JP2002082705A - Robot controller - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate track deviation caused in relation to an override or a temporary stop. SOLUTION: An operation planning part for software of a robot controller generates a plan for the operation of a robot without considering the override and outputs it to an interpolation processing part. The interpolation processing part performs interpolation processing in each computation processing cycle to calculate and output a movement quantity of each ITP to a filter part. The output of filtering performed by the filter part for acceleration/deceleration control is processed by an override processing part provided with a dynamic shutter. The outputs after processing applied with an override value β (0<=β<=1) common to respective axes are obtained by multiplying a speed and acceleration by β and β2. At a corner part where two operations overlap with each other, the acceleration and deceleration of both the operations are equally dealt with, so no track error is generated. Even when the override is changed during passage through the corner part, the same result is obtained. Further, the override is gradually varied in deceleration and acceleration before and after the temporary stop to eliminate even track deviation accompanying the temporary stop.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for controlling an industrial robot (hereinafter, simply referred to as "robot"). The present invention relates to a robot control device that is improved.

[0002]

2. Description of the Related Art When it is necessary to move a robot at a speed lower than a speed specified by an operation program (hereinafter, referred to as a "program speed"), the operation speed is set at a fixed rate to the program speed. Is widely used. For example, with respect to a newly created operation program or a modified operation program, a regeneration operation (test run) is performed by designating a low rate of override, and the movement locus is confirmed while avoiding danger.

[0003] The operation program whose trajectory is confirmed to be appropriate is reproduced in the control device of the robot in charge of the actual work. The override specified during the reproduction operation is, of course, the movement at the programmed speed. Is usually set to “100%” or a high rate close to “100%”.

However, when the conventional robot controller is used, there is a deviation between a trajectory realized when the moving speed is reduced by the override and a trajectory realized when the override 100% is designated (no speed reduction). Had occurred. This trajectory shift has the property of increasing as the designated override value decreases. In other words, if the robot approaches the corner during operation under low rate override,
Except in the case where positioning is performed at the corner, the trajectory of the tool tip point greatly changes depending on the specified override value in each axis, straight line, and circular motion.

[0005] Therefore, when checking the actual trajectory,
At present, an attempt is made to confirm an accurate trajectory by repeating a test run while gradually increasing the override, starting from an operation in which a low-rate override is designated. Such a trajectory shift occurs not only when a low rate override is specified,
It occurs in the same manner when the override value is changed during operation or before and after the robot temporarily stops (during deceleration and acceleration).

After all, in the prior art, after position teaching,
It took a long time to repeat the test run before operating at 100% override. In addition, when a pause process is performed during the program reproduction operation, the tool tip point position is reduced along with the deceleration, and the tool tip point locus drawn during normal work without performing the pause process (hereinafter, “normal locus”) And stop at a position that is not normally on the trajectory. In this specification, "pause"
It does not include "emergency stop", which means that the movement command output to the servo is instantaneously stopped by the alarm output.

Similarly, the trajectory of the tool tip point immediately after resuming the program reproduction after the pause is also off the normal trajectory, and requires a considerable moving distance until it coincides with the normal trajectory. Such a shift in the trajectory has also caused interference between an end effector such as a robot hand and an external object (a jig, a work, another installed device, or the like).

FIG. 1 is a diagram for explaining a framework of a conventional processing which is performed in a robot controller from an operation plan to a movement command output to a servo at the time of reproducing a program in order to consider the above-mentioned problem. is there. The framework of normal processing according to the conventional method is as shown in FIG.
The entire framework is roughly divided into an operation planning unit, an interpolation processing unit, and a filter unit, and each unit can be divided into a sequence for each axis operation and a sequence for orthogonal operation (linear operation, arc operation, etc.). .

The motion planning section is for creating a motion plan of the robot in accordance with the contents specified by the motion program (moving target position, motion format, program speed, etc.). The motion format specified by the motion program is for each axis motion. For example, an operation plan is created by each axis operation planning unit, and if the motion type specified by the operation program is orthogonal operation (linear operation, arc operation, etc.), the operation plan is created by the orthogonal operation planning unit.

The override is considered in the operation planning unit, and the product of the program speed multiplied by the override (percentage) is the command remote output to the interpolation processing unit. Between the operation planning unit and the interpolation processing unit, there is provided a block functioning as an operation shutter for controlling sending of the command speed output from the operation planning unit to the interpolation processing unit. The operation shutter is opened each time a commanded far output is generated, and passes the command far output to the interpolation processing unit.

In response to the command distance output, the interpolation processing section performs an interpolation process on each axis or a rectangular coordinate system at each calculation processing cycle (ITP), and calculates and outputs a movement amount for each ITP. You. In the subsequent filter unit, the output of the interpolation processing unit is filtered with a predetermined time constant. This process is performed to control the acceleration / deceleration of the robot and to smooth the acceleration / deceleration movement.

The outputs filtered by the respective axis filter processing units are also filtered, and then filtered for each axis (J1 to J6).
Is output to the servo. In the case of the orthogonal operation, a so-called inverse conversion process for performing conversion from the orthogonal coordinate system to each axis coordinate system is required. In the inverse conversion process, the inverse conversion is often performed before the filtering, but may be performed after the filtering.

[0013] When a temporary stop command is output inside the robot controller during the reproducing operation, the interpolation processing section is held by the inversion of the operation shutter, and the interrogation processing is interrupted. The output created by the interpolation process before the hold command is issued is filtered by the filter unit,
Output to servo. When the movement command output to the servo is consumed by the movement of each axis, the input to the servo is interrupted, and the robot stops.

When a path that smoothly connects the two operations is specified by the operation program, the two operation commands are processed in a superimposed manner. At this time, if the moving tangent direction at the end point of the preceding operation of the two operations is different from the moving tangent direction at the start point of the following operation, a rounded corner is generated. The overlapping amount of the motion at this time affects the inner shape of the corner.

