JPH04355805A - Learning controller - Google Patents

Abstract

PURPOSE:To eliminate a need of a target position command, a position loop, a manual operation function, etc., in a learning controller applied to the servo system where the repeated operation of a prescribed pattern of a machine tool, a robot, or the like is performed. CONSTITUTION:A learning controller 1 is provided with an operation means which performs corrected operation to output a corrected speed command obtained at the time of learning and uncorrected operation to store an inputted speed command in a memory 4 and to output it as the corrected speed command as it is, a measuring means which measures the dynamic characteristic of a motor control system, an editing means which edits various data and parameters of an external information processor, a mode selecting means which designates arbitrary one of learning, operation, measuring, and editing means, a corrected speed command selecting means which designates arbitrary one of plural corrected speed commands, a correction/non-correction selecting means which designates corrected operation or uncorrected operation, a start means which start learning, operation, and measuring means, and a means which outputs the state of the inside of the learning controller.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a learning control device for machine tools, robots, etc. that repeatedly perform predetermined patterns of motion.

0002

2. Description of the Related Art As a learning control device, there is one described in Japanese Patent Laid-Open No. 2-231602, which was previously filed by the present applicant. This device has two modes: an edit mode in which a machining program is input from an external information processing device, stored in memory, and edited; a learning mode in which the machining program is corrected by learning control and the corrected machining program is stored in memory; A learning control device comprising a target operation mode and a manual operation mode to be controlled, mode selection means for specifying any one of the modes, and a plurality of machining programs and correction machining programs. and a starting means for starting the learning mode or the driving mode, and when the mode selecting means specifies the learning mode, the starting means starts the learning mode. In this case, the learning means executes learning control based on the machining program or correction machining program specified by the program selection means until the deviation falls within a predetermined range, and starts when the mode selection means specifies the operation mode. When the means starts the operation mode, the operation means executes the machining program or correction machining program specified by the program selection means.

Therefore, this device can operate with various specifications by combining the functions of the mode selection means and the program selection means, and furthermore, the learning mode and the driving mode can be started independently by the starting means. Therefore, these modes can be activated at any time, and as a result, they can be adapted to various purposes.

0004

[Problem to be solved by the invention] However, since this device corrects the machining program, that is, the motor position command, the correction calculation by learning control must be performed before the position loop, and the learning device itself It was necessary to obtain position commands, have a position loop, and also have manual operation functions, home return functions, etc. SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a learning control device that does not require a target position command, a position loop, a manual operation function, etc. inside the learning control device.

[0005]

[Means for Solving the Problems] In order to solve the above-mentioned problems, the present invention provides a combination of a position control device and a speed control device that executes a position loop of a motor control system that repeats a follow-up operation for the same pattern of target position commands. In the learning control device inserted between the devices, a speed command that is the current position deviation multiplied by a constant is input from an external position control device, stored in memory, corrected by learning control, and then outputs the corrected speed command. A learning means that stores the command in a memory and outputs it to an external speed control device; and an operation with correction that inputs a speed command, stores it in the memory, and outputs the corrected speed command obtained during learning without performing the correction calculation. and an operating means that performs non-correction operation by storing the input speed command in memory and outputting it as a corrected speed command as it is, and outputting a step signal as a corrected speed command by an external command signal and inputting the speed command. a measuring means for measuring the dynamic characteristics of the motor control system by performing the above operations; an editing means for inputting and outputting various data and parameters with an external information processing device and storing and editing them in memory; mode selection means for specifying any one of the editing means; correction speed command selection means for specifying any one of the plurality of correction speed commands; The present invention is characterized in that it includes a correction presence/absence selection means for specifying one of them, a starting means for starting the learning, operation, and measurement, and a means for outputting the internal state of the learning control device.

[0006]

[Operation] With the above means, the position loop control utilizes the existing NC device, and the learning control device itself does not need to acquire a target position command and does not have a position loop or manual operation function. The motor can be operated according to the target position command given to the position loop controller.

[0007]

EXAMPLES The present invention will be specifically explained below based on examples. FIG. 1 is a block diagram of an embodiment of the learning control device of the present invention, and FIG. 2 is a block diagram of the entire control system when the learning control device of the present invention is applied to a motor control system. For comparison, a block diagram of a normal motor control system is shown in FIG. 3, FIG. 4, FIG. 5, FIG. 6, and FIG. 7 show the time of learning, operation with correction, operation without correction, and measurement of the device in FIG. 1, respectively.
FIG. 6 is a conceptual diagram illustrating a method of outputting a corrected speed command at other times. Further, FIG. 8 shows how the speed command decreases due to learning.

