JPH0469706A - Industrial robot controller - Google Patents

Abstract

PURPOSE:To easily obtain the control of a robot executed by a smooth operation and with high operation accuracy by a simple traching operation by setting in advance several transfer systems required in a teaching point, and selecting one of these transfer systems by a prescribed mark added to the teaching point. CONSTITUTION:In an initialization processing, a processing program stored in a bubble memory 35 is loaded onto a RAM-A 32, initialization of each apparatus is executed, and subsequently, a judgement of a mode is executed. The judgement of this mode is selected by a changeover switch on a play-back console (PBC) 24. Next, in the case of a teaching mode, a prescribed mark is added to teaching point data inputted by teaching, and in at least two kinds of different transfer systems, one kind of the transfer systems is selected in accordance with the mark put to the teaching point data. In such a way, the teaching operation can be simplified.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a teaching-playback type robot, and particularly to a control device suitable for a welding robot.

[Prior Art] In a teaching-playback robot using a manipulator having at least two degrees of freedom, it is possible to control of transfer routes;
That is, route interpolation control is necessary.

To explain such a robot control device using the teaching/playback method, first, the robot is moved, multiple points are determined on the trajectory where the work point (usually the robot's hand) should move, and those points are is stored on the robot control device as a teaching point. At this time, information on how to interpolate between teaching points, information on how to move at what speed, and information on the robot's work, such as opening and closing of the hand, start of welding, and end of welding, etc. is also memorized at the same time. The above is teaching.

Next, the robot control device controls the robot's degrees of freedom based on the above-mentioned stored information, and moves the work point at a specified speed along the initially determined trajectory. This is playback.

Here, for information on how to move the robot between each teaching point at high speed and smoothly, please refer to Paza MIT Press, Robot Manipulator (1981)' p130~ p
133 (The HIT Press Robot M
anipulators) PA JP-A-177286
This is discussed in detail in Japanese Patent Application Laid-Open No. 62-84983.

In these documents, a transfer route is newly generated near the robot's teaching point, and the robot's actual movement path, actual movement speed, and actual movement acceleration are optimized to increase the robot's movement speed. I try not to go below the specified speed.

[Problem to be solved by the invention] The above-mentioned conventional technology does not take into account the simplification of the teaching work, and it is difficult to make the robot operate smoothly.
In addition, in order to maintain the desired trajectory accuracy, it is necessary to set a wide variety of parameters for each teaching point, which poses the problem of requiring extremely complicated operations.

An object of the present invention is to provide an industrial robot control device that can easily control a robot with smooth motion and high precision through simple teaching operations.

[Means for solving the problem] In order to achieve the above purpose, in advance, necessary at the teaching point,
Several transfer methods are set, and one of these transfer methods is selected by a predetermined mark added to the teaching point.

Here, the transfer method employed in the present invention will be explained.

First of all, since there is no object without mass, in general, in order to change the path of an object moving in a straight line at a constant velocity, it is necessary to apply a force to the object for a predetermined amount of time. .

By the way, as a matter of course, robots also have mass, so they are included in this category.

So, if you try to change the direction of a robot at a single point, you will need an infinite amount of force at this point. However, it goes without saying that such a motor (actuator) should never exist.

For this reason, in actual robot control, a predetermined force is applied over a predetermined period of time, and the path changes by drawing a predetermined curve at corners.

Now, if we can approximate the mass of the robot's movable parts as concentrated in its hands, the following equation holds.

fFmaxf)d+=M(V, -V,)---(1
) Here, ■ 2 Velocity vector before transfer v2; Speed vector M after transfer: Approximate mass Fmax(t); Torque (force) that can be generated. This integral value is the velocity vector V before transfer and the speed vector after transfer. Since it is determined by the velocity vector v2 and the approximate mass M, this value does not change no matter what transfer method is used.

Therefore, the following three transfer methods can be considered.

Transfer method ■ The simplest method, where the speed vector before transfer is O
Torque pmax(t) is applied in the opposite direction to this speed vector v1 until the speed vector v1 reaches 2, and then the speed vector v2 is accelerated from 0 to the speed vector v2.

