JPS61148507A - Control device of robot - Google Patents

Abstract

PURPOSE:To always optimize a coefficient of forecasting speed by correcting successively the coefficient of forecasting speed, based on a hysteresis data which has been stored this time and a hysteresis data which has been stored in the previous time. CONSTITUTION:A target speed command value VCP from a central processing part 1 is updated and stored successively in a target speed command value data latch 2, and a forecasting speed coefficient MTCn is updated and stored in a forecasting speed coefficient data latch 3. A data entry switch 4 for setting a position correcting coefficient outputs a position correcting coefficient delta which has been set by operating a switch. A target position command value PCP is updated successively in a target position command value data latch 5, and a present position value PP from a pulse counter 9 which is detecting a present position of a movable part by counting a feedback pulse FP in accordance with a rotational direction of a driving motor 7 is updated and stored successively in a present position value data latch 6. In this way, the coefficient of forecasting speed is corrected successively, based on the target position command value of the time of a necessary operation, which has been detected whenever a movable part of a robot is brought to a necessary operation, the present position value, and the present time and the previous time of a hysteresis of the position deviation.

Description

[Detailed Description of the Invention] [Industrial Application Field] This invention provides a method for driving and controlling the movable parts of a robot in accordance with a speed command value based on a value obtained by multiplying a target speed command value by a predicted speed coefficient. This paper relates to improved technology for robot control devices.

[Conventional technology]

2. Description of the Related Art Conventionally, a robot control device is known that calculates a speed command value for driving and controlling a movable part of a robot using the formula 0.

Q gap m order 1υ = (me side - treatment IDX (pre-soak pox roasted field +
Position correction phase detection 0......■ Here, the predicted speed coefficient (gain) refers to the target speed command value, the control ability of the hardware of the servo control drive system, and the robot body including the drive motor. It is a correlation coefficient between the actual speed value of the robot movable part (or drive motor) based on the thick movement capacity of the mechanism part, etc. 1 position correction coefficient is the position deviation between the target position command value and the current position value. This refers to a speed command value correction coefficient (gain) for decreasing the speed command value, and is set by operating a data end switch for setting coefficient data provided in each robot's control device in advance.

[Problem to be solved by the invention] -By the way,
In the conventional robot control device as described above, the predicted speed coefficient is fixedly set by a data entry switch, which causes the following problems.

In other words, not only do the robot main body mechanisms and the hardware of the servo control drive system change over time, but the birdware of the servo control drive system is an electronic circuit, so the control values are subject to temperature drift due to the influence of the environmental temperature. If the predicted speed coefficient is set fixedly by the data entry switch, the correlation between the target speed command value and the actual speed value will be
The problem was that it could not respond to subtle changes.

In addition, if it is recognized that the robot's motion has become difficult to regroup due to a change in the correlation, and the predicted speed coefficient is reset, the robot that is currently operating on the production line will stop the line. There were also issues such as what I had to do.

This invention attempts to solve the above problems.

[Means for solving problems]

Therefore, in the present invention, as shown in FIG. 1, the movable parts of the robot are controlled in accordance with a speed command value based on a value obtained by multiplying the target speed command value as described above by a predicted speed coefficient. The control device A for the robot includes a history detecting means B for detecting a history of positional deviation between a target position command value and a current position value at the time of the required operation each time a movable part of the robot is operated as required, and this history detecting means A storage means C for storing the history of the positional deviation detected by B, and a correction means for sequentially correcting the predicted speed coefficient based on the history data stored this time and the history data stored last time in the storage means C. Establish and configure.

[For production]

In such a configuration, each time the history detection means B detects the history of positional deviation, the detected history is stored in the storage means C, while the correction means stores the history data stored in the storage means C this time and the previous time. Since the predicted speed coefficient is sequentially corrected based on the stored history data, the predicted speed coefficient can always be optimized.

〔Example〕

Embodiments of the present invention will be described below with reference to FIGS. 2 and 3 of the drawings.

FIG. 2 is a block diagram showing an embodiment of the present invention.

The robot control device shown in FIG. 2 controls a one-axis robot consisting of a plurality of moving parts (not shown), but the -axis is shown as a representative, and the following explanation will focus on the two Only the axis will be explained.

However, it goes without saying that the same thing can be said about the other axes as explained below. .

In the figure, numeral 1 is a central processing unit, and the system is composed of a microcomputer and a memo for storing teaching data. In addition to various functions, by executing a program to be described later, it serves as a history detection function, a storage function, a correction function, etc. according to the present invention.

The central processing unit 1 is sequentially inputted with the current position value pp of the movable part (axis) of the robot, and the central processing unit 1 is also inputted with the target speed command value vCP and the predicted speed coefficient M V
Cn, target position command value PCP, etc. are sequentially output.

2 is a target speed command value data latch, and central processing unit 1
The target speed command value vcp' from is sequentially updated and stored.

3 is a predicted speed coefficient data latch in which the predicted speed coefficient M V Cn from the central processing unit 1 is updated and stored.

4 is a data entry switch for setting a position correction coefficient, and outputs the position correction coefficient δ set by switch operation.

