JPH07104823A - Numerically controlled working system - Google Patents

Abstract

PURPOSE:To make it possible to perform high-accuracy work from the biginning after the second word by preserving final learnt result data on the side of a numerical controller and performing work while using the preserved latest data at the time of the next work. CONSTITUTION:At the time of the first work, a parameter and a working command are paired and dispatched to a servo system 12 by a numerical controller 11. According to the working command, the servo system 12 starts a working operation with the parameter as the initial value of learning control. The servo system 12 stores the learnt result data after the first learning control provided by the learning control at the time of work in a learnt result storage means 13. When continuously working the same shape, the servo system 12 is controlled based on the learnt result data stored in the learnt result storage means 13 and when the new learnt result data are similarly provided, the contents of the learnt result storage means 13 are updated into those data. Thus, the contents of the learnt result storage means 13 are updated by learning at every time of work.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a numerical control machining method, and more particularly, to the same pattern in which a machining section such as a piston lathe or a cam grinder is non-circular and one round of the next round is repeated in the past. The present invention relates to a numerical control machining method in which a numerical control device is applied to a machine tool that repeatedly performs machining.

[0002]

2. Description of the Related Art Numerical control devices are sometimes applied to applications such as piston lathes and cam grinders. In such a case, the piston or cam to be processed has a non-circular shape such as a shape close to an ellipse, an egg shape, or a diaper shape when the cross section is viewed. When such non-circular machining is performed, it is necessary to control a tool such as a cutting tool or a grindstone to be put in and out at high speed in a direction perpendicular to the spindle according to the rotation of the spindle.

Conventionally, a learning control technique has been applied when processing a shape having a non-circular cross section. Even if the machining cross section such as the piston and cam is non-circular, there is a periodicity in what timing the blade should be taken in and out at a certain rotation position, so if you try trial cutting several times , It becomes clear that the desired shape can be obtained by moving the blade at what timing. Learning control is to store such a processing method and use it for the next processing, and as a result, it is possible to perform processing with higher accuracy than the previous processing.

FIG. 5 is a diagram showing the configuration of a conventional numerical control machining system. In the figure, 1 is a numerical control device, and 2
Is a servo system having a learning control function for machine tools. When machining a non-circle, the numerical control device 1 passes a parameter A representing the starting point of learning control and a machining command to the servo system 2 and passes the parameter A to the servo system 2 in accordance with the machining command. The machining operation is started as a value. The servo system 2 has a learning result storage means 3 for storing learning result data obtained by learning control during machining, and learning result data Aα after the first learning control is the learning result storage means 3.
Stored in. The learning result storage means 3 is composed of, for example, a random access memory (RAM). Subsequently, when machining the same shape, the servo system 2 uses the learning result data Aα stored in the learning result storage means 3.
It is operated based on. At this time as well, learning control is performed in the same manner, learning result data with a smaller processing error is obtained, the contents of the learning result storage means 3 are updated, and the next processing is prepared.

[0005]

As described above, according to the conventional numerical control processing method, the learning result data obtained by the learning control is continuously updated while the processing of the same shape is being performed, and is updated. However, if the machining of that shape is finished and the machine tool is turned off, or if you move to another machining area and start machining of another shape, the learning result obtained in the previous machining area The data was lost. Therefore, it is necessary to restart the learning from the beginning each time when the machining is started or each time when the machining of a different shape is started, and there is a problem that the material required for the trial cutting and the extra time for that are consumed at the time of the machining start. It was

The present invention has been made in view of the above circumstances, and an object thereof is to provide a numerical control machining method capable of reusing the learning result data obtained last time for the next machining of the same shape. To do.

[0007]

In order to solve the above-mentioned problems, the present invention comprises a numerical control device and a servo system having a learning control function, and an initial value parameter and a machining command given by the numerical control device. According to the numerical control machining method in which the servo system operates while performing learning control, the learning result data created by the servo system is taken in at the end of machining, and the latest learning result data previously taken in at the start of machining of the same shape. A numerical control processing method is provided, in which the numerical control device has a learning result storage means for sending to the servo system as the parameter.