Note that this method takes twice as long as it would take if the vehicle simply traversed the route at a constant speed.

Transfer method ■ This is the most efficient method, and the torque pmax(t) is always applied in the direction of the vector (V, -V, ). Note that in this method, the route becomes a parabolic shape.

Transfer method ■ This is a method that can minimize changes in linear speed (absolute value of moving speed), and torque Fmax(t) is always applied perpendicularly to the path. Note that in this case, the route draws an arc.

And, Fmax=M r tJ=M11'/r Therefore, the radius r of this circular arc becomes r=MV2/Fmax.

Note that both the transfer methods (1) and (2) perform smooth operation, but the moving route does not pass through the teaching point.

[Function] During playback, a predetermined mark added to a teaching point automatically controls the transfer path necessary for the robot at that teaching point, so during teaching, inputting the teaching point and the predetermined mark The teaching operation can be easily simplified without sacrificing smooth movement or accuracy.

[Example] Hereinafter, regarding the industrial robot control device according to the present invention,
This will be explained in detail with reference to the illustrated embodiment.

Figures 1 and 2 show an example in which the present invention is applied to a welding robot that performs arc welding work.
The figure shows the overall equipment configuration, and in the figure, 20 is a robot control device that controls all the devices. The main processing of the present invention is also carried out by this robot control device 20.
It will be held in

2] is a robot body (manipulator), and in this embodiment, a robot with a six-axis configuration is used, and is driven by six servo motors. A welding torch 29 is attached to the tip of the robot's wrist.

A welding machine 22 generates welding current and voltage based on commands from the control device 20, and also functions to give a wire feeding command to the wire feeding device 28.

A programming unit (PGU) 23 is provided with a liquid crystal display and various keys, and the PGU 23 is used to teach the robot its posture, operating conditions, and the like.

24 is a playback console (PBC), and it is this PBC that starts automatic operation of the robot and switches to teaching mode.
This is done by the switch on 24.

25 is a gas cylinder that supplies shielding residue for welding, and 26 is a welding wire supply device.

A floppy disk driver 30 is used to save teaching data and to load teaching data created by another device into the robot control device 20.

Reference numeral 27 is a target workpiece, which is to be welded using the equipment described above.

Next, the inside of the robot control device 20 and the robot body 21 will be explained in more detail using FIG. 6.

The inside of the robot control device 20 is roughly divided into three parts: a main CPU section 20a, a servo CPU section 20b, and a servo amplifier 20c.

The main CPU section 20a controls peripheral devices, man-machine interface, automatic driving sequence control, and the like. The main processing of the present invention also takes place here.

Next, the servo CPU section 20b controls the operation of the robot.

The servo amplifier 20c functions to supply drive current to the servo motor within the robot body 21.

Next, each of these parts will be explained in more detail.

First, the inside of the main CPU section 20a will be explained.The main CPU section 20a is an ICPU-A, which controls each device according to a program written in a ROM or RAM.

31 is a ROM-A, which stores a program for initialization performed by the CPU-A 30 when the power is turned on.

32 is a RAM-A, which stores processing programs other than initialization and intermediate results of calculations and the like executed by the CPU-A 30. And this RA, M-A32 has R
A processing program is loaded from the bubble memory 35 by the program of the OM-A 31 when the power is turned on.

33 is a communication interface, CPU-A30 is a PG
When exchanging information with U23 and PBC24, it is done through here.

A welding machine interface 34 is used to send various commands to the welding machine 22 and to read the status of the welding machine 22. As mentioned above, 35 is a bubble memory, which is a non-volatile auxiliary storage device, and R
It is used to store information that should not be lost even when the power is turned off, such as processing programs and teaching data to be loaded into the AM-A32.

36 is a dual port RAM, main CPU fJ2
It is used for exchanging information between 0a and the servo CPU section 20b.

45 is a floppy disk interface, which interfaces with the floppy disk driver 30; Here, the floppy disk is used to back up teaching data and to load teaching data taught by another device into this control device.

A data bus 43 interconnects the devices described above.

