JPS5911921B2 - numerical control device - Google Patents

Description

[Detailed Description of the Invention] The functions of a numerical control device require extremely high processing speed for basic functions such as interpolation calculations and follow-up servo control, and the processing speed is on the order of μsec. , functions such as reading other command data, preprocessing for interpolation calculations of read data, displaying internal data, and sequence control of the operation panel and the machine-side power panel, which is often controlled, are not that fast. There is no need for such processing, and the processing time may be approximately on the order of mSec.

Moreover, recent needs tend to require increasingly high-speed processing for the former function. On the other hand, in the past, numerical control devices were constructed mainly using so-called random logic, which was logically designed according to the purpose, but the cost reduction of small computers meant that instead of the conventional random logic-based structure, small computers were constructed based on numerical Even if it is used in a control device, there is a possibility that it will be cost-effective, and in fact, it is used in a variety of ways.

However, with random logic, various types of information in the numerical control device were processed in parallel at the same time, but when a computer was introduced, it was necessary to process the information once in a while, so small computers (mini computers, micro computers, etc.) ), the problem was that the processing speed was insufficient to realize basic numerical control functions.

As a result, the basic function 15 deteriorates, such as by making the interpolation accuracy rough, or processing that requires high speed relies on an additional arithmetic mechanism constituted by hardware, which results in higher costs. As an example, in Figure 1, the computer reads the command data and pre-processes the read data, and sends the normalized data as a result of the pre-processing to an interpolation unit specially designed by hardware. This figure shows a numerical control device configured to supply the output of a calculation unit as a command pulse to a servo system to a follow-up servo mechanism made of ordinary hardware. The first purpose of this invention is to demonstrate a method for realizing high-speed processing of basic numerical control functions using a computer in order to improve cost performance. The second object of this invention is to show the configuration of a 30 computer that meets these objectives, and to show that this configuration can be realized by ordinary means without using special means.

The effects of this invention can be summarized in that by using such a computer, complicated additional hardware mechanisms can be eliminated as much as possible, high-speed processing of basic numerical control functions can be realized, and costs can be reduced. .

The configuration of the present invention will be explained below with reference to embodiments of the present invention.

The basic configuration of the computer is shown in FIG.

The program to be executed is stored in advance in the 3 main memory, and the program is executed in the J2 main memory or 4 according to the program.
The data in the output section are processed in the first calculation, and the results are stored in the second main storage section L or output from the fourth input/output section.

A series of operations for sending control signals to 2 to 4 in accordance with the contents of the decoded program and regulating the flow of data are performed by the 1 control section. Although the control section can be configured only with hardware, it is possible to provide a storage section within the control section (
ζ2 is also possible in which the control operation is defined by the control memory (which is called a control memory to distinguish it from the main memory).

This method is called a microprogramming method.
In the microprogramming method, control operations are executed by a program (referred to as a microprogram) stored in a control memory, so it is highly flexible, as the architecture of the computer can be changed by changing the microprogram.

On the other hand, since the execution of a program stored in the main memory (called a scroll program) is sequentially controlled by microprograms, it has the disadvantage that the processing speed is generally slower than a control method consisting only of hardware. ing. Now, the instructions used in microprograms (called microinstructions) directly control the internal operations of the computer;
Compared to micro-instructions (referred to as MAG-0 instructions), micro-instructions have a larger number of bits, have more powerful functions, and each individual micro-instruction is executed at a much higher speed.

For example, small computer computers generally have one or several registers for calculations called accumulators, but at the microprogram level, the number of registers that can be directly accessed is increased, and the register access methods and calculations are There are many types. The execution speed of microinstructions is several hundred Nsec to 1 μsec in a computer configured with bibolar elements.

FIG. 3 shows an example of a microprogram type computer.

2CS (control memory) is 1Micr0Pr0gramC
The contents specified by 0unter are given to each part 3 to 7 as a control signal.

By this signal, 50perati0nReji
The contents of ster, 6MS (main memory), or 7I/0 are read out, processed by 4ALUs (arithmetic units), and the results are written again to 5-7. 1Micr0P
r0gramC0unter is 3Micr0Pr0g
ramC0unterC0ntr01 for each microinstruction, and accordingly, MicroOPrO
gram is executed.

FIG. 4 shows an example in which a macro program is executed by a micro program.

Here, in 0perati0nRegister, R1 is the program counter of the macro program, and R2 is the program counter of the macro program.
Assume that we use this as the macro instruction register. By executing the microprogram from A to B, one macro instruction is executed, and in a normal microprogrammed computer, after B, it immediately returns to A and executes the next macro instruction. . Now, assume that the program does not immediately return to A from B, but returns to A after executing another microprogram C after B.

At this time, microprogram C is executed every time one instruction of the macroprogram is executed. As a matter of course,
Macro program execution speed becomes slower. Therefore, it appears that the low-speed macro program and the high-speed microprogram C are executed simultaneously. In this case, microprogram C must change the data that controls the macro program (contents of Rl, R2, etc.) or the data that the macro program directly accesses (usually stored in the MS). For example, a macro program and a micro program C run completely unrelated, but conversely, if a macro program and a micro program access certain data in common, they can be made to run in synchronization. It is possible. This is equivalent to a high-speed computer and a low-speed computer operating independently while exchanging data. Furthermore, the flexibility of the microprogramming method is
To easily create special macro instructions using a microprogram.