Here, the relationship between the level of the override and the amount of overlap of the operation will be considered. When the override is low, the command speed naturally becomes low, but the time constant of the acceleration / deceleration processing of the filter does not change, so that the overlap amount of the operation becomes small. This means that the inward turning amount of the corner becomes small (that is, the curvature of the roundness becomes small). Conversely, when the override is high, the command speed naturally increases, and the overlap amount of the operation increases under the time constant of the acceleration / deceleration processing of the same filter. As a result, the inward turning amount of the corner increases (that is, the curvature of the roundness increases).

FIG. 2 is a diagram illustrating these phenomena in a simple example. In the example of FIG. 2, as the operation program, an operation of linearly moving at 2000 mm / sec toward point [1] and an operation of linearly moving and stopping (positioning) at 1000 mm / sec toward point [2] are defined as 100. Think of something that connects with% smoothness. In order to consider the phenomenon associated with the corner passing, as shown in the figure, it is assumed that the straight path toward the point [1] is in the −X-axis direction, and the straight path toward the point [2] is in the + Y-axis direction. The transition of the command speed output after the filtering process of each operation is also shown in the upper right part, and the symbols a and b generally represent the overlap of the operations.

Now, assuming that the high override is 100% and the low override is 50%, the overlaps a1 and b after the filter processing when the override is 100% are assumed.
1 and the overlaps a2 and b2 after the filtering process when the override is 50% are as shown in the drawing. That is, although the override is different by a factor of two, filtering is performed with the same time constant, so that there is a large difference in the amount of overlapping operation. As a result, the trajectories realized are different as shown by c1 and c2. The roundness of the corner is large in c1 and c2
Then it is small.

Next, in the case of 100% override, consider a case where a hold is applied due to a temporary stop at point P during a corner represented by spot [1]. From the time when the hold is applied, the output of the interpolation processing unit corresponding to the operation toward the position [2] is sharply reduced as compared with the case where the hold is not applied (indicated by a broken line d), and only the residual command output is output.
For this reason, an overlap between the two operations after the filtering process has an effect of rapidly decreasing during the movement of the corner.

As a result, the trajectory shifts from the point P corresponding to the time when the hold is applied to the corner representative point 1 [2], as shown by c3, and deviates from the trajectory c1 when the hold is not applied. Stop at Q (note that the action going to Ichi [1] is digested). When the reproduction is resumed after the pause, the robot trajectory follows the trajectory indicated by c4 from point Q (coincident with c1 from the middle).

It is apparent that a similar trajectory shift phenomenon also occurs when the override is changed during the passage through a corner. Because the filtering is performed under the same time constant, the amount of overlap between the two operations changes depending on the magnitude of the output of the interpolation processing unit.
This is because the trajectory changes.

In other words, in various cases in which the trajectory shift occurs (override level, switching during override operation, before and after temporary stop), the transition of the input level of the filtering process is different from that in the normal state. There are factors. Since the filtering process is performed under the same time constant even though the input level of the filtering process is changed, it can be considered that there is a difference in the operation path in a portion where different operations in two directions overlap. .

[0022]

SUMMARY OF THE INVENTION It is therefore an object of the present invention to provide a robot control apparatus which can eliminate a trajectory shift caused by the above-mentioned common factor in the prior art. That is, an object of the present invention is to provide a robot control device capable of preventing a decrease in trajectory accuracy before and after a change in override operation, a change during override operation, and a temporary stop.

The present invention is also intended to eliminate the deviation between the trajectory during the test run after teaching and the trajectory during operation under a high override, thereby improving the efficiency and safety of the trajectory checking operation. Is what you do. Further, the present invention aims to improve the accuracy and safety of the robot operation by eliminating the trajectory deviation due to the switching or the temporary stop during the operation of the override.

[0024]

SUMMARY OF THE INVENTION The present invention provides a robot control device equipped with software capable of eliminating a trajectory shift generated due to the above-mentioned factors, and a servo system of a servo motor for driving each axis of the robot. The present invention is applied to a robot control device having software means for outputting a movement command to a robot.

According to the present invention, the software means provided in the robot control device obtains the amount of movement of the robot to the target operation position, interpolates the obtained amount in a certain cycle, and accelerates / decelerates the interpolated output. It has a means for performing a filtering process for control, and a dynamic speed control means for inputting a movement command after the filtering process and correcting the movement command according to a specified command common override value for each axis.

The dynamic speed control means includes means for outputting, in response to the movement command, a value whose speed is proportional to the command override value and a value whose acceleration is proportional to the square of the command override value. .

Here, when the target override value is different from the current override value, the robot controller according to the present invention changes the command override value for the dynamic speed control means from the current override value to the target override value. It is possible to have a command override calculating means for changing smoothly.

Further, the robot control device according to the present invention sets the target override value to 0 when the robot is temporarily stopped at the time of executing the program, and sets the target override value to 0 by the command override calculation means. The robot can be configured to decelerate to a stop by smoothly changing the command override value from the current override value to 0.

Further, the robot control device according to the present invention sets both the target override value and the command override value to 0 and sets the movement command output to 0 instantaneously when the robot is to be emergency stopped during execution of the program. Alternatively, the robot may be configured to stop immediately.

On the other hand, in the case where the robot is temporarily stopped or an emergency stop is performed during the execution of the program, and the operation is resumed after the stop, the robot control device according to the present invention uses the command override calculation means to execute the command override calculation means. The robot can be configured to accelerate by smoothly changing the override value from 0 to the target override value.

Further, using the command override calculation means, the command override value for the dynamic speed control means can be changed stepwise by dividing into a specified number of divisions from a current override value to a target override value. You can also.

Here, regarding changing the command override value for the dynamic speed control means using the command override calculation means, a smooth change using an exponential function from the current override value to the target override value is performed. You may make it implement | achieve.

In the vicinity of the temporary stop,
The operation data may be stored in the robot controller, and the stored operation data may be used when the operation is restarted.

In the present invention, the operation plan based on the operation program may include an orthogonal operation plan for performing an orthogonal operation on at least one operation section. Further, the operation plan based on the operation program may include each axis operation plan for performing each axis operation for at least one operation section.