The learning control device 1 of this embodiment includes driving means,
Learning means, measuring means, editing means (hereinafter collectively referred to as mode control means), mode selection means, correction speed command selection means, correction presence/absence selection means (hereinafter collectively referred to as selection means).
, a starting means, an internal state output means, a speed command, a corrected speed command, measurement data, and a memory for storing various parameters, and further includes auxiliary means necessary for each mode control means to execute control. We are prepared. Each mode control means functions when the CPU 2 executes a program stored in the ROM 3. Converter 8 is an input/output interface for serial communication with personal computer 13, and converters 9 and 10
is for isolation of the start point interrupt signal and I/O signal from the NC device. Table 1 shows specific examples of I/O signals. The starting point interrupt signal is input to the CPU 2 to control the timing of starting the trial (see FIG. 8). This interrupt signal may be generated using the signal from the motor position detector 23, or when performing synchronized operation with another axis, the signal from the motor position detector of the other axis may be used. You can also make it.

The NC device 11 outputs a value obtained by multiplying the deviation between the target position command and the actual motor position by a constant as a speed command, and this speed command is input to the learning control device 1 by the A/D converter 5. be done. The corrected speed command is output from the D/A converter 6 and input to the external speed controller 12. The timing of input/output by the A/D converter 5 and the D/A converter 6 is determined by a timer interrupt signal from the timer 7 to the CPU 2. When performing synchronized operation with other axes, a signal from an external position detector of the other axis motor may be used instead of the signal from the timer 7. RAM
4 is a battery-backed nonvolatile memory, which includes a speed command memory 30, a measurement data memory 34, a correction speed command memory 31, and others shown in FIGS. 3, 4, 5, and 6. It includes memory for various parameters (not shown). Speed command memory 3 of this embodiment
0 and measurement data memory 34 respectively store speed commands and measurement data input from the A/D converter 5 and parameters related thereto. The corrected speed command memory 31 stores three types of corrected speed commands and parameters related to each.

Mode selection, correction speed command selection, correction presence/absence selection, and startup are performed using I/O signals from the NC device (Table 1
(see Input). The mode selection means sets one of the driving mode, learning mode, measurement mode, and editing mode according to the mode designation signal. The corrected speed command selection means includes a corrected speed command selection switch 32, which is switched according to the corrected speed selection signal. The internal state output means outputs the internal state of the device as an I/O signal to the NC (see output in Table 1).

[0011]

[Table 1]

To perform learning (FIG. 3) with the apparatus of this embodiment, the learning mode is designated with the mode designation signal, the corrected speed command is selected with the corrected speed selection signal, and the start signal is turned ON. When a start point interrupt signal is input to the CPU 2 in this state, a trial run is started. The speed command is inputted by the A/D converter 5 and stored in the speed command memory 30. The CPU 2 performs a correction calculation 33 based on learning control, and the correction result is stored in the specified correction speed command memory 31 and outputted by the D/A converter 6. Note that when storing the data in the memory 31, the correction speed command selection switch is switched in accordance with the correction speed selection signal. Through repeated trials, when the speed command falls below a preset value or when the start signal is turned off, the learning is completed and the state returns to the state shown in FIG. 7.

To perform operation with correction (FIG. 4), specify the operation mode with the mode designation signal, and turn the correction presence/absence selection signal O.
N, select the correction speed command with the correction speed selection signal,
Turn on the start signal. When the start point interrupt signal is input to the CPU 2 in this state, the operation with correction is started. The speed command is inputted by the A/D converter 5 and stored in the speed command memory 30. No correction calculation is performed here,
D/A corrected speed command stored in memory 31 during learning
The converter 6 outputs the signal. When the start signal is turned OFF, the operation with correction is ended and the state shown in FIG. 7 is returned. To perform the operation without correction (FIG. 5), the operation mode is specified with the mode designation signal, the correction presence/absence selection signal is turned OFF, and the start signal is turned ON. In this state, the start point interrupt signal is
When input to , operation without correction starts. The speed command is inputted by the A/D converter 5, and is stored in the speed command memory 30.
The data are stored in the data and outputted as they are by the D/A converter 6. When the start signal is turned OFF, the non-correction operation is ended and the state shown in FIG. 7 is returned.

In order to perform measurement (FIG. 6), the measurement mode is designated by a mode designation signal, and the measurement is started by turning on the start signal. Here, a step command is outputted from the D/A converter 6 as a corrected speed command, and subsequent speed command data is inputted to the A/D converter 5 and stored in the measurement data memory 34. This measurement data is used as dynamic characteristic information of the motor control system during learning. When editing other than the above (Figure 7), the speed command is A/D.
It is inputted by the converter 5 and outputted as it is by the D/A converter 6. In other words, it has the same function as the conventional system shown in FIG. In edit mode, speed commands, corrected speed commands, measured data,
Sends and receives various parameters, etc. This means that personal computers can monitor and change these data.