5 is a target position command value data latch, and the central processing unit 1
The target position command value PCP from is sequentially updated and stored.

Reference numeral 6 denotes a current position value data latch, which outputs a feedback pulse FP from a pulse generator (pulse encoder) 8 attached to the output shaft of a drive motor 7 that moves the movable parts of the robot (not shown) according to the rotational direction of the drive motor 7. The current position value PP from the pulse counter S, which detects the current position of the movable part by counting up or down, is successively updated and stored.

The current position value PP latched in the current position value data latch 6 is input to the central processing unit 1. 10 is a speed command value calculation unit, which is actually controlled by the system by a microcomputer. Its functions are shown as a hard block: a multiplier 11. Subtractor 12
9 multiplier 13. and an adder 14.

Then, this speed command value calculation unit 10 calculates the target speed command value vcp from the target speed command value data latch 2, the predicted speed coefficient MVCn from the predicted speed coefficient data latch, the position correction coefficient δ from the data entry switch 4, the target The speed command value Vx is calculated by inputting the target position command value pcp from the position command value data latch 5 and the current position value PP from the current position value data latch 6, and performing the calculation of the above-mentioned formula 0. and output it.

Multiplier 11 outputs VCPXMVCn, and subtracter 1 outputs VCPXMVCn.
2 is PCP-PP, and multiplier 13 is (Pcp-pp)x
s, addition m14 is VCPXMVCn+ (PCP-P
P)
-PP) Outputs Xδ as Vx.

Reference numeral 15 denotes a speed command value data latch in which the speed command value Vx from the speed command value calculating section 10 is sequentially updated and stored.

16 is a D/A converter, and speed command value data latch 1
The speed command value Vx from 5 is converted into an analog speed command signal Sx
Digital/analog conversion is performed and output.

Reference numeral 17 denotes an F/V converter, which converts the frequency of the feedback pulse FP from the pulse generator 8 into a voltage signal, detects the actual speed of the drive motor 7 or the robot movable part, and outputs it as an actual speed signal Sv. .

18 is a servo amplifier, which receives the speed command signal Sx from the D/A converter 16 and the actual speed signal S from the F/V converter 17.
A drive current based on the deviation from v is output to the drive motor 7.

Next, the operation of the embodiment configured as described above will be explained with reference to the flow diagram of FIG.

However, the following explanation assumes that the robot driven and controlled by this control device is installed in a predetermined production line and repeats the required operations.

Further, the target position command value PCP and target speed command value vCP output from the central processing unit 1 are actually values that have been subjected to necessary interpolation processing, but a detailed explanation of the interpolation processing will be omitted.

CPU of the microcomputer that constitutes the central processing unit 1
First, as shown in FIG. 3, the following servo processing is performed.

That is, the target position command value pcp based on the teach data in the teach data storage memory is written into the target position command value data latch 5 in FIG. is the original position data of the robot movable part) and then divides the value by the travel time between both target positions, and the target speed command value vc is based on the value obtained by the calculation.
Write p to target speed command value data latch 2.

Then, the predicted velocity coefficient MVC is set in advance according to the state of the servo control drive system and robot main body mechanism shown in Fig. 2.
n is written to the predicted speed coefficient data latch 3, and the current position value pp of the robot movable part latched to the current position value data latch 6 is read, and the position deviation ΔS from the current target position command value PCP is calculated. .

However, it is assumed that the predicted speed coefficient M V Cn written in the predicted speed coefficient data latch 3 is not updated unless M V Cn is changed by adding a correction to be described later.

After performing such a series of servo processing, the CPU determines the movement direction of the robot movable part by checking the sign of the difference between the current value and the previous value of the target position command value PCP calculated by the servo processing. However, if the operating direction is the direction corresponding to the sign of 10 (+ direction), the position deviation ΔS calculated by servo processing is added to the value (S n +) of the child direction register whose initial value is zero. , if the operating direction is the direction with respect to the negative sign (unidirectional), then the process of adding the positional deviation ΔS to the value (Sn-) of the one-sided register whose initial value is zero is also performed.

Then, each process from the servo process to the addition process of the positional deviation ΔS is repeated until the robot movable part performs the required motions and one cycle of motion is completed.

Thereby, the robot movable part is controlled by the speed command value calculation unit 10.
The drive motor 7, which is rotationally driven in accordance with the speed command value Vx sequentially calculated by An integrated value ΣΔS of ΔS regarding the movement direction is stored.

Then, when the ■cycle ends, the CPU first sets the absolute value 1 (Sn+) of the value (Sn+) of the child direction register.
The absolute value 1 (Sn
-) Calculate the sum with 1 and write the calculation result to the history register Sn. Note that after this writing process is completed.

(S n +) and (S n −) are reset to zero value.