[0035]

FIG. 3 is a main block diagram illustrating the hardware configuration of a robot system including a robot controller to which the present invention is applied. As shown in the figure, the robot control device generally designated by reference numeral 30 controls a robot RB in which the hand 10 is attached to the arm tip 1 as an end effector. Here, the robot RB is a six-axis robot. End effector (hand 1
The tool tip point TCP which is a point representing the position of 0) is:
It is set at the center of the hand 10.

The robot controller 30 has a host CPU 3
1. Shared RAM 32, servo CPU 33, servo amplifier 34, memory 35, teaching operation panel interface 36
Also, an input / output device 38 for a general external device is provided.
The memory 35 stores the RO in which the system program is stored.
M, a RAM for temporarily storing data, and a system (a non-volatile memory in which various program data defining the operation of the robot RB are stored).

The teaching operation panel 37 connected to the teaching operation panel interface 36 is used for inputting, correcting, and registering program data, and for manual input of a manual feed ((jog feed) command, a regeneration operation command, and the like). In addition to the mechanical brake of the robot RB, various external devices (for example, a welding power supply device) corresponding to the application are connected to the external device input / output device 38.

Regeneration operation or manual feed (jog feed)
At the time of execution, the host CPU 31 creates a movement command for each axis of the robot RB and outputs it to the shared RAM 32. The servo CPU 33 reads this out at a predetermined short cycle, executes servo processing based on a position signal (feedback) sent from a position detector (pulse coder) of each axis of the robot, and supplies a current command to the servo amplifier 34 of each axis. And drives the servo motors of each axis of the robot. The servo CPU 33 periodically writes the current position of each axis of the robot to the shared RAM 32 based on a position signal (feedback) sent from a position detector (pulse coder) of each axis of the robot.

The configuration and function described above are not particularly different from those of a normal robot system. The difference between the robot controller 30 used in the present system and the conventional robot controller is that the processing steps from the reading of the program data by the host CPU 31 to the creation of the movement command for each axis of the robot RB are improved. This is to prevent the robot's trajectory deviation occurring due to the various factors (override level, switching during override operation, and processing before and after the temporary stop). The improved program data and related setting values are stored in the memory 35. Hereinafter, description will be made focusing on this improvement.

FIG. 4 is a diagram for explaining, in a form similar to FIG. 1, a framework of processing performed in the robot controller from an operation plan to a movement command output to a servo during program reproduction according to the embodiment of the present invention. is there.

The entire processing framework shown in FIG. 4 includes an operation planning unit, an interpolation processing unit, and a filter unit, each of which has a block for each axis operation and a block for orthogonal operation, as in the related art. Each unit can be divided into a series for each axis operation and a series for orthogonal operation (linear operation, circular arc operation, etc.).

The motion planning section is a portion for creating a motion plan of the robot in accordance with the contents specified by the motion program (moving target position, motion format, program speed, etc.). The motion format specified by the motion program is for each axis motion. For example, an operation plan is created by each axis operation planning unit, and if the motion type specified by the operation program is orthogonal operation (linear operation, arc operation, etc.), the operation plan is created by the orthogonal operation planning unit.

In the interpolation processing section, upon receiving the command distance output, the interpolation processing is performed on each axis or the orthogonal coordinate system at a predetermined calculation processing cycle (ITP), and the movement amount for each ITP is calculated. Is output. In the subsequent filter unit, the output of the interpolation processing unit is filtered with a predetermined time constant for acceleration / deceleration control. The output from each axis interpolation processing unit is filtered by each axis filter processing unit.

Although the above block configuration is common to the conventional one, the step of considering the override is basically different from the conventional one. That is, as described above, in the conventional processing, the override is considered in the processing of the operation plan, and the output after being multiplied by the override value is passed to the interpolation processing unit via the operation shutter. In the present embodiment, according to the features of the present invention, it is provided at the subsequent stage of the filter unit.

In accordance with this, the block of the operation shutter has also been moved to the subsequent stage of the filter section.

The override is considered in the override processing unit with a dynamic shutter (also referred to as a dynamic speed control unit; the same applies hereinafter). The output of the filter unit multiplied by the override (percentage) is calculated for each axis (J1). ~ J6) are distributed and output to the servos. The output from the orthogonal interpolation processing unit is
Filtering is performed by an orthogonal filter processing unit, and after being subjected to inverse transformation, distributed to and output to servos of each axis (J1 to J6). The output from the orthogonal interpolation processing unit is subjected to an inverse conversion process at the time of distribution to each axis.

When a temporary stop command is output inside the robot controller during the reproduction operation, the hold is applied not to the interpolation processing unit but to the override unit (output of the filter unit) due to the inversion of the operation shutter.

The processing which characterizes the present embodiment is divided into an override-related processing and a temporary stop-related processing, and the outline thereof is as follows.

(1) Outline of Override-Related Processing In this embodiment, as shown in the lower half of FIG. 4, the operation planning unit does not consider the override, and performs the interpolation processing for the input at the program speed. Done. The output after the interpolation processing is subjected to filtering processing in a subsequent filter unit under a predetermined time constant. At this stage, it is clear that there is no room for the process to depend on the override. Therefore,
Whether the override is high or low, or if the override rate is switched during operation, the output of the filter unit is not affected at all.

In the following override processing, the output of the filter section is modified so as to satisfy the following conditional expressions (1) and (2). Note that the override value is β% (0 ≦ β ≦ 10
0), the speed represented by the output of the filter unit is vflout, the acceleration is αflout, and the speed represented by the output after the override processing is v
ovout and acceleration are represented by αovout.

Αovout = αflout × (β / 100) 2 (1) vovout = vflout × (β / 100) (2) FIG. 5 shows that the processing of the override is applied to the output of the filter unit by 50%.
FIG. 7 is a diagram illustrating an example in which an override is applied. In the override processing, an instruction is divided and an output is generated while synchronizing all axes. Now, if the output of the filter unit is represented by seven segments sg1 to sg7 (seven movement commands), the segments sg1, sg2,.
Sg7 is divided into a total of 15 segments using 15ITP and output.