Further, the speed command correction calculation may be performed in real time while the motor is operating, or may be performed in the background while the motor is operating, or a rest period may be provided between each trial operation, and during that time, The calculations for the next one trial may be performed all at once. The method of calculating the corrected speed command for the next trial during the pause time is to first perform operation with correction, then perform the correction calculation for one trial using the learning calculation signal, and then wait until the learning signal is o.
This can be carried out by repeating the operation of performing the operation with correction after ff. In addition, by performing data transfer from the A/D converter 5 to the memory 30 and from the memory 31 to the D/A converter 6 using a DMA circuit or a separate dedicated CPU, the CPU that performs the correction calculation 33
It is also possible to reduce the load of 2 and shorten the calculation time.

[0016]

As described above, according to the present invention, the position loop control utilizes the existing NC device, and the learning control device itself does not need to obtain a target position command, and the position loop control and manual operation It is possible to operate the motor according to the target position command given to the position loop controller by learning control without having any function or return-to-origin function. Depending on the combination, it is possible to operate with a variety of specifications.Furthermore, since the speed command is directly input to the speed controller when no repetitive operation is performed, the NC device can operate normally without being aware of this learning control device. It has the effect of being able to do the following.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of a learning control device of the present invention.

[Figure 2] Block diagram of the entire control system when the learning control device of the present invention is applied to a motor control system

[Figure 3] A diagram showing a method of outputting a corrected speed command during learning of the device in Figure 1.

[Fig. 4] A diagram showing a method of outputting a corrected speed command when the device in Fig. 1 is operated with correction.

[Fig. 5] A diagram showing a method of outputting a corrected speed command when the device in Fig. 1 is operated without correction.

[Figure 6] A diagram showing a method of outputting a corrected speed command during measurement using the device in Figure 1.

[FIG. 7] A diagram showing a method of outputting a corrected speed command at times other than those shown in FIGS. 3 to 6 for the device in FIG. 1.

[Figure 8] Diagram showing how speed command decreases due to learning

[Figure 9
]Block diagram of conventional normal motor control system

[Explanation of symbols]

1 Learning control device 2 CPU 3 ROM 4 RAM 5　A/D 6 D/A 7. Timer 8, 9, 10 converter 11 NC device 12 Speed control device 13 Personal computer 21 Motor 22 Speed detector 23 Position detector 30, 31 Memory 32 Correction speed command selection switch 33 Correction calculation 34 Memory

Claims (4)

[Claims] [Claim 1] A learning control device inserted between a position control device and a speed control device that executes a position loop of a motor control system that repeats a follow-up operation for a target position command of the same pattern. Learning means inputs a speed command obtained by multiplying the current position error by a constant, stores it in the memory, corrects the speed command by learning control, stores the corrected speed command in the memory, and outputs it to an external speed control device. , operation with correction in which a speed command is input and stored in memory and the corrected speed command obtained during learning is output without performing the correction calculation, and operation with correction is performed in which the input speed command is stored in memory and corrected as is. an operating means for performing uncorrected operation outputting as a speed command; a measuring means for outputting a step signal as a corrected speed command based on an external command signal and measuring dynamic characteristics of the motor control system by inputting the speed command; Editing means that inputs and outputs various data and parameters to and from an external information processing device, stores them in memory, and edits them, and mode selection that specifies any one of the learning, operating, measuring, and editing means. correction speed command selection means for specifying any one of the plurality of correction speed commands;
The present invention is characterized by comprising a correction presence/absence selection means for specifying either the operation with correction or the operation without correction, a starting means for starting the learning, operation, and measurement, and means for outputting the internal state of the learning control device. learning control device. 2. The learning control device according to claim 1, further comprising means for inputting a timing for starting a repetitive operation as an interrupt signal. Claim 3: Any or all of the following: means for performing speed command correction calculation in real time while the motor is operating, means for performing the correction calculation in the background during motor operation, and means for performing correction calculation for one pattern at once when the motor is stopped. 3. The learning control device according to claim 1, further comprising means for controlling the learning control device. 4. The state signal inside the learning control device is a signal indicating that the speed command for one trial is within a preset range, a signal indicating completion of measurement, a signal indicating an abnormality in the learning control device, or 4. The learning control device according to claim 1, further comprising at least one signal indicating that a correction calculation is in progress.