JPH09128024A - Method for optimizing operation program of robot having redundant axis - Google Patents

Abstract

PROBLEM TO BE SOLVED: To optimize the operation program of a robot having a redundant axis. SOLUTION: The initial value of a fine adjustment variable δ for the axial value (posture W around the X axis in this case) of the redundant axis and the predicting accuracy condition of a cycle time are set (S1, S2) and an original program A is read out (S3). Then a teaching point index value (i) is initialized (S4) and a teaching point position for calculating a cycle time is set (S5). In a first processing cycle, the teaching point position is set in accordance with the positional data of the program A. Then the cycle time is calculated under a condition with the axial value Wi of an i-th teaching point finely adjusted and not finely adjusted based upon the set data (S6) and the axial value of a redundant axis corresponding to a condition giving the shortest time is taken as an axial value after executing fine adjustment processing (S7). Then fine adjusting processing for n teaching points is executed through the updating of an index value (S8) and the checking of the existence of a remaining teaching point (S9), the value δ is updated down (S10) and processing is similarly repeated. When the value δ becomes lower than a previously set value, δmin, a program B adopting the axial value Wj of the redundant axis is outputted (S12) to end the processing.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for optimizing an operation program of an industrial robot (hereinafter, simply referred to as "robot"), and more specifically, it relates to a robot having an allowable redundant degree of freedom. The present invention relates to a technique for optimizing a program that defines an operation. The technology of the present invention
For example, spot welding robot, arc welding robot,
Effectively applied to sealing robots.

[0002]

2. Description of the Related Art Generally, when teaching a robot to move a path, a desired number of points are selected along the path to be realized, and the position and orientation of the robot (usually the position of the tip of the tool) is selected for each point. Is represented by the posture) and a motion program is created through this teaching work. The created motion program includes position data that uniquely specifies the position and orientation of the robot in the three-dimensional space.

By the way, depending on the contents of the work performed by the robot, there are many cases where the posture or position of a specific axis is not directly restricted by the work contents. FIG. 1 is a diagram for explaining this. FIG. 1A shows a case of a spot welding robot and FIG. 1B shows a case of an arc welding robot.

In the case of the spot welding robot shown in FIG. 1 (1), the attitude (arrow 4) about the pressure axis (double arrow) of the spot gun 2 is seen from the viewpoint of performing welding work at that position. As long as there is no particular restriction, it is impossible to uniquely determine the optimum posture immediately.

As described above, an axis for which redundancy is allowed in determining the posture or position from the viewpoint of work content is also called a "redundant axis". In the illustrated case, the tool coordinate system of the spot welding robot is set such that the pressure axis direction of the spot gun 2 and one coordinate axis (for example, X axis).
Since they are performed so that the directions of the (axis) coincide with each other, the axis (W axis) representing the rotation around the X axis of the tool coordinate system becomes the redundant axis.

Further, in the case of the arc welding robot shown in FIG. 1 (2), regarding the attitude of the welding torch 3 mounted on the robot arm 1 in the axial direction, there are restrictions on the aiming angle and the forward / backward angle. Therefore, the optimal posture is uniquely determined. On the other hand, the posture (arrow 5) around the axis of the welding torch 3 is not particularly limited from the viewpoint of performing welding work at that position.

Therefore, the axis of the welding torch 3 corresponds to the redundant axis. Since the tool coordinate system of the arc welding robot is often set so that the axis of the welding torch 3 coincides with the Z-axis direction, in this case, the axis representing the rotation around the Z-axis of the tool coordinate system (R-axis). Is the redundant axis.

In this sense, when the robot has a redundant axis, the position (the angular position in the case of the posture. The same applies hereinafter) about the redundant axis can be taught for the time being, and therefore the teaching work is burdened accordingly. Would be reduced. However, if the position related to the redundant axis is freely specified, even if the teaching position itself at each teaching point does not cause any inconvenience in performing the work, considering the time required for the robot to move the route, the optimized operation is performed. It is generally insufficient to make it happen.

For example, when the teaching is performed such that the angular position (orientation) of the redundant axis changes unnecessarily during the movement between the adjacent teaching points, the time required for the robot to move the path is obviously long. Turn into. In order to avoid such a situation, it is sufficient to teach the angular position (orientation) with respect to the redundant axis in consideration of shortening the route moving time, but it is optimal to shorten the route moving time for many teaching points. Creating a program that defines the position or posture of
It is extremely difficult in practice. Moreover, even if a program close to the optimum program can be created, a large increase in the burden for that is inevitable.

Despite this situation, there is no known suitable technique for automatically optimizing the operation of a robot having redundant axes from the viewpoint of shortening the path movement time. It was an unsolved issue in applications such as arc welding robots and sealing robots.