When you want to execute a certain process in a macro program at high speed, you can store that process in the CS as a microprogram and create a macro instruction to call the microprogram. From the perspective of a macro program, this can be said to be equivalent to adding a special additional calculation mechanism. In a normal microprogrammed computer, the microprogram only executes the macro program, and the functions of the system are defined by the contents written in the macro program.

However, by assigning part of the system's functions to a microprogram, it is possible to achieve the effect of two low-speed and high-speed computers operating in parallel, or the effect of adding a special additional calculation mechanism. Here, the computer itself is characterized by a microprogramming method rather than a special configuration, and there is no need to add special hardware. Now, in a numerical control device, a micro programmable computer such as the one described above is used to execute numerical control functions that require high-speed processing using microprograms stored in the CS. Let us explain, by giving an example, that if the macro programs stored in the MS are used to execute the functions that are not executed, high-speed processing of basic numerical control functions can be realized and a device with high cost performance can be constructed.

FIG. 5 is an example of such a numerical control device.

The microprogram stored in the CS performs processing such as execution and control of the macro program, interpolation calculations, and the servo system. These processes realize the basic functions of numerical control, and although the program size is small, high speed is required. The macro program stored in the MS performs processing such as reading command data, preprocessing the read data, displaying internal data, and inputting/outputting bit interface data.

The peripheral hardware of a computer is limited to converting input and output signals into a format that can be processed by the computer, and all substantial processing is performed within the computer. Now, how the micro program and macro program are executed will be explained using FIG. 6.

FIG. 6 is an example of a macro program and a micro program when performing linear interpolation.

Reading of command data, preprocessing of the read data, etc. are performed by a macro program, and interpolation calculations are performed by a micro program. Here, Xe and Ye are the end point registers,
X and Y are registers that count the distribution to each axis, D is a register for determining interpolation, and the distribution permission flag, distribution flag, straight line flag, etc. are synchronized between the macro program and the micro program. This is a necessary flag. These data are stored in a data area within the MS and are accessed by both macro programs and micro programs. Now, the interpolation is D
=Xe-Y-Ye-X=0.

Now, assume that the distribution is in the first quadrant (in the case of other quadrants, by rotating in advance in the preprocessing, it can be reduced to the case of the first quadrant.) Now, when X=XiY=Yi, Dij≧ If 0, distribute to x, and Xi+1=Xi+1 Di+1,j=Xe-Yj-Ye(Xi+1)=Di,
If j-Ye or Dij is 0, then distribute to Y, and Yi+1=Yi+1 Di, j+1=Xe(Yj+1)-Ye-Xi=Di,
By performing calculations such as j+Xe, it is possible to distribute while tracing a desired straight line, as shown in FIG.

Assume that the macro program is now in position 4.

The macro program reads command data and preprocesses the read data. At this time, the distribution flag is 0, so the microprogram returns from Δ to ▲.
Does not perform interpolation calculations. Next, assume that the macro program has been executed up to 7. At this time, registers such as D, X, Y, Xe, Ye, etc. can be used by the microprogram. Then, when the macro program executes 8 and the distribution flag becomes 1, the micro program starts executing steps ▲ and thereafter. While the microprogram performs interpolation calculations and distributes as shown in FIG. 7, the macroprogram returns to 1 through steps 1 to 3. There is a variety of information that the numerical control device must know during dispensing (emergency stop, stop bell, etc.), but the macro program monitors this information at relatively low speed, so the micro program can process it at high speed. Can only be involved in processing such as interpolation calculations that require the following. When the microprogram executes the distribution to the end point and executes ``X'', the distribution flag becomes 0, and the macro program then proceeds to step 4 and onward to execute the next program.

In this way, the high-speed program and the low-speed program perform processing independently and complementarily to satisfy the functions of the entire numerical control device.

Microprograms are capable of high-speed processing, but because the number of bits is large, the unit cost per 1W0rd is high.
Conversely, macro programs are relatively inexpensive, but their processing speed is slow. Taking advantage of this feature in numerical control equipment,
Maximum cost efficiency can be obtained by performing high-speed processing of basic functions of numerical control using a microprogram and performing other low-speed processing using a macro program. Having such a configuration with one computer has a great effect.

[Brief explanation of drawings]

Figure 1 shows an example of a conventional numerical control device consisting of a computer and additional hardware.

Claims (1)

[Claims] 1. The commands for executing the basic functions of high-speed real-time numerical control, such as high-precision high-speed interpolation calculations and control of follow-up servo systems, are composed of microprograms and stored in control memory, and reading of command data that does not require high speed is possible. , a microprogrammable computer in which execution instructions for functions such as preprocessing of read data, display of internal data, input/output of bit interface data, etc. are configured as ordinary programs and stored in main memory or external memory. A numerical control device characterized by the following: