orientation command values Pc1m to Pc6m cannot be converted back into the unique corrected axis command values θc1m to θc6m, they are stored in the memory 3 at the three control timings immediately before the control timing Tm. The estimated correction value DPnm was calculated using the correction value Dnm. However, the present invention is not limited to this, and the estimated correction value DPnm may be calculated using the correction values Dnm stored in the memory 3 at a plurality of control timings immediately before the control timing Tm.

First modified example of the first embodiment.
FIG. 6 is a flowchart showing an estimated correction value calculation process according to the first modification of the first embodiment of the present invention. In step S31 of FIG. 6, the difference value DD1 is calculated by subtracting the correction value Dnm-2 from the correction value Dnm-3 stored in the memory 3, and in step S32, the correction value DDnm-2 stored in the memory 3 is calculated. The correction value DDnm-1 is subtracted from the difference value DD2. In step S33, the difference value DD3 is calculated by subtracting the difference value DD1 from the difference value DD3. Further, in step S34, by adding the difference value DD3 to the correction value Dnm-1 stored in the memory 3, an estimated correction value DPnm of the axis value of the joint Jn at the control timing Tm is calculated, and the robot control of FIG. Return to processing. According to this modification, when the corrected position / orientation command values Pc1m to Pc6m cannot be converted into the unique corrected axis command values θc1m to θc6m at the control timing Tm, the immediately preceding control timings Tm-3, Tm-2, and Tm-1 are obtained. Since the estimated correction value DPnm is calculated using the difference values DD1 and DD2 of the correction values Dnm-3, Dnm-2, and Dnm-1 stored in the memory 3, the same effects as in the first embodiment are obtained. Play.

  In this modification, the estimated correction value DPnm is calculated using the difference values DD1 and DD2 of the correction value Dnm stored in the memory 3 at the three control timings immediately before the control timing Tm. However, the present invention is not limited to this, and the estimated correction value DPnm may be calculated using each difference value of the correction value Dnm stored in the memory 3 at a plurality of control timings three or more times immediately before the control timing Tm. .

Second modification of the first embodiment.
FIG. 7 is a flowchart showing an estimated correction value calculation process according to the second modification of the first embodiment of the present invention. In step S41 in FIG. 7, the correction value Dnm-1 stored in the memory 3 is set as the estimated correction value DPnm, and the process returns to the robot control process in FIG. According to this modification, when the corrected position / orientation command values Pc1m to Pc6m cannot be converted into the unique corrected axis command values θc1m to θc6m at the control timing Tm, the correction stored in the memory 3 at the immediately preceding control timing Tm−1. Since the value Dnm-1 is set as the estimated correction value DPnm, the estimated correction value DPnm can be estimated in a short time compared to the first embodiment.

Third modification of the first embodiment.
FIG. 8 is a flowchart showing an estimated correction value calculation process according to the third modification of the first embodiment of the present invention. In step S51 of FIG. 8, a predetermined function is applied to the correction value Dnm stored in the memory 3 at a predetermined number of control timings immediately before the control timing Tm stored in the memory 3, and the applied function is The estimated correction value DPnm is used to calculate, and the process returns to the robot control process of FIG. Specifically, for example, a cubic function is applied to each of the correction values Dnm-5 to Dnm-1 stored in the memory 3 at five control timings Tm-5 to Tm-1 immediately before the control timing Tm. The estimated correction value DPnm at the control timing Tm is calculated using the fitted cubic function. According to this modified example, when the corrected position / orientation command values Pc1m to Pc6m cannot be converted into the unique corrected axis command values θc1m to θc6m at the control timing Tm, they are stored in the memory 3 at the predetermined predetermined number of control timings. Since the estimated correction value DPnm is calculated using each correction value Dnm, the same effects as those of the first embodiment can be obtained.

Fourth modified example of the first embodiment.
FIG. 9 is a flowchart showing an estimated correction value calculation process according to the fourth modification of the first embodiment of the present invention. In step S61 of FIG. 9, the weighted average value Davew1 of the correction values Dnm-3 to Dnm-1 stored in the memory 3 using the weights W3, W2, and W1 set in advance so that W1>W2> W3. Is calculated. Specifically, the weights W1, W2, and W3 are preset to 0.5, 0.3, and 0.2, respectively. In step S62, the weighted average value Dave1 is set as the estimated correction value DPnm, and the process returns to the robot control process of FIG. According to the present modification, the weighted average value Davew1 is calculated by giving a larger weight Wm to the control timing close to the control timing Tm, so that the estimated correction value DPnm can be calculated with higher accuracy than in the first embodiment. I can guess.

  In this modification, the estimated correction value DPnm is calculated using the correction value Dnm stored in the memory 3 at the three control timings immediately before the control timing Tm. However, the present invention is not limited to this, and the estimated correction value DPnm may be calculated using the correction values Dnm stored in the memory 3 at a plurality of control timings immediately before the control timing Tm.