Next, the servo CPU section 20b will be explained.
is CPU-B, which controls the operation of the robot.

38 is a ROM-B in which a processing program to be executed by the CPU-B 57 is stored.

39 is RAM-B, which is used for temporary storage of intermediate results of calculations, etc.

40 is a timer, which is set for a certain period of time (sampling period)
It functions to interrupt the CPU-B57 every time.

4] is a D/A conversion converter, in which the current command for each servo motor is converted from a digital quantity to an analog quantity and sent to the servo amplifier 20c.

A counter 42 detects pulses from an encoder (described later) and the rotation angle of each servo motor.

44 is a data bus, which connects each of the devices mentioned above.

Finally, the servo amplifier 20c has six amplifiers built-in, and the motor part 21a of the robot M]
Generates current to drive six motors from M6 to M6.

21b is an encoder section, and represents encoders E1 to E6 attached to each of the motors M1 to M6.

In addition, each of these motors M1 to MG and encoders E1 to E6 respectively correspond to the rotation axis of the robot body 21,
One is attached to the arm shaft, forearm shaft, rotation shaft, bending shaft, and twisting shaft.

Next, we will explain the actual processing.

Figure 7 shows the overall processing flow, and this flow is CPIJ-
A.30.

First, when the power is turned on, initialization processing is performed in process 100. In this initialization process, the processing program stored in the bubble memory 35 is loaded into the MA 321, and then each device is initialized.

Next, in process 01, the mode is determined. This mode is selected by a changeover switch on the PBC 24. That is, the CPU-A 30 uses the communication interface 33 to read the status of the switch on the PBC 24 . Then, when the mode is parameter setting, the parameter setting process of process 102 is executed, and the PGU
23" Outputs a message prompting the second user to enter a parameter, and prompts the user to enter the value using the key. The following values are input at this time.

Maximum allowable error while not working (during air cutting): Da (mm) Maximum allowable error while working (during welding) D w (mm) These values are stored in the RAM-A32 and at the same time in the bubble memory. It is also stored in 35.

This completes the parameter setting process.

Next, when the mode is teaching, a teaching process 103 is performed, and then a transfer process 104, which is a characteristic and important process of the present invention, is performed, and the process in the teaching mode is ended. Note that this process 104 will be explained in detail later.

Finally, if the mode is automatic operation, automatic operation processing of process 105 is performed. A detailed explanation of this process will be given later.

Next, the teaching process will be explained using FIGS. 8 to 0.

Figure 8 shows the workpiece 27 that is about to be welded and the teaching points for welding work.
The explanation will be based on the welding work shown in FIG. 8.

FIG. 9 shows the flow of the teaching process.
input. This program number is PCI, 723
This is used to identify the teaching data to be entered and taught using the key, and in this embodiment, program numbers from 1 to 100 can be selected.

In process 111, it is determined whether the PGU23, ., l- keys have been pressed.

First, if the robot operation key is pressed, process 1]2 to operate the robot. In addition, in this embodiment, 18 robot operation keys are installed in PG IJ 23''2, and each key corresponds to the following operation: namely, rotation axis direction, one direction. , upper arm axis + direction, one direction, arm cutting axis ten direction, one direction, rotary axis direction, one direction, bending axis direction, one direction, twisting axis + direction, one direction, rotor axis l-x axis,
When any one of the child direction, one direction, and each key of each of the y-axis and Z-axis is operated, the CPU-A 30 writes a command to move the corresponding position by a certain amount on the DPRAM 36. The servo CPU unit 2 (lb moves the corresponding axis of the robot by a certain amount according to the command. Furthermore, when the X-axis, y-axis, and Z-axis keys are pressed, the Ml-M3 motor that controls the robot's posture multiple units operate simultaneously to move the robot's wrist tip in that direction.

If the pressed key is a teaching key, the teaching contents are registered in the teaching data in process 113. The teaching keys include a position teaching key, an interpolation type designation key, a speed designation key, and a work designation key (welding start, end, etc.).

If the position teaching key is pressed now, the current position of the robot is captured and registered in the teaching data.

Similarly, in the case of other keys, the command corresponding to that key is registered in the teaching data.

When the registration is completed, the teaching program is stepped up by one step in step 114.

Then, when the end key is pressed, the teaching process is ended.

Figure 10 shows that using the functions shown in the flow of Figure 9,
This is an example of teaching data when teaching the welding work shown in FIG. 8.

The command "M OV Pi" is a command to move to point 21, and when the position teaching key is pressed, this command is registered in the teaching data. In this example, as mentioned above, the eighth
Points P5 to P1 in the diagram. This shows the case where the section is welded.

Next, in this command, I'p lI+ internally represents the position and orientation of the robot. That is, Pi= (xj
, yi, zi, ai, βj, yi) Note that α1. βi,
γ1 is a parameter representing the posture of the robot, and in this embodiment, Euler angle is used.

Command “E OP I+ is end of Pro
A gram indicates the end of one program.

Next, details of the transfer process 104 shown in FIG. 7 will be explained.

The second transfer process is performed using the teaching data (
This is the process of taking program 10) as input and generating a new program with information about the transfer section added.
The processing flow is shown in FIGS. 11 to 13.

Here, symbols commonly used in these three types of flows will be briefly explained.

S: Step code of the program in Figure 10 1: Point of the program in Figure 10 ATR, when working ``w'' When not working, it becomes II all Attribute d: Step code of the newly generated program Points of the generated program: Teaching points of the program shown in FIG.

of the newly generated program in q point.

FIG. 11 is an overall processing flow of the transfer processing.

First, initialization is performed in process 120. Set ATR to a (air cut), and set sr '' + d + J to l each. N is used to count how many teaching points [・ are in the same attribute. Set this N to O. .

Processing] In step 21, the instruction of the S-th step of the program lo (hereinafter referred to as the source program) shown in FIG. 10 is examined.

In the case of a work-related command, the current attribute ATR is compared with the attribute newly specified by the command in step 122. In the case of welding work, it is defined as ``W''.

If the attributes are different, a 5TOP command is entered into the S-th step of the new program in step 123. This 5TOP command is a command that indicates to stop once at that point. Processing] In step 24, set N to ○, change ATR to the current attribute, and add a 1-step instruction to the new program, so set d to +
Do 1. Thereafter, in process 125, the S step of the source program is copied to the d step of the new program. At step 26, one step of processing of the source program and new program has been completed, so S and d are each set to 10. After that, the process returns to step 121.

Next, if the command is a speed command command in process 121, the specified speed is entered into V in process 127.

After that, jump to step 25.

If the instruction is M O■ in process 121, process 12
At 8, add 1 to N. Next, in step 129, it is checked whether N, 3 or more. If there are only two teaching points in the same attribute, transfer processing does not occur. Therefore, N
is less than 3, in step 130 the MOV is
Enter the command qj to copy pi to qJ. Thereafter, in process 131, 1 and J are incremented by 1. Then, the process jumps to step 126.

Also, if N is 3 or more in process 129, perform the 3-point process in 132, and then go to process 121.If there is another command in ash process 121, jump to process 125 and copy one step. .

If there is no command in process 12], it is determined that the process has ended, and in process 133, the new program is stored in the bubble memory, and the transfer process is ended.

FIG. 12 is a processing flow of three-point processing.

In this three-point processing, the way to transfer at the midpoint is calculated from the data of three continuously operating teaching points.

First, in process 140, vectors 8 and π are calculated.

λ=p+-I Q+-2 (X +-I X 1-21 yL-1-yj-1
+Zi-I Z''-a+ β1-1 (Zj-+
+β 1-1- βj-2+ γ1-1- γ, -2
B=p 1-p = (X= xi,, yI 5'+-z z
, Zα1-β1-1. β1-β1-1. γ, −γ
1-1) Further AI), =-p, λ0. Define π0 as follows.

AI) = (X knee + X, i-xi 5't-+
-yI-4+Z4-I Zj-2) πp (Xr
1-1-βi-2+γ1-1-γ, -4) πo-(α・-β1-1.β3-β1-1.γ, -γ,
-1) Let λp of the first toe be a part representing the position of λ, and 80 be a part representing the attitude of λ.