[0011]

Therefore, the present invention is to provide a technique for providing an operation program optimized for a robot having redundant axes from the viewpoint of minimizing the time. Further, the present invention intends to improve the efficiency of the teaching work and shorten the cycle time of the work by the robot.

[0012]

According to the present invention, in order to solve the above-mentioned problems, software processing for optimization is applied to a created operation program that defines the operation of a robot having redundant axes. By doing so, an operation program optimized from the viewpoint of shortening the time for route movement is created.

That is, in the method of the present invention, the software processing of the robot controller is used to (1) input the operation program to be optimized, and (2) reduce the cycle time. A step of repeatedly finely adjusting the position data of the teaching points to be adjusted included in the input operation program with respect to the redundant axis, and a step (3) of outputting the operation program for which the iterative fine adjustment has been completed are executed. Here, the iterative fine adjustment in the step (2) is executed while increasing the level of fine adjustment with respect to the redundant axis, that is, while correcting the parameter indicating the level of fine adjustment downward.

More specifically, in the step (2) of the iterative fine adjustment, (2-1) in order to calculate the cycle time, a position after adjustment up to that point is set for each teaching point. And (2-2) a condition in which the axis value of the redundant axis for one adjustment target teaching point is finely adjusted upward by a predetermined amount based on the set teaching point position data, and downward fine adjustment by a predetermined amount. And the step of calculating the cycle time under each condition without fine adjustment, and (2-3) of the above conditions, the axis value of the redundant axis corresponding to the condition giving the shortest time is the axis after the fine adjustment processing. (2-1) ~
The step (2-3) is repeatedly executed while changing the adjustment target teaching points, and (2-1) to (2-) for all the adjustment target teaching points included in the input operation program.
Each time the step 3) is completed, the predetermined amount of upward fine adjustment and the predetermined amount of downward fine adjustment are updated downward, and the process is repeated until the fine adjustment amount reaches a sufficiently small limit.

In the preferred embodiment, the predetermined amount of upward fine adjustment and the predetermined amount of downward fine adjustment are described by a parameter δ representing the level of fine adjustment. By using this parameter δ, the fine adjustment condition of the redundant axis is determined as the condition for finely adjusting the axis value by + δ and the condition for fine adjustment of −δ. When the fine adjustment conditions of these redundant axes conflict with the limitation of the motion range set in advance in the robot, the cycle time is calculated under each condition excluding the conflicting conditions.

[0016]

The operation program optimizing method of the present invention operates in the direction of shortening the cycle time while repeating the fine adjustment on the redundant axis with respect to the created operation program which defines the operation of the robot having the redundant axis. It has a basic feature in that it performs software processing that modifies a program.

In order to converge the modification of the operating program in the direction of optimization, the level of fine adjustment with respect to the redundant axis is gradually raised. A way to change the level of fine adjustment for the redundant axis is to change the amount of fine adjustment when calculating the cycle time. In a preferred embodiment adopting this idea, the condition that the axis value of the redundant axis is finely adjusted by + δ,
−δ The cycle time is calculated and compared between the condition where the fine adjustment is performed and the condition where the fine adjustment is not performed, and the axis value of the redundant axis corresponding to the condition of a shorter time is sequentially adopted.

It is considered that there are few cases in which an operation program optimized from the viewpoint of shortening the cycle time conflicts with the restriction of the operation range set in the robot. However, when calculating and comparing cycle times, by eliminating the conditions that conflict with the robot's operating range limits, the output of the operating program including position data that conflicts with the operating range limits can be eliminated more reliably. To be done.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION The method of the present invention can be carried out using a robot controller having a normal hardware configuration. FIG. 2 is a block diagram showing a representative configuration of the representative configuration.

In FIG. 2, the robot controller, which is generally designated by the reference numeral 30, is equipped with a processor board 31, and the processor board 31 is a central processing unit (hereinafter referred to as CPU) 31 including a microprocessor.
a, ROM 31b and RAM 31c.

The CPU 31a controls the entire robot controller according to the system program stored in the ROM 31b. In the RAM 31c, in addition to the created operation program (unoptimized), various setting values, etc., a program for executing optimization processing described later and related setting values, and operation program data after the optimization processing are stored. Is stored. A part of the RAM 31c is used for temporary data storage for calculation processing executed by the CPU 31a.
Note that a hard disk device or the like prepared as an external device is appropriately used for storing the program data and the set values.