Fifth modification of the first embodiment.
FIG. 10 is a flowchart showing an estimated correction value calculation process according to the fifth modification of the first embodiment of the present invention. In this modification, when the correction value Dnm of each joint Jn is stored in the memory 3 in step S11 of FIG. 3, whether the correction value Dnm was calculated by the process of step S7 of FIG. A flag indicating whether it has been calculated is stored together. In step S71 of FIG. 10, weights W3, W2, and W1 set in advance so that the correction value calculated by the process of step S7 is larger than the weight for the correction value estimated by the estimation correction value calculation process. The weighted average value Dave2 of the correction values Dnm-3 to Dnm-1 stored in the memory 3 is calculated. Specifically, the weight for the correction value calculated by the process of step S7 is preset to a value twice the weight for the correction value estimated by the estimated correction value calculation process. In step S72, the weighted average value Dave2 is set as the estimated correction value DPnm, and the process returns to the robot control process of FIG. According to the present modification, the weight for the correction value calculated by the process of step S7 is set larger than the weight for the correction value estimated by the estimation correction value calculation process. Therefore, compared to the first embodiment, The estimated correction value DPnm can be estimated with high accuracy.

  In this modification, the estimated correction value DPnm is calculated using the correction value Dnm stored in the memory 3 at the three control timings immediately before the control timing Tm. However, the present invention is not limited to this, and the estimated correction value DPnm may be calculated using the correction values Dnm stored in the memory 3 at a plurality of control timings immediately before the control timing Tm.

Second embodiment.
FIG. 11 is a flowchart showing the robot control process according to the second embodiment of the present invention, and FIG. 12 is a flowchart showing the estimated correction value calculation process executed in step S86 of FIG. Compared with the robot control process according to the first embodiment, the robot control process according to the present embodiment is performed at the control timing Tm at which the corrected position / orientation command values Pc1m to Pc6m can be converted back to the corrected axis command values θcnm. It is determined whether or not the control timing Tm is within three control timings from the latest control timing at which the corrected position / orientation command values Pc1m to Pc6m cannot be converted back into the respective correction axis command values θcnm, and the corrected position / orientation command values Pc1m to When the control timing at which Pc6m can be converted back to each corrected axis command value θcnm is within three control timings from the latest control timing at which corrected position / orientation command values Pc1m to Pc6m cannot be converted back to each corrected axis command value θcnm From each axis command value θctnm obtained by inversely converting the corrected position / orientation command values Pc1m to Pc6m, each axis The command value θnm is subtracted, and each correction value Dtnm−1, Dtm− stored in the memory 3 at the provisional correction value Dtnm of the subtraction result and the two control timings Tm−1 and Tm−2 immediately before the control timing Tm. The estimated correction value DPnm of each joint drive shaft is calculated based on 2.

  In FIG. 11, first, in step S <b> 1, position / orientation command values P <b> 1 m to P <b> 6 m at each control timing Tm are read from the memory 3. In step S81, the parameter k is set to an initial value 0, and in step S2, the parameter m is set to an initial value 1. In step S3, the controller 2 reversely converts the position / orientation command values P1m to P6m at the control timing Tm into an axis command value θnm which is the rotation amount of the drive shaft of each joint Jn. When the position / orientation command values P1m to P6m cannot be converted back into the axis command value θnm, the controller 2 controls the driving device 5 so as to stop the operation of the robot 7 by performing a predetermined error process. The control process ends.

  Subsequently, in step S4, the controller 2 corrects the position / orientation command values P1m to P6m using the high-precision robot model, and calculates the corrected corrected position / orientation command values Pc1m to Pc6m. Next, in step S5, it is determined whether or not the corrected position / orientation command values Pc1m to Pc6m can be inversely converted into the unique correction axis command values θc1m to θc6m. If YES, the process proceeds to step S82, whereas if NO, Proceeding to step S84, the parameter k is set to 0, and for each joint Jn, the estimated correction value DPnm is estimated by the “estimated correction value calculating process” of FIG. 4, and the process proceeds to step S9.

  In step S82, 1 is added to the parameter k. In step S83, it is determined whether or not the parameter k is 3 or less. If YES, the process proceeds to step S86. If NO, the process proceeds to step S6. In step S6, the corrected position / orientation command values Pc1m to Pc6m are inversely converted into corrected axis command values θc1m to θc6m. In step S7, the axis command value θnm is subtracted from the corrected axis command value θcnm, thereby controlling each joint Jn. Each correction value Dnm at timing Tm is calculated. Then, the process proceeds to step S11.