For reference, the graphical sizes and directions of the partial vectors Wp and πp are shown in FIG. 14.

Next, in process 141, these three points (Q, -+, I) pl)
Calculate the angle O formed by

e = cos-' ((-Ap) ・Bp/(lApl
・fBpl)) Here, the symbol ・represents an inner product, and cos' is set to return a value in O≦0≦180°.

Next, in process 142, it is determined whether O is greater than Δθ. Δθ is a small positive value, in this example approximately 3°.

If θ is less than or equal to Δθ, these three points are turned around, so it is inevitable to stop at point pl-1. Therefore, in process 143, 5TOP is added to the d-th step.
Input the command, step up d in process 144,
In process 146, the MOV qj command is entered in the next step, and pl is copied to q and J. after that,
In process 147, i, j, and d are incremented by 1, and in process 57, S is +], and the process ends.

Next, when 0>△O, in process 145, 0>=180−
80 or not, and when 0≧180−ΔO, the three points are almost lined up in a straight line, so there is no need to consider the transfer, so in process 146, MOVqj is added to the d-th step. , copies pi to qj, increments each pointer by 1, and ends the process.

Next, in process 145, if O≧180−△θ does not hold, it means that the three points intersect at an angle of 80, △e<e<18o−, so a transfer process is necessary. Become.

Therefore, first, in process 148, the route error is calculated.

This path error can be expressed as D = r /5jn(e /2) -7, where the maximum acceleration (corresponding to the motor torque) that the robot can produce is a. 13, then γ-v'/B・5j
n(θ/2)).

The state of these γ and D is shown in FIG.

By the way, this path error represents how much the path deviates from the teaching point when the robot turns this curve at maximum acceleration.

Next, in process 149, the route error just calculated is compared with the allowed error set in process 102 of FIG. Here, as the comparison value DAT3, Dw is used during work, and Da is used during air cutting.

In process 149, if D>D ATR is not satisfied,
This means that it is possible to turn the corner with this radius γ, so 158 points are generated.

Therefore, if D > D A T R, process 15
0, the current speed V is saved to V'.

This is a value smaller than the initial ■.

Next, in process 152, a command specifying this speed ■ is inserted into the d-th step of the new program, and in process 153, d is +
Do 1.

Through this series of processing, the speed is corrected to a speed that allows curve movement within the setting error.

Next, in step 154, point generation processing is performed. After this, in order to return the speed after turning the corner to the original speed, in step 155, a command specifying "Sokudo ■'"' is entered in the d-th step. Then, in process 156, d is incremented by 1, and 14
Skip to step 6.

FIG. 13 shows the point generation process.

First, in process 160, the position of q,-1 is moved from the position of I)l-1 to the starting point of the transfer route.

γ” ” / a m a, Q = y / tan (e /2) q” - + - p + - + - α ° λ / 18 p Next, in process 161, a circular interpolation command is added to the d-th step of the program. Add.

In process 162, qJ is generated, and M○V is added to the d+] step.
Add qj instruction.

Here, qj is calculated using the following formula.

(Position of qj) - (Position of pl-+) (Position of qj) - (Position of pl-t) 10 [(πp/lπ
pl)-(Ap/1Apl)]・DAs mentioned above, D = v' (]-5in(θ/2))/(a
,,,A, ・5in(θ/2)).

Next, in process 163, q,+1 is generated and the MOVq,+, instruction is added to step d+2.

Here, q and +1 are calculated using the following formula.

qh++=p+-+10α・B/lπp In addition, q, + −+ + Q + + Q J + 1
The positional relationship is as shown in FIG.

Next, in process 164, an interpolation command is added to the cl+3 step. This is used to restore the specification of circular interpolation to the d step.

In steps 165 and 166, the number of points plus J and d is incremented by the number of instructions, and the process ends.

FIG. 15 shows a part of the new program obtained by processing the source program 10 through the above processing.