For example, sg1 is output in half at sh2, and is output at quarters at sh1 and sh3. Also, s
g4 is half output at sh8, and is 1/1/7 at sh7 and sh9.
Four are output. In general, each segment sh at the time of output, except for both end portions, has a movement amount derived from the segment sg of a plurality of filter unit outputs overlapping. Now s
The movement amounts corresponding to g1, sg2, and sg3 are vectors m
When represented by 1, m2, and m3, the output segmants sh1 to sh5 derived therefrom represent the moving amounts represented by the vectors n1 to n5 as shown. As the route, a locus with increased smoothness is obtained as the override is lower, but the basic line of the locus does not change before and after the override process.

The formula for calculating the movement amount for each ITP in the override processing (output segment = movement amount per ITP and corresponding to the instantaneous speed value) will be described later, but the main points are as follows.

1-1. A movement amount is calculated by multiplying each movement amount (input segment) of the output of the filter unit that has been subjected to the acceleration / deceleration processing by the override rate (β / 100).

1-2. 1-1. If the movement amount calculated in is too large (if output as it is, the above equation (1)
If the left side of (2) is too large to satisfy the equation, the excess is subtracted and output. In the first ITP, the above 1-1. With the movement amount calculated in the above, there is no case where the movement amount is too small to satisfy the above expressions (1) and (2) (because β ≦ 100, the left side of the above expressions (1) and (2) is too small). Never be.)

2. In the next ITP, a movement amount is calculated by multiplying the non-output movement amount again by the override rate (β / 100). However, if the amount of movement is insufficient (if the above equations (1) and (2) cannot be satisfied), the shortage is output. Insufficient outputs are selected with priority given to unoutputted old movement commands (input segments). In the second and subsequent ITPs, the amount of movement does not become excessive in the above calculation in which the amount of movement that has not been output is multiplied again by the override rate (β / 100).

3. Further, in the next ITP, for a movement command having a movement amount that has not been output, a movement multiplied again by the override rate (β / 100) is output. Then, when the movement amount is insufficient (when the above equations (1) and (2) cannot be satisfied), the shortage is added and output. Insufficient outputs are selected with priority given to unoutputted old movement commands (input segments). Hereinafter, the same processing is repeated, and when the entire movement amount of the entire movement command is output, the processing is ended.

In such processing, the movement amount used for acceleration, the movement amount of the acceleration unit, and the movement amount of the deceleration unit become substantially constant. This means that even when the override changes, the overlap amount when the two operations overlap is constant. Therefore, regardless of the level of the override, a substantially constant trajectory is realized even when passing through the corner.

FIG. 6 illustrates the effect of the override process when the operations represented by the segments sg1 to sg7 and the operations represented by the segments sk1 to sk7 overlap in a synchronous relationship between sg6 and sg7 and sk1 and sk2. FIG. Now, for example, override 50
%, As described in FIG. 5, the segment sg
1 to sg7 are output after being divided into sh1 to sh15. At this time, the movement amount represented by sg6 is sh11, s
The moving amount represented by sg7 is distributed to sh12, sh13, and sh15.

Similarly, the segments sk1 to sk7 are also represented by s
The output is divided into k1 to sk15. And sk1
Is moved to sq1, sq2, and sq3, and the movement amount represented by sk2 is sq3, sq4, and sq.
5 is distributed. When viewed after the override processing, the two operations overlap at sh11 to sh15 and sq1 to sq7.

As described above, the digestion speed of the movement amount at the time of the overlap of the movements is reduced equally in both operations (here, half), and the total movement amount to be digested is overridden by 50%.
Has not been changed by the process. Such a property does not change even if the override rate is other values.

Even when the override is switched during the operation, a constant locus can be similarly maintained by smoothly changing the division ratio of the segment of the filter unit output and performing the override processing according to the above example. It is.

In this case, the new and old override rates are set to (βN /
100) and (βO / 100), the acceleration is (βN /
100) 2 is changed to (βO / 100) 2 on the way. Therefore, the output of the override processing, which was to be output in the ITP immediately after the override change command, is further reduced to 1 / [β
N / (βO × 100)] times.

New and old override rates (βN / 100),
In order to avoid the effect of a sudden change in (βO / 100), it is preferable to gradually change the override rate from (βN / 100) to (βO / 100) in a stepwise manner. Therefore, in practice, a register for setting the target override value is set in the memory, and the user updates and inputs the new override value βN to this register, thereby changing the current override value to the target override value (here, Then, the process of asymptotically reaching the new override rate βN) is started.

In calculating the movement amount in each ITP, the old override target value (β) is used as an override value from the ITP immediately after the update of the override target value.
The movement amount is calculated by using a value obtained by gradually dividing the interval between the override target value (βN) of O) and n times (n ≧ 2).
For example, in the j-th ITP immediately after the update of the override target value, β = βO + jΔβ. Where Δβ =
(ΒN−βO) / n, j = 1, 2,..., N.

(2) Outline of temporary stop-related processing; FIG.
It is a figure for explaining temporary stop relevant processing in this embodiment. Here, as an example, sg1 to sg
With respect to the output of the filter unit represented by the 18 segments (the output after the override processing may be performed), the output to the servo up to the segment sg4 is completed, and the hold command is internally output by the ITP for the output of the segment sg5. Consider the case.

In this case, as shown in the figure, the data representing the movement amount after the segment sg5 is not output to the servo as it is, but is stored in a predetermined memory (a buffer preset in the memory 35 in FIG. 3). ) Is stored once. However, a predetermined time constant (here, 3ITP)
The segments sg5, sg6, and sg7 corresponding to are processed dynamically by the override processing unit, and the axes are decelerated and stopped eventually. As the dynamic processing for deceleration / stop, the override target value is set to 0 by applying the above-mentioned method of applying the override, and the override target value is gradually reduced to n which is the override target value n times (n ≧ 2). Can be adopted.