[0008]

According to the above-mentioned means, the learning result data learned by the servo system is sent to and stored in the learning result storage means provided in the numerical controller at the end of machining. In the next machining of the same shape, the learning result data stored in the learning result storage means of the numerical controller is used as the parameter of the initial value passed to the servo system together with the machining command.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a configuration of a numerical control processing system of the present invention. In the figure, the numerical control machining system includes a numerical control device 11 and a servo system 12 of a machine tool. Further, the servo system 12 has a learning result storage means 13 for storing learning result data by a learning control function, and numerical control. Device 1
Learning result storage means 14 for storing learning result data in 1 as well.
Have respectively. The learning result storage means 13 of the servo system 12 has a capacity capable of storing the learning result data necessary for one processing, and is composed of, for example, a random access memory (RAM). On the other hand, the learning result storage means 14 of the numerical control device 11 has a capacity capable of storing a plurality of learning result data, and the numerical control device 11
The contents can be retained even if the power of the
For example, it is configured by a non-volatile type memory. In particular, the servo system 12 of the machine tool performs repetitive machining of the same pattern such that the machining cross section such as a piston lathe and a cam grinder is non-circular, and the machining of the next one round is a repeat of the past one round. For example, a linear axis is repeatedly controlled to move in the same pattern in synchronization with a rotary axis to perform machining.

When machining a first shape such as a piston or a cam whose machining cross section is non-circular, the numerical controller 11 passes the parameter A and the machining command to the servo system 12 as a set at the time of the first machining. . The servo system 12 starts the machining operation with the parameter A as the initial value of the learning control according to the machining instruction. The servo system 12 stores the learning result data after the first learning control obtained by the learning control during processing in the learning result storage means 13. Then, when machining the same shape, the servo system 12 is controlled based on the learning result data stored in the learning result storage means 13, and when new learning result data is obtained in the same manner, the learning result storage is performed. The contents of the means 13 are updated with the data. In this way, in the servo system 12, the contents of the learning result storage means 13 are updated after learning every machining, and the machining control is converged to the one having the smallest error.

At the end of processing, the learning result storage means 13 stores the last learned learning result data Aα, and this learning result data Aα is stored in the first learning result storage means 14 of the numerical controller 11. After storing in the storage area for machining the shape, the machine tool is stopped or machining of another shape is started.

Next, when the machine tool is started to machine the first shape, or when another shape is machined and the machine is switched to the first shape, the numerical controller 11 sets the initial value to the initial value. The parameter Aα and the machining command are combined and sent to the servo system 12 of the machine tool, and the servo system 12 performs the learning control based on the previous learning result data Aα and can perform the highly accurate machining from the beginning. .

FIG. 2 is a block diagram showing an embodiment of the present invention. In the figure, reference numeral 11 denotes a numerical control device for controlling a machine tool, which has a learning result storage means 14 capable of storing a plurality of learning result data. The numeral 21 receives various commands to the servo motor output from the numerical control device 11 and passes them to the processor of the digital servo circuit 22, or receives the learning result data from the learning result storage means 13 of the digital servo circuit 22 to perform numerical control. Device 11
It is a shared memory for passing to the learning result storage means 14. The digital servo circuit 22 is a processor, a read-only memory (ROM), and a part of the learning result storage means 13
Which is composed of a random access memory (RAM), etc., which controls the position, speed, and current of the servo motor 24 that drives the linear axis by a processor, and performs processing such as repeated learning control.
It is a signal processor. Reference numeral 23 is a servo amplifier including a transistor inverter, 24 is a servo motor for driving a linear axis, and 25 is a pulse coder as a position detector for detecting the rotational position of the servo motor 24 and feeding it back to the digital servo circuit 22. Further, 26 is a spindle motor that drives a rotary shaft (spindle), 27 is a pulse coder as a position detector that detects the rotational position of the rotary shaft, 28 is a spindle amplifier, and 29 is a spindle control circuit.

Parameters and machining commands from the numerical controller 11 are given to the digital servo circuit 22 via the shared memory 21. In the digital servo circuit 22, the processor makes initial settings at the start of processing. As one of them, the learning result storage means 13 stores a parameter from the numerical control device 11 as an initial value of learning control. The processor obtains a command position based on the positions of the rotary axes and the linear axes from the pulse coders 27 and 25 and the parameters of the learning result storage means 13, and drives and controls the servo motor 24 via the servo amplifier 13.

FIG. 3 is a diagram showing an example of the learning result storage means of the digital servo circuit. Digital servo circuit 2
The learning result storage means 13 of No. 2 makes one round of repeated machining n times.
It is created as a table T showing the correspondence between the divided rotation axis position θi (i = 1 to n) and the linear axis position ri at each rotation position θi, and is updated for each repeated machining. The learning result storage means 14 of the numerical control device 11 is similarly realized in the form of a table, but the learning result storage means 13 of the digital servo circuit 22 is one, whereas the numerical control device 11 is provided. The learning result storage means 14 is created for each processing of a plurality of shapes.