In addition, in this example, Da=180 and Dw=0.005mm, a, , -=4900mm/sec' (=0.5G)
That is.

Here, the relationship between the teaching point and the generation point near P2 is shown in FIG. 16, and the relationship near P7 is shown in FIG.

Next, the automatic driving process of process 105 in FIG. 7 will be explained. In this automatic operation, the robot operates based on a new program.

Fig. 18 shows the flow of automatic operation processing.

This flow is executed by the CPU-A 30.

In step 170, a program number is input. In this example, P
Enter 310 from above BC. Next, in process 1, 71, it waits for the activation button to be pressed. l) When the start button of I'(,:hi is pressed, the specified teaching program is taken out from the bubble memory 35 in process 172 and stored in the RAM -
A. Place it on 35. In process 173, the teaching step counter S is set to 1. Processing] In step 74, the instruction of the S-th step is checked.

In the case of a speed command, the specified speed is substituted for ■ in step 175. S T OI) Do nothing when commanded. At the time of a pokan command, input the specified hokan type as the hokan type with processing +76. M OV command (J, ) Point processing is performed at 77. When it is a work-related command, wait for movement completion at step 178, and then issue commands such as arc ON to peripheral equipment (in this example, welding machine). Note that whether or not the movement is completed is determined by checking the movement completion flag of RAM 36-h.

When these processes are completed, S is stepped up in process +79, and the process goes to process 174. When it is a command or E○P, the automatic operation process is ended.

Next, point processing will be explained using FIG. 19. In this point processing, servo CP U91120
Issue an interpolation command to b to operate the robot.

First, in process 180, the point number at the current step (S) is taken out and assigned to 1. Next, in process 181, M
By the time the OVq,... command comes out, Sokudo and ST○
Check whether there is a speed command such as ①). If there is a command, the speed is entered in the end speed V8 in step 182. For the 5TOP instruction, set v,,=O.

If there is no speed command, in step 183, the speed ■ previously specified by the speed command is substituted into ■. Next, process 1
In step 84, check the hokan type. If it is not circular interpolation, wait for the completion of the movement in step 185. This is a separate CI) Servo CP which is Ll. The Ll section is performing the movement process up to q, so this is the process of waiting for its completion. It is.

When the movement is completed and the robot reaches ql-1, process 18
At step 6, an interpolation command up to ql is issued.

This command: Interpolation type: straight line, joint, arc 1] Gauge: q1 Steady speed: V Ending speed:～・・ Auxiliary point: q′ (only for circular interpolation) command.

In the Zaho CPIJ section, the robot moves from the point where the robot is currently located to the target point according to the interpolation type using the second command. In the speed processing, if the current speed is different from the steady speed \I, it is accelerated or decelerated to \I according to the acceleration of a,1.

Also, as you get closer to the target point, when you reach the target point, the speed will be V
Accelerate and decelerate so that F is reached.

Next, if it is a hokan type or a circular arc in processing 184,
In process 187, it is checked whether ql,1 is also circular interpolation. If there is no poking command specifying another hocantile by q1+1, the process waits for the completion of movement in process I88, specifies the auxiliary point as 91+1, and issues an interpolation command to the servo CPIJ section in process 189.

If q, +1 is not circular interpolation, mover holding is performed in process 190, the auxiliary point is designated as C1 in process 189, and an interpolation command [ku] is issued to the servo CP I-: section.

If q+++ is not circular interpolation, processing 190 waits for movement completion, and subsequent processing 191 issues an interpolation command with the auxiliary point ql-2.

Point processing ends with "-".

The route when the program 10 is automatically operated by the processing described so far is schematically drawn in Figure 1 (3).In this figure, the three dimensions are shown two-dimensionally (7) for ease of understanding. .

Among teaching points P1 to P1c, P,, P,, P,
.

Pl. l P141 Operating on a new route of transfer at each teaching point of pH.

The speed along the motion path at this time is the second
The result will be as shown in the figure.