In the illustrated example, the segments sg5, sg
6, sg7 is distributed to sw5 to sw9. After all,
Servo has segments sg1 to sg4 (before hold)
Then, sw5 to sw9 are output, and the robot stops when the movement represented by these segments is completed. It should be noted here that sw5 representing the deceleration process is used.
The total amount of migration digested by ~ sw9 is segment sg5,
It is smaller than the total movement amount represented by sg6 and sg7.

The surplus movement amount is consumed during the acceleration process when the pause is released and the operation is resumed. That is, when the suspension is released, the data indicating the movement amount after the segment sg5 is read from the memory and output to the servo while accelerating to the original command speed.

In the acceleration process, the segments sg5, sg
6. Of the total movement amounts represented by sg7, those obtained by subtracting sw5 to sw9 digested in the deceleration process are digested. For this purpose, by applying the above-described method of applying the override, a process of setting the override target value to the override value before suspension (for example, 100) and gradually increasing the override target value n times to the override target value can be adopted. As the dynamic processing, a processing corresponding to gradually increasing the override to 100% in n times (n ≧ 2) by applying the above-described method of applying the override can be adopted.

After all, when the operation is resumed, sw10 is applied to the servo.
To sw15, and then sg8 and thereafter. Segments sw5 to s representing deceleration process and acceleration process
The total amount of movement digested at w15 is the segment sg5, sg
It becomes equal to the total movement amount represented by g6 and sg7.

By performing such processing before and after the temporary stop, the robot decelerates and stops while maintaining the trajectory when the hold command is output, and accelerates the robot while maintaining the trajectory from there when restarting. Can be shifted to the original operation.

At the time of hold, it goes without saying that the operation plan data, interpolation data, filter data, etc. at the time of deceleration stop are frozen and stored as they are in preparation for the restart of the operation. At the time of the resuming operation, the operation is started from a state where the override is set to 0, and the data stored at the time of the hold is restored. Thereafter, the override is gradually increased while smoothly synchronizing all axes again. As a result, it is possible to perform the deceleration stop and restart operations by changing the override while synchronizing all the axes while performing processing as if there was no hold up to the operation planning unit, the interpolation unit, and the filter unit. Will be possible.

When performing an emergency stop by an alarm, do not use the above-mentioned temporary stop method as it is.
Stop the output to the servo immediately and stop the robot as soon as possible. In the example of FIG. 6, sw6 to sw9 forming the deceleration process are not output to the servo.

In this case, since the trajectory is deviated due to the emergency stop, the trajectory may return to the original trajectory at a low rate of override (temporary positioning), and then accelerated by applying the above-described override changing method. Data for calculating the amount of movement required to return to the original trajectory (in the example of FIG. 6, data of sw6 to sw9 which have not been output to the servo segment due to an emergency stop), data of program speed, etc. Is stored in the memory near the point of emergency stop. Various data for determining the operation after returning to the original trajectory are also stored in the memory near the point of emergency stop.

The above is the outline of the embodiment. Below,
The processing of the dynamic speed control executed to avoid the deviation of the robot's trajectory will be described in more detail by dividing the processing into an override processing and a pause / operation restart processing. In the description, as an example of an override rate different from the override rate of 50% (β = 50), an override rate of 30% (β = 3)
Consider case 0).

FIG. 8 is a view for explaining, in a slightly different form from FIG. 4, a framework of processing performed in the robot controller from the operation plan to the output of a movement command to the servo during program reproduction according to the embodiment of the present invention. It is. In this drawing, while the division of the series for each axis operation and the series for the orthogonal operation (linear operation, arc operation, etc.) is omitted, the override processing unit with a dynamic shutter is described as a dynamic speed control unit, The memory used in connection with dynamic speed control has been simplified and added to the block display.

Referring to FIG. 8, the entire system includes an operation planning unit, an interpolation processing unit, a filter unit, and a dynamic speed control unit, and the output of the dynamic speed control unit is a servo system (motor) for each axis. Is displayed). In the processing flow, a memory is interposed on the input side and the output side of the dynamic speed control unit. A large number of register areas are prepared in advance in the memory so as to conform to the processing described below.

Each register area is also simply called a memory. Also, the n-th movement command after the filtering processing output from the filter unit is denoted by Dn. In the figure, it is described that the movement command Dn is stored in the memory Mn on the input side of the dynamic speed control unit. Each block described in a vertical line means one unit of memory (buffer register area). D1 in input side memory M1, D in M2
2, Mn stores Dn (n = 1, 2), respectively.

The output memory O1 of the dynamic speed control unit
It is illustrated that D2, D2 ', D1, and D1' are stored in ~ O4, respectively. Here, D2 ′ and D1 ′ are values obtained by multiplying D1 and D2 by an override value (β / 100), respectively. Details of how to use the input side and output side memories will be described in the following processing description.

[Override Processing] FIG. 9 is a flowchart illustrating the override processing performed by the dynamic speed control unit. The main points of each step are as follows. Note that i is a labeling index of the interpolation cycle, and the initial value is i = 1. A register for setting a current value β and a target value γ of the override is prepared, and a desired override target value γ0 is preset in the latter. As an initial value of the current value β of the override, a value that matches the initially set target value γ0 is automatically set.

(Step S1): It is checked whether the movement instructions Di and Di + 1 are stored in the memories M1 and M2, respectively. If not stored, the process proceeds to step S2, and if stored, the process proceeds to step S3.

(Step S2): Move instructions Di and Di + 1 are stored in memories M1 and M2, respectively, and move instructions Dm1 and
Dm2. Here, m1 and m2 are memories M1,
This is a label indicating that the data is stored in M2.

(Step S3): It is checked whether the value of the movement command Mm1 stored in the memory M1 is 0.
If yes, go to step S4; if no, go to step S5.

(Step S4): The movement command Dm2 stored in the memory M2 is copied to the memory M1, and Dm1
(Update of the value of the movement command Dm1). And Di +
The next dynamic command Di + 2 is read out, stored in the memory M2, and set as Dm2 (update of the value of Dm2).