In step S86, the estimated correction value DPnm is estimated for each joint Jn by the “estimated correction value calculation process” of FIG. Specifically, in step S201 of FIG. 12, the corrected position / orientation command values Pc1m to Pc6m are inversely converted to temporary correction axis command values θct1m to θct6m. In step S202, the temporary correction value Dtnm for each joint Jn is calculated by subtracting the axial command value θnm from the temporary correction axis command value θcnm. Further, in step S203, the temporary correction value Dtnm and the correction values Dnm-1 and Dnm-2.
The average value Dave1 is calculated, and in step S204, the calculated average value Dave1 is set as the estimated correction value DPnm, and the process returns to the robot control process.

  In step S9, the correction value Dnm of each joint Jn is set to the estimated correction value DPnm. In step S10, the correction value Dnm is calculated by adding the correction value Dnm to the axis command value θnm, and the process proceeds to step S11. .

  In step S11, the correction value Dnm of each joint Jn is stored in the memory 3, and in step S12, the controller 2 controls the robot drive device 5 to operate the robot 7 based on the correction axis command value θcnm. Next, in step S13, it is determined whether or not the control timing Tm is a control timing before the last control timing TM. If YES, the robot control process is terminated. If NO, the parameter m is determined in step S14. 1 is added to and the process returns to step S3.

  FIG. 13 is a graph showing an example of the axis command value θnm, the corrected axis command value θcnm, and the correction value Dnm calculated by the robot control process of FIG. In FIG. 13, at the control timings T1 to T4 and T7 to T10, the corrected position / orientation command values Pc1m to Pc6m can be inversely converted to the unique corrected axis command values θc1m to θc6m, and YES in step S5 of FIG. It has been judged. However, at the control timings T5 and T6, the corrected position / orientation command values Pc1m to Pc6m cannot be converted into the unique corrected axis command values θc1m to θc6m, and it is determined as NO in step S5 of FIG. Therefore, for example, at the control timing T5, DPn5 that is an average value of the correction values Dn2, Dn3, Dn4 at the control timings T2, T3, T4 is used as the correction value Dn5. The control timings T7, T8, and T9 are each controlled within three times after the corrected position / orientation command values Pc1m to Pc6m are changed from a state where the corrected position / orientation command values Pc1m to Pc6m cannot be converted back to the unique corrected axis command values θc1m to θc6m. It is timing. For this reason, the estimated correction value DPnm is calculated by the estimated correction value calculation process of FIG. Further, the control timing T10 is the fourth control timing after the correction position / orientation command values Pc1m to Pc6m are changed from the state in which the correction position / orientation command values Pc1m to Pc6m cannot be converted back to the unique correction axis command values θc1m to θc6m. The correction value calculation process is not performed.

  As described in detail above, according to the present embodiment, within three times after the corrected position / orientation command values Pc1m to Pc6m are changed from the state in which the corrected position / orientation command values Pc1m to Pc6m cannot be converted back to the unique corrected axis command values θc1m to θc6m. At each control timing, the estimated correction value DPnm was calculated by the estimated correction value calculation process of FIG. Therefore, when the corrected position / orientation command values Pc1m to Pc6m are changed from the state in which the corrected position / orientation command values Pc1m to Pc6m cannot be converted back to the unique correction axis command values θc1m to θc6m to the state in which they can be converted back, the corrected axis command values are compared. The change in θcnm becomes small, and the positioning accuracy can be improved by operating the hand 17 smoothly.

  In the present embodiment, the correction value is estimated by the estimated correction value calculation process of FIG. However, the present invention is not limited to this, and the first embodiment uses the temporary correction value Dtnm at the control timing Tm and the correction value Dnm stored in the memory 3 at least at one control timing immediately before the control timing Tm. The estimated correction value DPnm may be calculated by the estimated correction value calculating process similar to the estimated correction value calculating process according to the first to fifth modifications. Further, the three control timings Tm-1 and Tm- immediately before the control timing Tm when the corrected position / orientation command values Pc1m to Pc6m are changed from the state where the corrected position / orientation command values Pc1m to Pc6m cannot be converted back to the unique corrected axis command values θc1m to θc6m. The correction value DPnm stored in the memory 3 at 2 and Tm−3, and the estimated correction value by the estimated correction value calculation process according to the first embodiment or the first to fifth modifications of the first embodiment. DPnm may be calculated.

Third embodiment.
In each embodiment and each modification described above, the computer 50 outputs control data including the position / orientation command values P1m to P6m representing the position of the control point A and the attitude of the hand 17 at each control timing Tm to the robot control device 1. did. In the present embodiment, the computer 50 outputs one control data of the following three control data Da, Db, and Dc to the robot controller 1.