As is clear from this figure, the work start point P6 and the end point P
, 2, the tip of the robot is not stopped. Moreover, the dotted line in the figure is when the speed is specified to the maximum value. At this time, increase the speed to the maximum rotation speed of each motor shaft. Therefore, with an articulated robot, the linear velocity will not be constant, but the takt time (work time) will be the shortest.

Furthermore, Fig. 21 shows the state of acceleration, and q2~
It can be seen that at q, q, ~q, the acceleration is at the maximum in order to change the direction of motion.

Therefore, this embodiment has the effect of greatly shortening the takt time (work time).

For example, in the section P1 to P1) in these figures, under the following conditions, the operating time is approximately 27% shorter than in the conventional system.

Conditions P-P, 500mm P2-P, 500 printings P, ~P4500 images/PIP2P. 90', /P, P, P, 90'''
Speed: 9800 marks Acceleration: 0.5G Tolerance: Da=200mm Furthermore, in this embodiment, the robot does not stop or decelerate unless necessary at the teaching point, so it has the effect of smoother movement. Furthermore, since the teaching method can be exactly the same as the conventional method, there is an advantage that the teaching does not become complicated even though detailed motion control is performed.

[Effects of the Invention] According to the present invention, all the conditions necessary for the robot to operate smoothly are automatically generated just by inputting a simple mark, which has the effect of simplifying teaching. .

[Brief explanation of drawings]

FIG. 1 is a schematic diagram of an actual motion path in an embodiment of the industrial robot control device according to the present invention, FIG. 2 is a velocity distribution diagram, FIG. 3 is a schematic diagram of an actual motion path in the conventional method,
Figure 5 is a speed distribution diagram of the conventional system, Figure 5 is an overall equipment configuration diagram of an embodiment of the present invention, Figure 6 is an internal block diagram of the robot control device, Figure 7 is a flowchart for explaining operation, Figure 8 The figure is a distribution diagram of teaching points, Figure 9 is a flowchart for explaining the operation, Figure 10 is an explanatory diagram of the teaching program, Figures 11, 12, and 13 are flowcharts for explaining the operation, and Figure 14 is a flowchart for explaining the operation. A process explanatory diagram, FIG. 15 is an explanatory diagram of the generation program, FIGS. 16 and 17 are explanatory diagrams of partial routes,
18 and 19 are flowcharts for explaining the operation, and FIGS. 20 and 21 are acceleration distribution diagrams.

Claims (1)

[Claims] 1. In a teaching-playback type industrial robot control device, means for adding a predetermined mark to teaching point data input by teaching, and a method defined by a plurality of consecutive teaching points. A control means for setting at least two different transfer methods is provided as a control method for transfer routes between consecutive partial routes, and when controlling the robot in accordance with the teaching point data, An industrial robot control device characterized in that one of the transfer methods described above is selected in accordance with the marked mark. 2. In the invention of claim 1, the control means has a predetermined data input means, and by operating the data input means, a predetermined parameter for determining the transfer method to be set is inputted to the control means. An industrial robot control device comprising: 3. In the invention of claim 2, the parameters for determining the transfer method include at least a numerical value that defines the maximum allowable error between the teaching point and the robot's actual motion path, and within this maximum allowable error range, the robot An industrial robot control device configured to determine the transfer route. 4. In the invention according to claim 1, the transfer route of the robot is determined so as not to exceed at least one of the maximum acceleration or the maximum speed that each movable axis of the robot can generate. Robot control device. 5. The industrial robot control device according to claim 3, wherein if the transfer route exceeds the maximum allowable error, the designated speed of the robot on the transfer route is suppressed. . 6. The invention according to claim 1, characterized in that at least one of the types of marks indicates that the robot is working, and at least one of the other marks indicates that the robot is not working. robot control device. 7. The industrial robot control device according to claim 1, wherein the mark is added to all of the teaching points. 8. The industrial robot control device according to claim 1, wherein the mark is added only to a predetermined point among the teaching points. 9. The industrial robot control device according to claim 1, wherein the control coefficients necessary for controlling the transfer route are calculated in advance at the end of the teaching. 10. The industrial robot control device according to claim 9, wherein the calculation of the control coefficient is performed by means different from means for controlling the operation of the robot.