(Step S5): The current value β% of the override is determined, and one interpolation per interpolation (1 IT) is performed under the determined β value.
Movement amount (per P) (Segment height indicating movement amount)
Di is calculated. The method of calculating Di will be described later. The override current value β may change during operation if the target value γ changes. The value of β changes during operation because of user input, change by external signal, pause (hold)
Time, a temporary stop (hold) or a restart of operation after an emergency stop. The process of determining the current value β% of the override will be described later together with the process of temporarily stopping (holding) and subsequently resuming operation.

(Step S6): Di calculated in step S5 and movement command Dm1 stored in memory M1
Compare with If Dm1-Di> 0, step S
8; otherwise (Dm1-Di ≧ 0), the process proceeds to step S7.

(Step S7): The movement command Dm1 stored in the memory M1 is copied to the memory O1 on the output side. Then, using these Dm1 and Dm2, Di = Dm
The value μ that satisfies 1 + Dm2 * μ is obtained. From this μ, Di ′ satisfying Di ′ = Dm2 * μ is output to the output memory O1.
Add to Also, Dm2 is calculated as Dm2 = Dm2-Dm2 *
μ, and copy it to memory M1 to create a new Dm
1 (update of Dm1).

(Step S8): The movement command Di determined in step S5 is written to the output side memory O1. Also, the movement command Dm1 stored in the memory M1 is updated to a new Dm1 = Dm1-Di.

(Step S9): The movement command Do of the output side memory is output to the motor (each axis servo system).

(Step S10): Hold, temporary stop, emergency stop, operation end command (issued when a new Di output does not come from the filter unit and the processing in step S2 cannot be performed), etc. Check whether the command is output inside the robot controller. If yes, execute the process for them (described below)
Stop the robot. If no, step S11
Proceed to.

(Step S11): Index value i of movement command
And returns to step S1.

Here, when the override (current value) in step S5 is β%, one interpolation (1I
The method of calculating the movement amount Di ′ (per TP) will be described.

The following formula is used to calculate the movement amount Di 'so that the speed change can be smoothly obtained even when the movement command Di output from the filter unit is not constant.

The movement command Di 'per i-th one interpolation time (ITP) t0 is represented by the following equation (3).

Di '= V (i) * t0 (3) Here, V (i) is the speed when the override in the i-th ITP is 100%. The acceleration A (i) per ITP when the override in the i-th ITP is set to 100% is temporarily expressed by the following equation (4).

A (i) = V (i) −V (i−1) (4) Using A (i) to smooth the speed transition,
The following equation (5) is adopted as a reference equation of the acceleration output in the i-th ITP at present.

Aout (i) = − A (i−1) / 2 + 3 * A (i) / 2 (5) Using this, the following equation (6) is used as the speed output at the i-th ITP at present. adopt.

Vout (i) = OVR * [V (i) + Aout (i) * (δ + OVR / 2−1 / 2) (6) where OVR = β / 100 and δ is V (i) Δ = 1 when the output of is output to the servo system.
If the output of (i) has not been fully output to the servo system, δ
= 0.

For example, when the override is 100%, Vout (i) = 1 * [V (i) + Aout (i)
* (0 + 1 / 2−1 / 2) = V (i), a natural result is obtained. When the override is 50%, Vout (i) = 0.5 * [V (i) + Aout (i)
* (Δ + 0.5 / 2-1 / 2) = V (i) / 2 + Aou
t (i) / 2 * (δ- ／), when δ = 1; V (i) / 2−Aout (i) / 8 When δ = 0; V (i) / 2 + Aout (i ) / 8.

The influence of the level of the override value on the trajectory when the path operation including the corner portion is performed while performing the above-described processing will be described with reference to FIGS. Will be explained.
As a path operation including a corner portion, as shown in FIG. 14, a linear operation from the position G1 to the position G2 and a linear operation from the position G2 to the position G3 are smoothly connected. The program speeds of both operations are assumed to be equal.

The crosses indicate the teaching points G1, G2, G3.
An asterisk indicates the position of the interpolation point (interpolation position) corresponding to the movement command finally output to the servo system. The corners are indicated by the symbol A. On the other hand, a range B represents a section of acceleration for starting and a section of deceleration for stopping (positioning).

In the case of this example, at the corner portion A, the marks are arranged at equal intervals at a constant speed, but in the acceleration / deceleration section B, the intervals of the marks are gradually enlarged or reduced. Signs q1, q2,
q3 is an interpolation point representing a locus in the corner portion A, and forms a moving section corresponding to the outputs D1 and D2 from the filter portion in the corner portion.

Symbols p1, p2, p3, and p4 are interpolation points representing trajectories in the acceleration section B, and form moving sections corresponding to outputs D1 and D2 from the filter section in the acceleration / deceleration section. . Note that the number of interpolation points (★ marks) actually generated is much larger, but is simply illustrated in a small number for display.

Referring to FIG. 15, as shown in the upper left part of the figure, the trajectory of the corner A when the robot is operated with the override 100% is as shown in FIG.
Interpolation points q corresponding to outputs D1 and D2 from the filter unit
A trajectory passing through 1, q2, q3 is obtained.

On the other hand, the locus of the corner A when the robot is operated with the override of 30% is as shown in the upper right part of FIG. By the above-described override processing (β = 30), outputs D1 and D
The interpolation points q1, q2, q3 corresponding to 2 are converted into a larger number of interpolation points qd1 to qd9 (signs are partially shown). The calculation formula for each section is as shown in the figure.

It is important to note that these interpolation points qd1 to qd1
qd9 means that the vehicle is riding on the trajectory at the time of the original override of 100%. This property is 30% override rate
If not, it is guaranteed as well. In general, a number of interpolation points that are inversely proportional to the override value are output on the basis of the interpolation point when the override value is 100%, and are formed so as to be arranged on the locus when the override value is 100%. Therefore, the same trajectory can be obtained in the corner portion regardless of the level of the override value.