(1) Control data Da.
Position / orientation command values P1m to P6m representing the position of the control point A and the attitude of the hand 17 at each control timing Tm are included.
(2) Control data Db.
It includes the axis command value θanm of each joint Jn that needs to be corrected for displacement caused by the inherent error of the robot 7 at each control timing Tm.
(3) Control data Dc.
The axis command value θnm of each joint Jn that does not require correction of displacement due to the inherent error of the robot 7 at each control timing Tm is included.

  Here, the correction axis command value θcnm that does not require correction of each joint Jn is an axis value that takes into account a positional deviation caused by an error inherent in the robot 7. Specifically, the teach pendant 51 is used. Then, the robot 7 is directly moved to the target position and posture for teaching, or the hand 17 is held by the human and moved to the target position and posture for teaching, and the taught position and posture are measured. It is obtained by measuring each axis value of each obtained joint Jn.

  FIG. 14 is a flowchart showing a robot control process according to the third embodiment of the present invention. In step S91, it is determined whether or not the control data in the memory 3 includes the position / orientation command values P1m to P6m at each control timing Tm. If YES, the robot control process of FIG. 3 is executed in step S92. While the robot control process ends, the process proceeds to step S93 when the determination is NO. In step S93, it is determined whether or not the control data in the memory 3 includes an axis command value θanm that needs to be corrected at each control timing Tm. If YES, the process proceeds to step S94. If NO, step S95 is performed. Proceed to

  In step S94, assuming that there is no error in the robot 7, the axis command value θanm is forward-converted to position / orientation command values P1m to P6m in the orthogonal coordinate system, and the process proceeds to step S92. On the other hand, in step S95, the robot driving device 5 is controlled to operate the robot 7 based on the axis command value θnm that does not require correction, and the robot control process is terminated.

  According to the present embodiment, as compared with the first embodiment, any control data of the control data Da, Db, and Dc input from the computer 50 is corrected based on the inherent error of the robot 7. The operation of the robot 7 can be controlled in accordance with the corrected axis command value θcnm. In particular, when the control data Dc is input, the mechanical stopper and the soft limit of each drive axis act as instructed, which is effective when it is desired to reliably grasp the limit of the operation area of each axis.

  Furthermore, for example, because the correction amount based on the error is not appropriate near the limit of the operation range of the robot 7, or because data such as a cable to be avoided when creating control data offline is leaked, The control data may be directly taught by a human. When the offline control data is corrected by a human, the offline control data and the direct teaching control data by a human are mixed. According to the present embodiment, since the control data Db and the control data Dc can be used, the positioning accuracy of the robot 7 is improved as compared with the prior art even when the off-line teaching data is corrected by teaching by a human. it can.

Fourth embodiment.
FIG. 15 is a flowchart showing a current position output process according to the fourth embodiment of the present invention. The controller 2 performs current position output processing at each control timing Tm in synchronization with the robot control processing according to the first or second embodiment. In step S301 in FIG. 15, the controller 2 acquires the measurement value αn at the timing Tm of the axis value of each joint Jn using the measurement device 6. In step S302, it is determined whether or not to reversely correct the acquired measurement value αn in accordance with an instruction command from the computer 50. If YES, the process proceeds to step S303. If NO, the process proceeds to step S307. Here, reverse correction refers to subtracting the correction value Dnm calculated in step S7 or S9 of the robot control process from the measured value αn. In step S307, the measurement value αn is output to the computer 50, and the current position output process is terminated. On the other hand, in step S303, the reversely corrected axis value θDn is calculated by subtracting the correction value Dnm stored in the memory 3 from the measured value αn, and the process proceeds to step S304. In step S304, it is determined whether to output the current position of the robot 7 as each axis value in accordance with an instruction command from the computer 50. If YES, the process proceeds to step S306. If NO, the process proceeds to step S305. In step S306, the reversely corrected axis value θDn is output to the computer, and the current position output process is terminated. In step S305, the inversely corrected axis value θDn is forward-converted to the display position / orientation value PDn and output to the computer 50, and the current position output process is terminated. The computer 50 outputs and displays the measured value αn of the axis value from the robot controller 1, the reversely corrected axis value θDn, or the display position / posture value PDn on the display 50d.

  According to the present embodiment, the axis value αn measured at the control timing Tm can be reversely corrected using the correction value Dnm stored in the memory 3 and output.

Fifth embodiment.
FIG. 16 is a flowchart showing a robot control process according to the fifth embodiment of the present invention. Compared with the robot control process (see FIG. 11) according to the second embodiment, the robot control process according to the present embodiment uses the corrected position / orientation command values Pc1m to Pc6m to represent the rotation amount of each joint drive shaft. It is determined whether or not it can be converted back to the corrected axis command value θcnm, and when it can be converted back, the corrected position / orientation command values Pc1m to Pc6m are converted back to the corrected axis command values θcnm and stored in the memory 3, When reverse conversion cannot be performed, the correction axis of each joint drive axis at the control timing Tm based on the correction axis command value θcnm of each joint drive axis stored in the memory 3 at the three control timings immediately before the upper control timing Tm The command value θcnm is calculated and stored in the memory 3. Specifically, the robot control process in FIG. 16 replaces the processes in steps S84 and S86 in FIG. 11 with the processes in steps S101 and S104, and replaces the processes in steps S9 and S10 with the processes in step S102. The process in step S11 is replaced with the process in step S103.

  If it is determined in step S5 that the corrected position / orientation command values Pc1m to Pc6m cannot be inversely converted to the unique correction axis command values θc1m to θc6m, the process of step S101 is executed. In step S101, for each joint Jn, the corrected axis command values θcnm-3 and θcnm− stored in the memory 3 at the three control timings Tm−3, Tm−2 and Tm−1 immediately before the control timing Tm. The estimated correction axis command value θPnm is estimated based on 2, θcnm−1, and the process proceeds to step S102. Specifically, the average value of the correction axis command values θcnm−3, θcnm−2, and θcnm−1 is set as the estimated correction axis command value θPnm.

  Further, in step S83, it is determined that the control timings are within three times after the correction position / orientation command values Pc1m to Pc6m are changed from the state where the correction position / orientation command values Pc1m to Pc6m cannot be converted back to the unique correction axis command values θc1m to θc6m. When it is determined, the estimated corrected axis command value θPnm is estimated for each joint Jn based on the corrected position / orientation command value Pcnm, the corrected axis command value θcnm-2, and the corrected axis command value θcnm−1. Proceed to S102. Specifically, the corrected position / orientation command value Pcnm is inversely converted into a temporary correction axis command value θt1nm, and the temporary correction axis command value θt1nm, the correction axis command value θcnm-2, and the correction are performed for each joint Jn. The average value with the axis command value θcnm−1 is set as the estimated correction axis command value θPnm.

  In step S102, the axis correction command value θcnm is set to the estimated correction axis command value θPnm. Further, in place of step S11 in FIG. 11, the correction axis command value θcnm of each joint Jn is stored in the memory 3 in step S103. This embodiment has the same effect as the second embodiment.

  In step S101, for each joint Jn, each correction axis command value θcnm-3 stored in the memory 3 at the three control timings Tm-3, Tm-2, Tm-1 immediately before the control timing Tm. The average value of θcnm−2 and θcnm−1 was used as the estimated correction axis command value θPnm. However, the present invention is not limited to this, and based on each correction axis command value stored in the memory 3 at at least one control timing immediately before the control timing Tm, for example, the first to fifth of the first embodiment. The estimated correction axis command value θPnm may be calculated by performing the same calculation as the estimated correction value calculation process according to the modified example. If YES is determined in the step S83, the provisional correction axis command value θt1nm at the control timing Tm and the correction axis command value θcnm stored in the memory 3 at at least one control timing immediately before the control timing Tm are used. Thus, the estimated correction axis command value θPnm may be calculated by the estimated correction value calculation process similar to the estimated correction value calculation process according to the first to fifth modifications of the first embodiment.

  In the above embodiments and modifications, the robot control apparatus 1 controls the operation of the six-axis articulated robot 7. However, the present invention is not limited to this, and a single-axis robot, a four-axis multi-joint robot, a five-axis multi-joint robot, or a seven-axis multi-point robot having a singular point that cannot reversely convert the corrected position / orientation command value into a unique corrected axis command value. You may control the operation | movement of the robot which a some link connects via at least 1 joint drive shaft, such as a joint robot.

  In each of the above embodiments and modifications, the controller 2 uses the simultaneous equation discriminant for inversely converting the corrected position / orientation command values Pc1m to Pc6m into the corrected axis instruction values θc1m to θc6m. It was determined whether or not the command values Pc1m to Pc6m can be inversely converted to the unique corrected axis command values θc1m to θc6m. However, the present invention is not limited to this, and whether or not the absolute value of the axis value of the joint J5 is a value near 0 degrees, or the position of the control point A in the XYZ coordinate system is within the movable range of the robot 7. The above determination may be made based on whether or not.

  Further, in each of the above-described embodiments and modifications, the controller 2 stores the control data from the computer 50 in the memory 3 and controls the operation of the robot 7 according to the stored control data. However, the present invention is not limited to this, and the control data may be read from the teach pendant 51 or the external storage device 52.

  Furthermore, in step S11 of the robot control process of FIGS. 3 and 11, the correction value Dnm of each joint Jn is stored in the memory 3, but the present invention is not limited to this, and in step S5, the corrected position / orientation command value Pc1m The correction value Dnm of each joint Jn may be stored in the memory 3 only when it is determined that .about.Pc6m can be inversely converted to the unique correction axis command values .theta.c1m to .theta.c6m. Further, in the robot control processing of FIGS. 3 and 11, when the corrected position / orientation command values Pc1m to Pc6m cannot be converted back to the unique corrected axis command values θc1m to θc6m, the control timing is three times immediately before the control timing Tm. Based on the correction value Dnm of each joint drive axis stored in the memory 3, the estimated correction value DPnm of each joint drive axis was calculated. However, the present invention is not limited to this, and when the corrected position / orientation command values Pc1m to Pc6m cannot be inversely converted into the unique corrected axis command values θc1m to θc6m, the memory 3 is stored in at least one control timing before the control timing Tm. The estimated correction value DPnm for each joint drive axis may be calculated based on the stored correction value Dnm for each joint drive axis.

  Further, in step S103 of the robot control process of FIG. 16, the correction axis command value θcnm of each joint Jn is stored in the memory 3. However, the present invention is not limited to this, and in step S5, the corrected position / orientation command values Pc1m to Pc6m are stored. The correction axis command value θcnm of each joint Jn may be stored in the memory 3 only when it is determined that can be inversely converted to the unique correction axis command values θc1m to θc6m. Further, in the robot control process of FIG. 16, when the corrected position / orientation command values Pc1m to Pc6m cannot be converted back into the unique correction axis command values θc1m to θc6m, the memory 3 is stored at the three control timings immediately before the control timing Tm. Based on the stored correction axis command value θcnm of each joint Jn, the estimated correction axis command value θPnm of each joint drive axis was calculated. However, the present invention is not limited to this, and when the corrected position / orientation command values Pc1m to Pc6m cannot be inversely converted into the unique corrected axis command values θc1m to θc6m, the memory 3 is stored in at least one control timing before the control timing Tm. The estimated correction axis command value θPnm for each joint drive axis may be calculated based on the stored correction axis command value θcnm for each joint Jn.

  As described above in detail, according to the robot control device and method of the present invention, it is determined whether or not the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft. When reverse conversion is possible, the correction position and orientation command value is converted back to each correction axis command value, and the correction value for each joint drive axis is calculated by subtracting each axis command value before correction from each correction axis command value. And when the reverse conversion cannot be performed, the estimated correction value of each joint drive axis based on the correction value of each joint drive axis stored in the storage means at at least one control timing before the control timing And calculating each correction axis command value by adding each estimated correction value to each axis command value before correction, and storing each estimated correction value in the storage means as a correction value for each joint drive axis, Each function according to each correction axis command value And controls to rotate the drive shaft, the singularity that can not inversely converted correction position and orientation command values to each unique correction axis command value, the positioning accuracy of the robot can be improved as compared with the prior art.

1 ... Robot controller,
2 ... Controller,
3 ... Memory,
4 ... Interface circuit,
5 ... Drive device,
6 ... Measuring device,
7 ... Robot,
9 ... the base,
10 ...
11-16 ... link,
17 ... hand,
50 ... computer,
50d ... display,
50i ... input section,
51 ... Teach pendant,
52. External storage device,
A: Control point,
J1-J6 ... Joints.

Claims (24)