The processor board 31 is connected to a bus 37, and commands and data are exchanged with other parts in the robot controller 30 via this bus connection. First, the digital servo control circuit 32 is connected to the processor board 31, receives a command from the CPU 31a, and drives the servo motors 51 to 56 via the servo amplifier 33. The servo motors 51 to 56 that operate the respective axes are built in the mechanical units of the respective axes of the robot RB.

The serial port 34 is connected to a bus 37, and is connected to a teaching operation panel 57 with a liquid crystal display and an RS232C device (communication interface) 58. The teaching operation panel 57 is used for inputting programs such as operation programs, position data, and other necessary set values. In addition, an input / output device (digital I / O) 35 for digital signals and an input / output device (analog I / O) 36 for analog signals are coupled to the bus 37.

Here, as one embodiment of the present invention, a procedure for performing optimization using the following program A as an original program by using the robot controller 30 having the above-described configuration and function will be described. Program A
Q1, Q2 ... Q7 as teaching points, X = constant value (148
It represents a program that moves on a straight line from the start point Q1 to the end point Q7 on the plane 0).

[Program A] SPEED = 2000 MOVE TO Q1 MOVE TO Q2 MOVE TO Q3 MOVE TO Q4 MOVE TO Q5 MOVE TO Q6 MOVE TO Q7 (position data) Q1 = (1480, 0, 990, 180, 0, 0) Q2 = (1480, 200, 990, 180, 0, 0) Q3 = (1480, 400, 990, 180, 0, 0) Q4 = (1480, 600, 990, 135, 0, 0) Q5 = (1480, 600, 790, 90, 0, 0) Q6 = (1480, 600, 590, 90, 0, 0) Q7 = (1480, 600, 390, 90, 0, 0) If it is displayed above and the posture around the X axis described by W is also indicated by an arrow, it becomes as shown in FIG. The direction of W = 0 was the + Y axis direction. As shown in the figure, the posture around the X axis specified in program A before applying the optimization processing is Q1 to Q3.
Up to W = 180 degrees, at corner point Q4 W = 135 degrees, Q
W = 90 degrees from 5 to Q7. In general, such teaching that does not involve a fine posture transition can be performed relatively easily.

In the present embodiment, the program A described above is used.
As can be seen from the axis values P and W of 0 being always 0, the Y axis and Z axis of the three-dimensional Cartesian coordinate system in which the robot operates always match the Y axis and Z axis of the tool coordinate system, respectively. The operation to keep the direction shall be performed. Further, it is assumed that a redundant degree of freedom is allowed with respect to the posture around the X axis of the tool coordinate system. Therefore, the redundant axis is the W axis, and it is the axis value W of the W axis that is adjusted for optimization.

Such conditions are consistent with the case where the program A is assumed to describe the path movement operation of the spot welding robot in FIG. 1 (1), for example. In the spot welding robot, all or most of the teaching points Q1, Q2 ... Q7 are usually selected corresponding to the welding points (Q1 and Q7 may be used as approach points and department store points, respectively).

Hereinafter, assuming that the program A having the above contents is already stored in the RAM 31c of the robot controller 30,
The process (optimization process) for optimizing the program A will be described with reference to the flowchart of FIG. First, the initial value of the parameter (fine adjustment amount of redundant axis) δ representing the level of fine adjustment is input to the liquid crystal display of the teaching operation panel 57 from the condition setting screen. The specified value of δ is internally set (step S
1). Similarly, when the condition regarding the prediction accuracy of the cycle time is input from the condition setting screen, the specified prediction accuracy condition is internally set (step S2). The condition regarding the prediction accuracy of the cycle time determines the condition at the time of calculating the cycle time in step S6 described later,
Here, any one of coarse / medium / fine is selected. Which condition is to be selected is preferably determined in consideration of the required precision of optimization, the prepared cycle time calculation program, and the like.

When a command indicating completion of condition setting is input from the teaching operation panel 57, the created program A is read (step S3), and the initial value i = 1 of the teaching point number index i.
Is internally set (step S4). Next, the teaching point position Qj = (Xj, Yj, Zj, Wj, Pj, Rj); j = 1,2 ... n when the cycle time calculation is executed in the subsequent step S6 is internally set ( Step S5). Here, it is the number of teaching points targeted for the optimization calculation processing, and n = 7 here. The teaching point position set in the first processing cycle matches the position data of the program A including the axis value Wj of the redundant axis.

In the subsequent step S6, based on the data set in step S5, the axis value Wi of the redundant axis of the i-th teaching point is finely adjusted by + δ (1) and -δ is finely adjusted (2). Also, the cycle time is calculated under the condition (3) in which no fine adjustment is made. However, although not described in the flow chart, the cycle time is not calculated when Wi + δ or Wi−δ is outside the operation range set for the robot.