In the lower left and lower right portions of FIG. 15, the manner in which the movement commands D1 and D2 at the output stage of the filter section are divided and output is displayed in segments. When the vehicle passes through the corner at a constant speed as in this example, the height of each segment is constant. When the program speed of the two operations for forming the corner portion is different, the height of each segment naturally changes stepwise. Even in that case, the trajectory created by the interpolation points is not affected by the override rate.

Next, taking the acceleration section as an example, the trajectory in the acceleration / deceleration section will be considered. Referring to FIG. 16, as shown in the upper left part of the figure, the trajectory of the acceleration / deceleration unit B (represented by the acceleration unit here) when the robot is operated with the override of 100% is as shown in FIG. 14. , P1, p2 corresponding to the outputs D1, D2, D3 from the filter unit
Acceleration toward the program speed is performed along a locus passing through 2, p3, and p4.

The trajectory of the acceleration section B when the robot is operated with the override of 30% is not different from the case of the override of 100% as shown in the upper right part of FIG. However,
By the above-described override processing (β = 30), the interpolation points p corresponding to the outputs D1, D2, and D3 from the filter unit are obtained.
1, p2, p3, p4 are more interpolation points pd1-q
It is converted to d11 (signs are partially shown). The calculation formula for each section is as shown in the figure.

Such a property has an override rate of 30%
If not, it is guaranteed as well. In general, a number of interpolation points that are inversely proportional to the override value are output with reference to the interpolation point when the override value is 100%, and are formed so as to be arranged on the locus when the override value is 100%.

In the lower left and lower right portions of FIG. 15, the manner in which the movement commands D1, D2, and D3 in the filter section output stage are divided and output is displayed in segments. In the acceleration / deceleration section, the height of each segment changes stepwise. This change represents a process of reducing the speed at 100% override to 30% (generally β%) and smoothly accelerating. The final reaching speed of acceleration is naturally 30% (generally β%) of the program speed.

When the override is changed during the operation of the robot, the override asymptotic / arrival processing shown in the flowchart of FIG. 17 is performed. This process is also used in a hold process and an operation restart process described later.

The essential points of each step of the override asymptotic / arrival processing are as follows. (Step L1); The override target value is updated to the new value βN. (Step L2); I next to update of override target value
Wait for the fth ITP from the TP. f is a positive integer value set to an appropriate size. (Step L3); Override current value from β to β + Δ
Update to β. Δβ is determined in advance, for example, as Δβ = 1 (%), or calculated by the following equation (7).

Δβ = (βN−β just before update of the override target value) / n (7) where n is the number of updates required until the override current value matches the target value, for example, n = 10, etc. Or separately determined from conditions (see the description of the hold processing described later). By performing such processing, β in the flowchart of FIG. 9 is changed stepwise toward a new target value.

[Hold Processing] FIG. 10 is a flowchart for explaining the hold processing performed by the dynamic speed control unit. The hold process is executed, for example, when the operator inputs a temporary stop command to the robot controller. The main points of each step are as follows.

(Step H1): It is checked for each ITP whether a hold command has been output inside the robot controller. If output, the process proceeds to step H2. (Step H2): From the current speed V in the dynamic control unit,
The deceleration override value OVd is controlled so that the speed becomes zero smoothly after the elapse of the time t1.

(Step H3): It is checked for each ITP whether or not the robot has stopped. When the stop is confirmed, the process proceeds to step H4.

(Step H4): The current position (stop position) at that time, the data of the movement command in the filter, and other data relating to the operation of the robot are stored in a memory (a register for storing at the time of holding), and The process ends. For the control of the deceleration override value OVd in the process of step H2, the override asymptotic / arrival process shown in the flowchart of FIG. 17 is used.

That is, the update target value βN of the override in step L1 is set to βN = 0. Further, s1 is divided by η (η is an appropriate positive integer other than 1), where s1 is the ITP conversion value of time t1, and f in step L2 is f = s1
/ Η. Then, n in the equation (7) for calculating Δβ in step L3 is calculated by βN = 0 and n = η.

FIG. 12 shows the speed transition when the above-mentioned hold processing is performed. The override gradually goes to 0, but the movement command needs to be completely output to the servo system in order to completely reflect the actual speed. Therefore, the speed transition does not drop linearly. Further, in order to make the operation smoother, the value of Δβ may not be a fixed value but may be gradually changed according to a smooth function (for example, an exponential function) to obtain a smooth transition of acceleration.

[Process of Resuming Operation] FIG. 11 is a flowchart illustrating the process of resuming operation performed by the dynamic speed control unit. The operation resumption processing is executed when, for example, an operation resumption command is input to the robot controller after the suspension. The main points of each step are as follows.

(Step H11): The stop position stored in the memory is read. (Step H12): The robot is operated from the current robot position (for example, moving by jog feed) to the stop position read in Step H11. (Step H13): Confirmation of completion of the operation in Step H12. If confirmed, the process proceeds to Step H14. (Step H14); Data necessary for the restart operation is read out from the memory, read out, returned to the operation planning unit, the interpolation unit, and the filter unit, and the state immediately before the hold is restored.

(Step H15) From the stop position, the interpolation operation is performed in the same manner as before the hold. Also, at the same time, the dynamic control unit smoothly changes the speed from the current override value (β = 0) to the original speed v after the elapse of the time t2,
The acceleration override value OVa is controlled. Step H15
The control of the acceleration override value OVa in the processing of
The override asymptotic / arrival processing shown in the flowchart of FIG. 17 is used.

That is, the update target value βN of the override in step L1 is set to βN = β immediately before the hold. Further, s2 is divided into η (η is an appropriate positive integer other than 1) with s2 as the ITP conversion value of time t2, and step L2
Is f = s2 / η. And step L
In the equation (7) for calculating .DELTA..beta. In (3), n is calculated by .beta.N = the value of .beta.

