JPH05210408A - High-speed skip system - Google Patents

Abstract

PURPOSE:To improve the precision of track for work processing and to shorten the cycle time by ending the execution of a block. CONSTITUTION:In executing the skipping block, the main CPU being the first CPU buffers the next block during the skipping block execution and transfers the command data of the next block on the shared data. The servo CPU being the second CPU can transfer the command data of the next block to the internal data from the shared data immediately when the skipping is ended. As a result, the distribution can be performed based on the next sampling processing.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention processes a dimension of a workpiece used in grinding or the like while measuring it, and inputs several stages of dimension arrival signals output from a measuring instrument to a numerical controller as skip signals. Change the feed rate with this signal,
To complete processing, that is, skip feed function,
The present invention relates to a high-speed skip system which is a numerical control device having.

[0002]

2. Description of the Related Art Normally, a numerical control (NC) device executes a current block execution when there is a time margin such as waiting for completion of movement so that the execution of the next block can be promptly performed after the execution of the current block is completed. During the process, data for the execution format of the next block is created, that is, buffering is performed. Especially when a block such as a cutting feed (G01) continues, for example, by using the high-speed buffering method disclosed in Japanese Patent Laid-Open No. 3-235102, the first C
A second CPU is used between blocks by using data buffered by a PU (hereinafter referred to as a main CPU).
Since (hereinafter, referred to as a servo CPU) can receive a command from the main CPU without waiting for a time, the servo CPU can move between blocks without interruption of pulse distribution.

[0003]

However, due to the configuration of the system in which the main CPU and the servo CPU have a shared memory,
In the skip-feed NC block, there is a problem that the buffering of the next block and subsequent blocks is suppressed until the execution of the block is completed. It was due to the following reasons. First, on the servo CPU side, if a skip signal is input during distribution in a skip feed block, distribution must be immediately stopped and movement completed, but at that time the target coordinate value instructed from the main CPU side. And that there is a difference between the current command coordinate value on the servo CPU side, and the second
In addition, in the main CPU, if the movement of the next block is an incremental command, for example, the incremental coordinate value from the current command position, that is, the position where the skip signal is input and stopped, becomes the current command coordinate value from the servo CPU side. That is, the coordinate calculation must be performed again based on the above, and the command data for the next block, that is, the command data for the servo CPU, that is, the servo code, must be newly created, and then the command for the next block must be issued. It was

For these reasons, the buffering after the next block is suppressed during execution of the skip feed block. Thus the main C
On the PU side, during skip feed block execution,
The buffering of the next block was to be started only when the completion of the block was recognized without buffering the next block. Especially the main C
When the PU recognizes that the servo CPU has completed execution of the block by inputting the skip signal, the servo C
Based on the current command coordinate value transferred from the PU to the shared memory, that is, the servo status data, the coordinate calculation is newly performed on the main side, and then the buffering is started.

Therefore, as shown in FIG. 5 showing the execution state between blocks in the conventional skip system and between the main CPU and the servo CPU, after the main CPU recognizes the completion of the block of the skip feed, the servo CPU Until the next block move command is issued, the servo CPU waits for a command due to the buffering process of the main CPU, and as a result, it takes 100 to 200 msec until the next block move command is received and distribution is started. There has been a problem that a time lag occurs, which leads to a decrease in trajectory accuracy in workpiece machining and eventually an increase in cycle time.

An object of the present invention is to complete execution of a block after a skip signal is input in a skip-forward block.
A means for eliminating the discontinuity of pulse distribution due to the command waiting time of the servo CPU until receiving the command of the next block from the main CPU and performing the distribution process, improving the trajectory accuracy for work machining, and further shortening the cycle time ( This is to provide a high-speed skip method which is a numerical controller having a high-speed skip method).

[0007]

Therefore, the present invention has solved the above-mentioned problems of the prior art by providing a high-speed skip system described in the claims.