The calculation results of step S6 are compared, and the axis value of the redundant axis corresponding to the condition giving the shortest cycle time is set as the axis value after the fine adjustment processing. For example, if the cycle time calculated under the condition (1) is the shortest, new Wi = old Wi + δ is set. Similarly, if the cycle time calculated under the condition (2) is the shortest, new Wi = old Wi−δ is set. If the cycle time calculated under the condition (3) is the shortest, new Wi = old Wi.

In the processing cycle of steps S5 to S7, the index value is updated in step S8 and step S8 is executed every time.
It is repeated n times in total (n is the number of teaching points), including the check for the presence of remaining teaching points in 9. Δ set in step 1
When the cycle time calculated under the condition (3) is the shortest for all the adjustment target teaching points under the initial value of, the fine adjustment under δ is completed. After completing the fine adjustment process,
In step S10, the value of δ is updated downward (updating of the fine adjustment level). Here, new δ = old δ × (1 /
2).

Then, after confirming that the downwardly updated δ value is not below the preset minute limit value δmin (step S11), the process returns to step S4. The processes after step S4 are as described above. However, it should be noted that in the position data set in step S5 as the reference data for the cycle time calculation, the axis value Wj of the redundant axis uses the value selected in the previous processing cycle.

By repeating the above processing a sufficient number of times,
The axis value Wj of the redundant axis of each teaching point converges in the direction that gives the shortest cycle time. Therefore, the program B that employs the axis value Wj of the redundant axis that is set when δ falls below the minute limit value δmin set to be sufficiently small is output (step S12), and the processing is ended. Output program B
Is a program obtained by optimizing the program A.

Normally, the smoother the redundant axis value changes, the shorter the cycle time tends to be. Therefore, the position data of the program B with respect to the program A having the above-mentioned position data is as follows. You can think.

(Position data of program A) Q1 = (1480, 0, 990, 180, 0, 0) Q2 = (1480, 200, 990, 160, 0, 0) Q3 = (1480, 400, 990, 140) , 0, 0) Q4 = (1480, 600, 990, 135, 0, 0) Q5 = (1480, 600, 790, 115, 0, 0) Q6 = (1480, 600, 590, 110, 0, 0) Q7 = (1480, 600, 390, 90, 0, 0) FIG. 5 shows the teaching point position and the attitude around the X axis for the program B having the above position data in the same format as FIG. Of course, among X, Y, Z, W, P, R,
Only W is adjusted to a different value from Program A. If the redundant axis is another axis, for example an axis representing rotation about the Z axis, the value of R is adjusted. Similarly,
If the axis represents translation about the Z axis, the value of Z is adjusted.

In the above, the optimization method has been described for the case where the redundant axis is the redundant axis which represents the rotation around the coordinate axis of the tool coordinate system in which the posture with respect to the spatial coordinate system is constant. It goes without saying that the method of the present invention can be applied as long as the above can be specified. For example, when the redundant axis is specified as one coordinate axis of the work coordinate system or an axis representing the posture around the coordinate axis, the method described above may be applied with the redundant axis as the redundant axis. If a specific robot axis Jk (for example, the sixth axis J6 of a 6-axis robot) can be designated as a redundant axis, the position data of the original program is converted into each axis value, and the axis value of the redundant axis is included therein. By repeating the fine adjustment process for, the original program can be optimized.

When the present invention is applied to the original program, a specific teaching point (for example, approach point)
With respect to, when the property of the redundant axis is lost (eg, the posture must be a specific posture), the teaching point may be excluded from the adjustment target teaching point in the above-described processing (steps S5 to S5). S5 is skipped over or the fine adjustment amount δ of the teaching point is set to δ = 0).

[0039]

According to the present invention, an operation program optimized from the viewpoint of shortening the cycle time is automatically created from the operation program created for a robot having redundant axes. Therefore, the efficiency of the teaching work at the time of creating the original program and the reduction of the cycle time of the work by the robot can be achieved at the same time.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a case in which a redundant axis occurs in a robot. (1) is an example of spot welding robot,
(2) represents an example of an arc welding robot.

FIG. 2 is a block diagram showing a hardware configuration of a robot controller that can be used to implement the present invention.

FIG. 3 is a diagram in which the positions of teaching points in the program A before the optimization process is applied are displayed on the ZY plane, and the postures around the X axis described by W are also indicated by arrows.

FIG. 4 is a flowchart outlining a program optimization process according to an embodiment of the present invention.

5 is a diagram illustrating the positions of teaching points in program B obtained by applying optimization processing to program A in the same format as FIG.