FIG. 13 shows the transition of the speed when the processing of the above step H15 is performed. The override gradually goes from 0 to the original value, but it is necessary to completely output the movement command to the servo system in order to reflect the actual speed. Therefore, the speed transition does not rise linearly. Further, in order to make the operation smoother, the value of Δβ may not be a fixed value but may be gradually changed according to a smooth function (for example, an exponential function) to obtain a smooth transition of acceleration.

[0127]

According to the present invention, it is possible to prevent a trajectory shift which has occurred in the prior art in connection with the level of the override, the switching during the operation of the override, or the temporary stop. Further, through this, it is possible to improve the efficiency, reliability, and safety of the trajectory checking operation by the test run after teaching. Further, since the trajectory deviation due to the switching or the temporary stop during the operation of the override is eliminated, the accuracy and safety of the robot operation are improved.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a framework of processing of a conventional method which has been performed inside a robot controller from an operation plan to a movement command output to a servo at the time of program reproduction.

FIG. 2 is a diagram for explaining the reason why a trajectory shift occurs using a simple example in relation to the processing of the conventional method shown in FIG. 1;

FIG. 3 is a main block diagram illustrating a hardware configuration of a robot system including a robot control device to which the present invention is applied;

FIG. 4 is a diagram illustrating, in a form similar to FIG. 1, a framework of processing performed in the robot controller from an operation plan to a movement command output to a servo when a program is reproduced, according to the embodiment of the present invention;

FIG. 5 is a diagram showing an example in which an override process is applied to a filter unit output by 50%.
FIG. 7 is a diagram illustrating an example in which an override is applied.

FIG. 6 shows an operation represented by segments sg1 to sg7 and an operation represented by segments sk1 to sk7,
It is a figure explaining the influence of an override process, when g6 and sg7 and sk1 and sk2 overlap in a synchronous relationship.

FIG. 7 is a diagram for describing temporary stop related processing in the present embodiment.

FIG. 8 is a diagram illustrating, in a form slightly different from FIG. 4, a framework of processing performed in the robot controller from an operation plan to a movement command output to a servo during program reproduction, according to the embodiment of the present invention;

FIG. 9 is a flowchart illustrating a point of an override process performed by a dynamic speed control unit.

FIG. 10 is a flowchart illustrating a main point of a hold process executed at the time of a temporary stop in the dynamic speed control unit.

FIG. 11 is a flowchart illustrating a main point of a process executed by the dynamic speed control unit when the operation is resumed after the suspension.

FIG. 12 is a diagram showing a change in speed when a hold process is performed.

FIG. 13 is a diagram showing a speed transition when a process of gradually increasing the override in resuming the operation is performed.

FIG. 14 is a diagram showing a route operation assumed as a premise for the description of FIGS. 15 and 16;

FIG. 15 is a diagram illustrating an influence of a level of an override value on a trajectory in a corner portion of the path operation illustrated in FIG. 14, using an example of an override rate of 30%.

16 is a diagram illustrating the influence of the level of the override value on the trajectory in the acceleration / deceleration unit in the path operation illustrated in FIG. 14, using a case of an override rate of 30% as an example.

FIG. 17 is a flowchart illustrating override asymptotic / arrival processing.

[Explanation of symbols]

 Reference Signs List 1 arm tip 10 hand 30 robot controller 31 host CPU 32 shared RAM 33 servo CPU 34 servo amplifier 35 memory 36 teaching operation panel interface 38 input / output device 37 teaching operation panel RB robot

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Atsuro Nagayama 3580 Kobaba, Oshino-mura, Oshino-mura, Minamitsuru-gun Yamanashi Prefecture Inside FANUC Co., Ltd. FANUC Corporation F-term (reference) 3C007 LT01 LU03 LU05 MT04 5H269 AB33 BB03 EE01 EE16 EE25

Claims (10)

[Claims] 1. A robot controller comprising software means for outputting a movement command to a servomotor for driving each axis of a robot, wherein the software means calculates a movement amount of the robot to an operation target position. Means for performing an interpolation process at a certain cycle, and filtering the interpolated output for acceleration / deceleration control; and inputting the movement command after the filtering process as a command override value common to each designated axis. Dynamic speed control means for correcting the movement command according to the speed of the movement command, the dynamic speed control means, for the movement command, the speed is a value proportional to the command override value, and the acceleration is the command override value The robot control device according to claim 1, further comprising means for outputting a value proportional to the square. 2. A command override calculating means for smoothly changing a command override value for the dynamic speed control means from the current override value to the target override value when the target override value is different from the current override value. The robot controller according to claim 1, wherein: 3. When the robot is temporarily stopped during execution of a program, the target override value is set to 0,
3. The robot controller according to claim 2, wherein the command override calculating means smoothly decelerates and stops the robot by changing a command override value for the dynamic speed control means from the current override value to zero. . 4. An emergency stop of the robot during execution of a program, wherein the target override value and the command override value are both set to 0, the movement command output is instantaneously set to 0, and the robot is immediately stopped. The robot control device according to claim 2, which performs the operation. 5. When a robot is temporarily stopped or an emergency stop is performed during execution of a program, and the operation is restarted after the stop, the command override calculating means smoothly changes the command override value from 0 to the target override value. The robot controller according to claim 2, wherein the robot is accelerated by changing. 6. The command override calculating means, wherein the command override value for the dynamic speed control means is changed stepwise by dividing into a specified number of divisions from a current override value to a target override value. The robot controller according to any one of claims 2 to 5, wherein 7. The command override calculating means smoothly changes a command override value to the dynamic speed control means by using an exponential function from a current override value to a target override value. The robot controller according to any one of claims 2 to 5. 8. The apparatus according to claim 3, wherein the operation data is stored in the robot controller near the time of the temporary stop, and the stored operation data is used when the operation is resumed. Item 6. The robot control device according to any one of items 5. 9. The operation plan according to claim 1, wherein the operation plan based on the operation program includes an orthogonal operation plan for performing an orthogonal operation for at least one operation section. 2. The robot controller according to claim 1. 10. The operation plan according to claim 1, wherein said operation plan based on said operation program includes each axis operation plan for performing each axis operation for at least one operation section. The robot control device according to claim 1.