(Operation) According to the high-speed skip method of the present invention, the first C
The main CPU side, which is a PU, buffers the next block while executing the skip feed block, transfers the command data of the next block on the shared data, and the servo CPU side, which is the second CPU side, skips the feed. Immediately after the completion, the command data of the next block can be transferred from the shared data to the internal data, and as a result, the distribution can be performed from the next sampling process. As a result, the servo CPU can perform pulse distribution between the skip feed block and the next block between the blocks without waiting time as in the cutting feed block, improving the trajectory accuracy and cycle time during work machining. Will lead to shortening of.

[0009]

Embodiments of the present invention will be described below with reference to the drawings and flowcharts. FIG. 1 is a block diagram showing the configuration of a numerical controller incorporating the high-speed skip system of the present invention.
1 is a block diagram specifically showing the flow of processing of each CPU in FIG. 1, FIG. 3 is a flowchart showing the interpretation of skip functional blocks and the flow of processing in the high-speed skip system of the present invention in the main CPU of FIG. FIG. 4 shows the servo C of FIG.
3 is a flowchart showing a flow of skip feed processing of a PU in the high-speed skip system of the present invention.

First, on the main CPU side (FIG. 3), when the skip feed function is instructed to the servo CPU, SKF
The ST (high speed skip valid) flag is set, and the next block is set to one block while waiting for the completion of skip feed.
Buffer. After completing the buffering, transfer the command data of the next block to the shared memory with the servo side,
The SDCRQ (next block servo code valid) flag is set.

The servo CPU side (FIG. 4) is the main CPU
The skip feed distribution process is performed based on the command data from. In this case, if a skip signal is input during movement, the distribution is stopped, but if the SKFST flag is set in the current command data and the SDCRQ flag in the shared memory is also set at that time, the next data is immediately transferred from the shared memory. Transfer the servo code of the block. At this time, servo CP
The following calculation is performed inside the U side. (A) When the end point coordinate of the next block is the incremental command, the end point coordinate is calculated by adding the movement amount data from the main CPU to the current command coordinate. (B) In the case of an absolute command, the movement amount data from the main CPU becomes the end point coordinates as it is.

By thus calculating the target coordinates of the next block by the servo CPU itself, the distribution can be performed from the next sampling, and the distribution of the next block can be started without waiting time. Also at this time the servo CP
U transfers the current command coordinate value as servo status data onto the shared memory. The main CPU side performs new coordinate calculation from the present command coordinate value. As a result, when there is a difference between the command coordinates on the first main CPU side, the difference (first target coordinate value-current command coordinate value) is stored in the command coordinates of the next block already buffered. provide feedback. With respect to the buffered command data, the coordinates may be shifted only when the command data is an incremental command, and other created data (for example, the target speed) is valid as it is.

The detection of completion of the high-speed skip on the main CPU side is performed by the SDCAK (completion of servo code transfer) flag instead of using the conventional SKIPED flag (skip completion by skip input) and FGSMEND (completion of distribution) flag. However, this method uses 1 per block
Since this is a shot process, the main CPU side will change to servo C due to changes such as speed override during skip feed.
When it becomes necessary to issue a command regarding the current block to the PU side again, the high-speed skip method is invalid for the skip feed of the current block, which is equivalent to the conventional skip method. In this case, the main CPU is the conventional SKIPED,
The end of block is recognized by the FGSMEND flag.

FIG. 6 shows the execution state of NC blocks between blocks and between the main CPU and the servo CPU in the high speed skip system according to the present invention. Here, Δt2 is the sampling time (distribution period) of the servo CPU, and it can be understood that the servo CPU can perform the distribution at the next sampling when the skip signal is input during the skip feed. On the other hand, for Δt1, after the main CPU recognizes the completion of the N1 block by the SDCAK flag, the current command coordinate on the main CPU side is calculated from the current command coordinate value from the servo CPU,
It is the time from the feedback of the deviation to the command data of the N2 block to the completion of the buffering of the N2 block. By the way, in this system, at the joint of blocks, before the main CPU completes the buffering of the block N2, the servo CPU completes N2.
Distribution of blocks will start, but the main CP
On the U side, the execution of the NC program is not stopped during Δt1, so there is no practical problem. In this way, the servo CPU can distribute the next block in the skip feed block without waiting time.