FIG. 4 is an explanatory diagram showing changes in parameters. This figure shows, for example, parameters and machining commands passed from the numerical control device 11 to the servo system 12 when machining three shapes, and learning by the digital servo circuit 22 of the servo system 12 and storage in the learning result storage means 13. And learning result data.

That is, when there is a machining command D1 for the first machining area from the numerical control device 11, it is passed to the digital servo circuit 22 through the shared memory 21 together with the initial value parameter A, and the digital servo circuit 22 produces the first command. The learning result data learned by the last processing by performing learning while processing the processing area becomes Aα and is stored in the learning result storage unit 13. Then, the learning result data Aα is sent to the numerical controller 11 at the time of finishing the machining of the first machining area, and stored in the learning result storage means 14. Next, when there is a machining command D2 relating to the second machining area from the numerical control device 11, it is passed to the digital servo circuit 22 via the shared memory 21 together with the parameter B, and the learning result data at the final machining here becomes Bα. , Are stored in the learning result storage means 14 of the numerical controller 11. Next, the machining command D3 relating to the third machining area is passed from the numerical controller 11 together with the parameter C to the digital servo circuit 22 via the shared memory 21, and the learning result data at the final machining here becomes Cα and becomes a numerical value. It is stored in the learning result storage means 14 of the control device 11.

At the second machining command for the first to third machining areas, the machining parameter D1 and the learning parameter A stored in the learning result storage means 14 in the first machining area.
The set of α, the set of the machining command D2 and the learning parameter Bα in the second machining region, and the set of the machining command D3 and the learning parameter Cα in the third machining region are passed to the digital servo circuit 22, respectively. Learning result data of Aβ,
They become Bβ and Cβ and are stored in the learning result storage means 14 of the numerical control device 11.

Similarly, at the time of the third time, the set of the machining command D1 and the learning parameter Aβ in the first machining region, the second
A set of a machining command D2 and a learning parameter Bβ in the machining area,
Then, the set of the machining command D3 and the learning parameter Cβ is passed to the digital servo circuit 22 in the third machining area, and the respective learning result data becomes Aγ, Bγ, Cγ and is stored in the learning result storage means 14 of the numerical controller 11. The operation of being performed is repeated. In this way, the numerical controller 11
The data stored in the learning result storage means 14 is updated every time, but after the second time, the machining accuracy is converged to a sufficiently high level by the learning control in the first machining, so that the temperature condition is different, for example. There is substantially no significant change in the learning result data except for the change factor due to

[0020]

As described above, according to the present invention, in the numerical control machining method in which the servo motor circuit repeatedly performs the same pattern in synchronization with the rotary axis, the learning control is repeated in the servo system. The last learning result data can be saved on the side of the numerical control device, and at the time of the next processing, the latest data saved can be used to start the processing. Therefore, after the second time, it becomes possible to machine with high accuracy from the beginning, and there is no need for trial cutting.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of a numerical control processing system of the present invention.

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

FIG. 3 is a diagram showing an example of learning result storage means of a digital servo circuit.

FIG. 4 is an explanatory diagram showing changes in parameters.

FIG. 5 is a diagram showing a configuration of a conventional numerical control processing method.

[Explanation of symbols]

 11 Numerical Control Device 12 Servo System 13 Learning Result Storage Means 14 Learning Result Storage Means 21 Shared Memory 22 Digital Servo Circuit 23 Servo Amplifier 24 Servo Motor 25 Pulse Coder 26 Spindle Motor 27 Pulse Coder 28 Spindle Amplifier 29 Spindle Control Circuit

Claims (5)

[Claims] 1. A numerical control machining method comprising a numerical controller and a servo system having a learning control function, wherein the servo system operates while performing learning control according to an initial value parameter and a machining command given from the numerical controller. In, the learning result data created by the servo system is taken in at the end of machining, and the latest learning result data previously taken in at the start of machining of the same shape is sent to the servo system as the parameter. A numerical control machining method characterized by having a numerical control device. 2. The servo system sends the learning result data to the learning result storage means of the numerical controller,
2. The numerical control processing method according to claim 1, further comprising a shared memory that receives the learning result data from the learning result storage means. 3. The numerical control machining method according to claim 1, wherein the servo system is a machine tool that performs machining by repeatedly controlling a linear axis in the same pattern in synchronization with a rotary axis. 4. The learning result data includes a position of the rotary shaft and a position of the linear shaft corresponding to the position of the rotary shaft, and is stored in a table. Numerical control processing method. 5. The numerical control machining method according to claim 1, wherein the table of the learning result data is updated by the learning control function while machining the same shape.