In a robot control apparatus comprising a control means for controlling the position and posture of a control point of a robot formed by connecting a plurality of links via at least one joint drive shaft at every predetermined control timing,
The control means includes
Based on the position and orientation command value representing the position and orientation of the control point at each control timing, at each control timing,
(A) The position / orientation command value is inversely converted into each axis command value before correction representing the rotation amount of each joint drive shaft,
(B) correcting the position / orientation command value based on a robot-specific error in the actual position / orientation when a predetermined position / orientation is instructed, and calculating a corrected position / orientation command value;
(C) Determining whether or not the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft;
(C1) When reverse conversion is possible, each joint drive is performed by reversely converting the corrected position / orientation command value into the correction axis command values and subtracting the axis command values before correction from the correction axis command values. While calculating the axis correction value and storing it in the storage means,
(C2) When reverse conversion cannot be performed, an estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing; The corrected axis command values are calculated by adding the estimated correction values to the uncorrected axis command values, and the estimated correction values are stored in the storage means as correction values of the joint drive axes. ,
(D) A robot control device that controls to rotate each joint drive shaft according to each correction axis command value. When the control means cannot perform the reverse conversion,
An estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction value is calculated before the correction. In addition to calculating each correction axis command value by adding to each axis command value, and storing each estimated correction value as a correction value for each joint drive axis in the storage means,
An estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction value is calculated before the correction. 2. The robot control apparatus according to claim 1, wherein each of the corrected axis command values is calculated by adding to each of the axis command values. The control means includes
At the control timing at which the corrected position / orientation command value can be converted back to the respective corrected axis command values, the control timing is a predetermined number of times from the latest control timing at which the corrected position / attitude command value cannot be converted back to the corrected axis command values To determine whether it is within the control timing of
The control timing at which the corrected position / orientation command value can be converted back to the corrected axis command value is within a predetermined number of times from the latest control timing at which the corrected position / attitude command value cannot be converted back to the corrected axis command value. When the correction position / orientation command value is inversely converted, each axis command value is subtracted from each correction axis command value obtained by inversely converting the correction position / orientation command value, and at least one control immediately before the correction value of the subtraction result and the control timing. 3. The robot control apparatus according to claim 1, wherein an estimated correction value for each joint drive axis is calculated based on each correction value stored in the storage means at a timing.   The control means calculates an estimated correction value for each joint drive axis based on correction values for each joint drive axis stored in the storage means at a plurality of control timings before the control timing when the reverse conversion cannot be performed. 4. The robot control apparatus according to claim 1, wherein an average value of the correction values stored in the storage unit is used as the estimated correction value.   The control means calculates an estimated correction value for each joint drive axis based on correction values for each joint drive axis stored in the storage means at a plurality of control timings before the control timing when the reverse conversion cannot be performed. 4. The estimated correction value of each joint drive shaft is calculated based on each difference value of each correction value stored in the storage means. The robot control device described in one. The control means calculates an estimated correction value of each joint drive axis based on the correction value of each joint drive axis stored in the storage means at one control timing before the control timing when the reverse conversion cannot be performed. when, the correction value stored in the storage means, the robot control apparatus according to any one of claims 1 to 3, characterized in that the above guess correction value.   The control means calculates an estimated correction value for each joint drive axis based on correction values for each joint drive axis stored in the storage means at a plurality of control timings before the control timing when the reverse conversion cannot be performed. A predetermined function is applied to each correction value stored in the storage means, and an estimated correction value for each joint drive shaft is calculated using the applied function. The robot control device according to any one of 1 to 3. The control means calculates an estimated correction value for each joint drive axis based on correction values for each joint drive axis stored in the storage means at a plurality of control timings before the control timing when the reverse conversion cannot be performed. A weighted average value of each correction value is calculated using a predetermined weight set in advance for each correction value stored in the storage means, and the weighted average value is calculated as each estimated correction value. The robot control apparatus according to any one of claims 1 to 3, wherein the robot control apparatus is configured as follows. The robot control apparatus according to claim 4 , wherein the plurality of control timings before the control timing are a plurality of control timings immediately before the control timing. In a robot control apparatus comprising a control means for controlling the position and posture of a control point of a robot formed by connecting a plurality of links via at least one joint drive shaft at every predetermined control timing,
The control means includes
Based on the position and orientation command value representing the position and orientation of the control point at each control timing, at each control timing,
(A) correcting the position / orientation command value based on a robot-specific error in the actual position / orientation when a predetermined position / orientation is instructed, and calculating a corrected position / orientation command value;
( B ) determining whether the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft;
( B 1) When reverse conversion is possible, the correction position / orientation command value is reversely converted into the correction axis command values and the correction axis command values are stored in the storage means.
( B2 ) When the reverse conversion cannot be performed, each joint drive axis at the control timing is based on the corrected axis command value of each joint drive axis stored in the storage unit at at least one control timing before the control timing. And the correction axis command value is calculated and stored in the storage means,
( C ) A robot control device that controls the joint drive shafts to rotate according to the correction axis command values. When the control means cannot perform the reverse conversion,
Based on the correction axis command value of each joint drive axis stored in the storage means at at least one control timing before the control timing, the correction axis command value of each joint drive axis at the control timing is calculated and Instead of storing in storage means,
Calculating a correction axis command value for each joint drive axis at the control timing based on a correction axis command value for each joint drive axis stored in the storage means at at least one control timing before the control timing; 11. The robot control apparatus according to claim 10, wherein 12. The robot control apparatus according to claim 10, wherein the at least one control timing before the control timing is at least one control timing immediately before the control timing. In a robot control method including a control step of controlling the position and posture of a control point of a robot formed by connecting a plurality of links via at least one joint drive shaft at every predetermined control timing.