The data on the shared memory is defined as follows. (Servo code) Flag SKFST High speed skip valid / invalid flag, which is set when the high speed skip method is valid. In the skip feed, the servo CPU side recognizes that the high-speed skip method is effective when this flag is set. SDCRQ Next block servo code transfer valid / invalid flag, which is set when the main CPU transfers the servo code of the next block to the shared memory. The servo CPU can transfer the command data of the next block from the shared memory when this flag is set.

The SKALM skip alarm valid / invalid flag is set when the main CPU side wants to make an alarm when no signal is input (when reaching the target point) in skip feed of the high speed skip system. When this flag is set, the servo CPU side does not transfer the command data of the next block from the shared data at the completion of the execution of the skip feed due to the completion of distribution. SKINC Next block movement incremental / absolute determination flag, which is a flag for the servo CPU side to determine the type of movement of the next block at high speed skip. Set when the movement of the next block is incremental, and reset when an absolute command is issued.

Data NSKP Next block movement amount, which is the movement amount of the next block in the high-speed skip method. When the SKINC flag is set, the servo CPU side calculates the end point coordinates by adding this data to the current command position. Also, SKINC
If the flag has been reset, this data becomes the end point coordinates as it is.

(Servo status data) flag SKPALM This is a skip signal non-detection flag, and is set when no signal is input by skip feed (when the target point is reached). SKALM is set on the main CPU side, and when this flag is set, an alarm can be generated. (Valid when you want to check the arrival of skip feed.) Data SKFSTP Servo C at the servo command coordinates when high speed skip feed is completed.
A command coordinate value when a skip signal is input and stopped while the PU is distributing skip feed. The main CPU side calculates the current command position coordinates on the main CPU side from this data.

[0018]

As described above, according to the present invention, in the block of the skip feed, the skip signal is input, the block is completed, and the command of the next block is received from the main CPU until the distribution process is performed. In this way, there is no interruption of pulse distribution between blocks due to waiting for a command from the servo CPU, and a high-speed skip method that can improve the trajectory accuracy for machining a workpiece and further shorten the cycle time is provided.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of a numerical controller incorporating the high speed skip system of the present invention.

FIG. 2 is a block diagram specifically showing a processing flow of each CPU in FIG.

FIG. 3 is a flowchart showing a flow of interpretation and processing of a skip function block of the first CPU of FIG. 1 in the high speed skip system of the present invention.

FIG. 4 is a flowchart showing the flow of skip feed processing of the second CPU of FIG. 1 in the high speed skip system of the present invention.

FIG. 5 is an explanatory diagram showing execution states between blocks in a conventional skip method and between a main CPU and a servo CPU.

FIG. 6 shows N between blocks and between a main CPU and a servo CPU in the high speed skip system according to the present invention.
Explanatory drawing which showed the execution state of C block.

Claims (1)

[Claims] 1. A first C for interpreting and processing an NC block.
In a numerical control device having a system configuration in which a PU and a second CPU that processes distribution based on a command from the first CPU transfer data via a shared memory, first, in the first CPU, A means for buffering the next block and transferring command data to the second CPU to the shared memory until the second CPU completes the block execution of the skip feed. Next, in the second CPU, in the sampling process of the current block in which the execution of the current block is completed due to a skip signal being input during the distribution of the block for skip sending or the completion of the distribution, Means for transferring movement of the next block quickly from the next sampling by having a means for transferring the command data to the internal data By providing a skip feed block used for grinding or the like, a skip signal is input to complete the execution of the block, and the block from the block to the movement after receiving the command of the next block from the first CPU A high-speed skip method characterized in that processing between blocks is performed at high speed by eliminating the command waiting time of the second CPU.