[Explanation of symbols]

 1 robot arm 2 spot welding gun 3 welding torch 4 arrow indicating the degree of freedom of posture around redundant axis 30 robot controller 31 processor board 31a main processor 31b ROM 31c RAM 32 digital servo control circuit 33 servo amplifier 34 serial port 35 digital signal Input / output device (digital I / O) 36 Analog signal input / output device (analog I / O) 37 Bus 51-56 Servo motor 57 Teaching operation panel with liquid crystal display 58 RS232C device (communication interface) Q1 ~ Q7 Teaching point

Claims (5)

[Claims] 1. A method of optimizing a motion program of a robot having redundant axes by software processing, comprising: (1) inputting a motion program to be optimized; and (2) shortening cycle time. In a direction to perform, the step of repeatedly finely adjusting the position data of the teaching points to be adjusted included in the input operation program with respect to the redundant axis, and (3) the step of outputting the operation program in which the iterative fine adjustment is completed, The method, wherein the iterative fine adjustment in the step (2) is performed while decreasing a value of a parameter representing a level of fine adjustment with respect to a redundant axis. 2. A method of optimizing an operation program of a robot having redundant axes by software processing, comprising: (1) inputting an operation program to be optimized; and (2) shortening cycle time. In a direction to perform, the step of repeatedly finely adjusting the position data of the teaching points to be adjusted included in the input operation program with respect to the redundant axis, and (3) the step of outputting the operation program in which the iterative fine adjustment is completed, (2) the step of (2) the iterative fine adjustment, (2-1) the step of setting the adjusted position for each teaching point up to that point in order to calculate the cycle time; and (2-2). According to the teaching point position data set above,
Calculating the cycle time under the condition where the axis value of the redundant axis is finely adjusted upward by a predetermined amount for one adjustment target teaching point, the condition is finely adjusted downward by a predetermined amount, and each condition is not finely adjusted; 2-3) The step of (2-1) to (2-3), including the step of setting the axis value of the redundant axis corresponding to the condition that gives the shortest time among the above conditions as the axis value after the fine adjustment processing, Is repeatedly executed while changing the adjustment target teaching point, and every time the steps (2-1) to (2-3) are completed for all the adjustment target teaching points included in the operation program,
The predetermined amount of the upward fine adjustment and the predetermined amount of the downward fine adjustment are updated downward, and under the updated fine adjustment condition, (2)
Steps -1) to (2-3) are repeatedly executed while changing the adjustment target teaching point, and the predetermined amount of the upper fine adjustment and the predetermined amount of the lower fine adjustment fall below a preset limit value. Then, the method, wherein the step (2) is completed. 3. In the step (2-2), the condition in which the axial value of the redundant shaft is finely adjusted upward by a predetermined amount or the condition in which it is finely adjusted downward is a limit of an operation range preset in the robot. The method for optimizing a motion program of a robot having a redundant axis according to claim 2, wherein in the case of conflict, the cycle time is calculated under each condition excluding the conflict condition. 4. A method of optimizing a motion program of a robot having redundant axes by software processing, comprising: (1) inputting a motion program to be optimized and a parameter δ representing a fine adjustment level of the redundant axis. And (2) in the direction of shortening the cycle time, iteratively finely adjusting the position data of the adjustment target teaching points included in the input operation program with respect to the redundant axis, and (3) the iterative fine adjustment is performed. The step of outputting the completed operation program, wherein the step (2) of the iterative fine adjustment includes: (2-1) the position which has been adjusted up to that point for each teaching point in order to calculate the cycle time. And (2-2) based on the teaching point position data set above,
Calculating a cycle time under conditions in which the axis value of the redundant axis is finely adjusted + δ, finely adjusted −δ, and conditions not finely adjusted for one adjustment target teaching point; (2-3) Of the conditions, including the step of setting the axis value of the redundant axis corresponding to the condition that gives the shortest time as the axis value after the fine adjustment processing, the steps (2-1) to (2-3) include the adjustment target. Each time the steps (2-1) to (2-3) are completed with respect to all the adjustment target teaching points included in the operation program, which are repeatedly executed while changing the teaching points,
The value of the parameter δ is updated downward, and under the updated fine adjustment condition, the steps (2-1) to (2-3) are repeatedly executed while changing the adjustment target teaching point, The method, wherein the step (2) is completed when the value of δ falls below a preset limit value δ min. 5. In the step (2-2), the condition in which the axis value of the redundant axis is finely adjusted by + δ or the condition in which −δ is finely adjusted conflicts with the limitation of the operation range preset in the robot. In this case, the method for optimizing a motion program of a robot having redundant axes according to claim 4, wherein the cycle time is calculated under each condition excluding the conflicting condition.