The above control steps are:
Based on the position and orientation command value representing the position and orientation of the control point at each control timing, at each control timing,
(A) The position / orientation command value is inversely converted into each axis command value before correction representing the rotation amount of each joint drive shaft,
(B) correcting the position / orientation command value based on a robot-specific error in the actual position / orientation when a predetermined position / orientation is instructed, and calculating a corrected position / orientation command value;
(C) Determining whether or not the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft;
(C1) When reverse conversion is possible, each joint drive is performed by reversely converting the corrected position / orientation command value into the correction axis command values and subtracting the axis command values before correction from the correction axis command values. While calculating the axis correction value and storing it in the storage means,
(C2) When reverse conversion cannot be performed, an estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing; The corrected axis command values are calculated by adding the estimated correction values to the uncorrected axis command values, and the estimated correction values are stored in the storage means as correction values of the joint drive axes. ,
(D) A robot control method characterized by controlling the joint drive shafts to rotate according to the correction axis command values. When the control step cannot perform the reverse conversion,
An estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction value is calculated before the correction. In addition to calculating each correction axis command value by adding to each axis command value, and storing each estimated correction value as a correction value for each joint drive axis in the storage means,
An estimated correction value for each joint drive axis is calculated based on the correction value for each joint drive axis stored in the storage means at at least one control timing before the control timing, and each estimated correction value is calculated before the correction. 14. The robot control method according to claim 13, wherein each of the corrected axis command values is calculated by adding to each of the axis command values. The above control steps are:
At the control timing at which the corrected position / orientation command value can be converted back to the respective corrected axis command values, the control timing is a predetermined number of times from the latest control timing at which the corrected position / attitude command value cannot be converted back to the corrected axis command values To determine whether it is within the control timing of
The control timing at which the corrected position / orientation command value can be converted back to the corrected axis command value is within a predetermined number of times from the latest control timing at which the corrected position / attitude command value cannot be converted back to the corrected axis command value. If it is, each axis command value is subtracted from each axis correction command value obtained by inversely converting the corrected position and orientation command value, and at least one control immediately before the correction value of the subtraction result and the control timing 15. The robot control method according to claim 13, wherein an estimated correction value for each joint drive axis is calculated based on each correction value stored in the storage means at a timing.   The control step calculates an estimated correction value of each joint drive axis based on the correction value of each joint drive axis stored in the storage unit at a plurality of control timings before the control timing when the reverse conversion cannot be performed. The robot control method according to claim 13, wherein an average value of each correction value stored in the storage unit is used as the estimated correction value.   The control step calculates an estimated correction value of each joint drive axis based on the correction value of each joint drive axis stored in the storage unit at a plurality of control timings before the control timing when the reverse conversion cannot be performed. The estimated correction value of each joint drive shaft is calculated based on each difference value of each correction value stored in the storage means. The robot control method described in one. The control step calculates an estimated correction value of each joint drive axis based on a correction value of each joint drive axis stored in the storage means at one control timing before the control timing when the reverse conversion cannot be performed. when, the correction value stored in the storage means, the robot control method according to any one of claims 13 to 15, characterized in that the above guess correction value.   The control step calculates an estimated correction value of each joint drive axis based on the correction value of each joint drive axis stored in the storage unit at a plurality of control timings before the control timing when the reverse conversion cannot be performed. A predetermined function is applied to each correction value stored in the storage means, and an estimated correction value for each joint drive shaft is calculated using the applied function. The robot control method according to any one of 13 to 15. The control step calculates an estimated correction value of each joint drive axis based on the correction value of each joint drive axis stored in the storage unit at a plurality of control timings before the control timing when the reverse conversion cannot be performed. A weighted average value of each correction value is calculated using a predetermined weight set in advance for each correction value stored in the storage means, and the weighted average value is calculated as each estimated correction value. The robot control method according to any one of claims 13 to 15, wherein: The robot control method according to any one of claims 16 to 20, wherein the plurality of control timings before the control timing are a plurality of control timings immediately before the control timing. In a robot control method including a control step of controlling the position and posture of a control point of a robot formed by connecting a plurality of links via at least one joint drive shaft at every predetermined control timing.
The above control steps are:
Based on the position and orientation command value representing the position and orientation of the control point at each control timing, at each control timing,
(A) correcting the position / orientation command value based on a robot-specific error in the actual position / orientation when a predetermined position / orientation is instructed, and calculating a corrected position / orientation command value;
( B ) determining whether the corrected position / orientation command value can be inversely converted into each corrected axis command value representing the rotation amount of each joint drive shaft;
( B 1) When reverse conversion is possible, the correction position / orientation command value is reversely converted into the correction axis command values and the correction axis command values are stored in the storage means.
( B2 ) When the reverse conversion cannot be performed, each joint drive axis at the control timing is based on the corrected axis command value of each joint drive axis stored in the storage unit at at least one control timing before the control timing. And the correction axis command value is calculated and stored in the storage means,
( C ) A robot control method comprising controlling the joint drive shafts to rotate according to the correction axis command values. When the control step cannot perform the reverse conversion,
Based on the correction axis command value of each joint drive axis stored in the storage means at at least one control timing before the control timing, the correction axis command value of each joint drive axis at the control timing is calculated and Instead of storing in storage means,
Calculating a correction axis command value for each joint drive axis at the control timing based on a correction axis command value for each joint drive axis stored in the storage means at at least one control timing before the control timing; 23. The robot control method according to claim 22, wherein 24. The robot control method according to claim 22, wherein the at least one control timing before the control timing is at least one control timing immediately before the control timing.

Images