Next, the predicted speed coefficient MVCn (current value) written to the predicted speed coefficient data latch 3 in the servo processing, the predicted speed coefficient MVCn written last time (this will be referred to as MVC old), and the predicted speed coefficient MVCn written this time to the history register Sn. value (this is written as (Sn)) and the value written last time (this is written as (Sn)).
Based on ((MVCn≧MVCn-+)A((Sn)<(Sr+
−+))) V ((MVCn<MVCn−+)A[(
If Sn)≧(Sn-+))), then the next predicted speed coefficient MVCn is the corrected value by adding a predetermined minute value Δ for correcting the predicted speed coefficient to the current predicted speed coefficient MVCn.
((MVCn≧MVCIM)△((Sn)≧(Sn-+
))) V ((MVCn<MVCn-+)△((Sn
)<(Sn-t)) ), then the current predicted speed coefficient M
A process is performed in which the above-mentioned minute value Δ is subtracted from VCn and the corrected value is set as the next predicted speed coefficient M CV n.

After performing this correction process and performing various processes not shown in the drawings, each process in FIG. 3 is repeatedly executed in order from the first servo process.

As a result, the robot can repeat the required movements, and the following effects can be achieved.

In other words, the hardware of the robot body mechanism and servo control drive system has not changed over time.

Moreover, the environmental temperature of the control system hardware is normal temperature (room temperature).
is constant, the predetermined predicted speed coefficient M V
If Cn is correct, the robot moves correctly as requested based on the teach data, and as long as the same operation is repeated, ■ The position deviation ΔS remains in the child direction register and one direction register each time the cycle operation ends as historical data. The integrated value ΣΔS is approximately constant, and (S n) = l (
S n +) l +l (S n -) l is also approximately constant.

Therefore, at the end of the first cycle of operation, M
VCn = MVC: n-+ and S n-+
= 0, a correction process is performed in which MVCn is subtracted by the minute value Δ, and at the end of the next cycle of operation, M V Cn < M V Cn-1 (M V Cn at this time).
(V Cn-+ is larger than M V Cn by Δ), but since S n "S n-+, M V Cn is then subjected to a correction process in which only a minute value Δ is added, and returns to its original value. return.

Therefore, the predicted speed coefficient M V Cn is equal to this minute value Δ
As long as the robot's operation is not affected even if the temperature changes, and the robot and its control system do not deteriorate and the temperature of its environment does not change, the robot will follow the initial prediction. It moves correctly with the speed command value Vx approximately based on the speed coefficient MVC:n.

However, if the environmental temperature changes significantly or the hardware of the robot main body components and servo control system deteriorates over time, even if you try to repeat the same operation, it will not move according to the command value, so one cycle operation Position deviation ΔS that remains in the child direction register and one-way register each time it ends
The integrated value ΣΔS as history data also changes.

For example, if we consider a case where, when moving in the + and one direction, each tends to move closer to the target position, in this case, both (Sn+) and (Sn) will be smaller than in the normal case.

Therefore, if there is a tendency like that,
(Sn)=l(Sn+)I+1(Sn)I is also smaller than the previous (Sn-1), in this case (Sn)
< As long as (Sn-+) continues, the predicted speed coefficient M
VCn increases by a minute value Δ every cycle.

As a result, this minute value increases by Δ< M
Since the speed command value Vx calculated based on V Cn also gradually increases from the normal value, the robot is automatically adjusted to gradually move correctly to the normal target position.

Of course, in the opposite case to the above trend, MVCn decreases by a minute value Δ every cycle, and the predicted speed coefficient M,
VCn can be optimally corrected.

In the above embodiment, the predicted speed coefficient M V Cn is corrected in real time during operation on the production line, but in addition to this, it is also possible to perform regular maintenance on the robots on the production line. By using this method, the robot is not corrected for M V Cn when it is in operation on the production line, but a trial is performed in which the robot repeats the required movements during maintenance, and the process shown in Figure 3 is executed during the trial. Therefore, MVCn may be optimized before the line is introduced.

〔Effect of the invention〕

As explained above, according to the present invention.

The predicted speed coefficient is sequentially corrected based on the current value and previous value of the target position command value, current position value, and position deviation history at the time of the required operation detected every time the robot's movable part is operated. Therefore, it is possible to respond to changes in the control amount due to changes in the robot body mechanism over time, changes in the control system hardware over time, temperature drift of the hardware, etc. Moreover, the predicted speed coefficient can be automatically adjusted without stopping the production line.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of the present invention, and FIG. 2 is a block diagram showing an embodiment of the present invention. FIG. 3 is a flow diagram of a program executed by the central processing unit 1 of FIG. 1... Central processing unit 2... Target speed command value data latch 3... Predicted speed coefficient data latch 4... Data entry switch 5... Target position command value data latch 6... Current position value Data latch 7... Drive motor 8... Pulse generator 9... Pulse counter 10... Speed command value calculation unit 15... Speed command value data latch 16... D/A converter 17...・F/V converter 1
8... Servo amplifier diagram 1

Claims (1)

[Scope of Claims] 1. A robot control device configured to drive and control a movable part of a robot according to a speed command value based on a value obtained by multiplying a target speed command value by a predicted speed coefficient, comprising: A history detecting means for detecting a history of positional deviation between a target position command value and a current position value at the time of the required operation each time the movable part is operated as required; and a history of the positional deviation detected by the history detecting means. A control device for a robot, comprising: a storage means for storing; and a correction means for sequentially correcting the predicted speed coefficient based on history data stored this time and history data stored last time in the storage means. .