JPS59172012A - Multi-shaft motor control system using computer - Google Patents

Abstract

PURPOSE:To attain a control system with a high speed and high resolution by adding a binary rate multiplier (BRM) or a digital differential analyzer (DDA) to the input side of a computer as an external hardware. CONSTITUTION:A multi-shaft motor M1 of an industrial robot, etc. is controlled numerically and in a pulse distribution type by means of a computer and through an interface I/O of said computer. In this case, a BRM (DDA) is added via the I/O as a pulse distributing operator for each shaft. The rate and the value to be integrated of each shaft are set to the BRM in response to the block signal. Then each BRM generates a pulse train of each shaft of the motor M1 in accordance with a common clock pulse. These generated pulse trains are counted and integrated by a present position register of each channel via an I/O port. At the same time, the motor M1 is driven through a counter of an output unit, an ROM, etc. and then positioned by said integration value as a terminal point of each block signal.

Description

[Detailed Description of the Invention] When performing pulse distribution type multi-axis numerical control of various automatic machines using a computer, if calculations for pulse distribution or digital control calculations for the machine are performed directly within the computer, the computer's ability will be reduced. Considering this, it is difficult to expect both high resolution and high speed. Therefore, the present invention provides a BRM as external hardware on the input side of the computer.
(BinaryRate Multiplier) or DDA (Digital Differenti)
By adding a sine and cosine transducer that is strong against the absolute position method, the control information can be continuously converted into sine and cosine codes on the output side as necessary, and a sine and cosine transducer that is strong against the absolute position method can be used. Counter and sine cosine conversion RO as external hardware
By adding M, an attempt is made to obtain a high-speed, high-resolution multi-axis control system.

The role of the computer in this case is to store and create data, create block signals for each channel during manual control and programming control, and transmit and receive information to the input/output side.

First, of the arithmetic units required on the input side, the BRM will be described.

The BRM has the role of distributing the number of continuous lock pulses according to the command. The principle of this circuit is shown in FIG. FIG. 1(a) is a block diagram showing the configuration, and giJi diagram (b) is a pulse distribution diagram. The figure shows a case where eight pulses are distributed for simplicity. n flip-flops F1. F2. F3.
...Fn are connected in cascade, and one output terminal of each flip-flop is connected to a gate circuit G1. G2
.

G3. ... are connected, and pulses are distributed by opening and closing these gate circuits. Now, flip-flop F1 receives eight equally spaced pulses 1 and 2.
,3. If . In other words, Figure 1 (a)
As noted above, among the eight input pulses, the odd-numbered pulses 1, 3° 5.7 appear at one output terminal, and the even-numbered pulses 2° 4.6.8 appear at the other output terminal. −
These even-numbered pulses are applied to the next stage flip-flop F.
2, and the odd-numbered pulses are guided to the gate circuit G1. F2. The same goes for F3, and the pulses with the numbers added in the figure are the gate circuits G2. Guided by G3. Therefore, pulses are distributed depending on how these gates 61 to G3 are opened and closed. In other words, as shown in Figure 1 (b), when the open/close states of gates 01 to 3 are set to a rate expressed in binary notation, the number of pulses corresponding to the binary number passes through the gate circuit and is output. It appears in the circuit.

Hereinafter, such an RM will be symbolized and shown as shown in FIG. 2 (a). In the present invention, when setting the rate of a plurality of BRMs to be handled, cents are set according to the MSB side of the BRM based on the rate with the maximum number of digits among them. ).

Next, DDA (Digital Different
tialAnalyzer). Its configuration consists of Y register, R register and Δ
It consists of an X gate. The integrand value is set in the Y register, an arithmetic clock is commanded to the ΔX gate, the contents of the Y register are added to the R register each time, and a carry pulse is output from the R register.

This is 0 or l for each operation. DDA is represented symbolically as shown in FIG. 4 (a). Note that when the external device of the present invention, which will be described later, is used as a pulse distribution calculator, it is used isometrically with the BRM, so the second symbolization representation of the BRM is used.
Match the diagram (b).

When setting the integral value of multiple DDAs, the integral value with the maximum number of digits among them is used as a reference and is set according to the MSB side of the DDA. It is symbolized and shown.

A basic embodiment of the invention is shown in FIG. Although this basic system uses a four-channel BRM as an external arithmetic unit, the number of channels is not limited to this. If you set up BRM from your computer, BR
M outputs a pulse train proportional to the set rate.

The roles played by computers in this basic system can be briefly summarized as follows: ■ BRM rate setting, ■ BR
Output pulse from M as parallel data by interrupt
(2) sine and cosine conversion of the counted data; detecting the end point by counting the number of zero interrupts; and so on. If the above functions were configured only with hardware, it would be very complicated, but
It can be configured easily and cheaply using a commercially available personal computer.

Further, when setting the rate to the BRM, the calculation speed of pulse distribution can be made constant by setting the rate according to the MSB side of the BRM based on the rate of the largest digit among the four channels.

Furthermore, the end point of each channel can be detected simultaneously by counting the number of interrupts. This will be explained using a simple example. Now, each setting value of 4 channels x, y, z, w for the block signal is (x)=11
0. (y) = 100° (Z) = 010, (W
) = 001 (see Figure 1 (b)). In this case, the maximum digit of each signal in this block is 3 digits, so each B
When setting the numerical value of RM (n bits), take the three digits from the MSB and give a rate setting value for each.

In the pulse distribution diagram of FIG. 1(b), (x)0'>(2)(W) is added to each setting rate on the left side to indicate the corresponding setting value. Note that the number of interrupts at this time is 8, which is the maximum value of 3 bits (111) plus 1.

A common clock f is given to the four-channel BRM set in this way. In this case, a pulse of 8111i1 is input. A part of the clock pulse is directly taken into the computer's interrupt register INT, and the output pulses from each BRM are received in parallel to the I10 port for each clock, and the presence or absence of input is checked each time. For example, they are counted and integrated in registers within the computer. The interrupt clock is also integrated at the same time, and when the set value (8) is reached, the interrupt is stopped. By this time, each boat's manpower has received the output pulses of each BRM under the pulse distribution situation shown in Figure 1, and the x, y
, z, and w channel registers accumulate pulses according to the set values. In this case, since the interval between each clock pulse is constant, the output pulse of each channel will eventually change as the displacement speed of each channel increases in proportion to the set rate, using 3 bits (111) = 7 as the standard for constant speed displacement. It becomes something missing. Note that in the above example, the end point determination was determined by the number of clock interrupts, but in this system, the feedback pulses of each channel are counted and integrated, so the rate setting value of a specific channel or multiple channels and the feedbank pulse The end point may be determined by comparing the integrated values.

Next, we will discuss the sine and cosine transformation method on the output side of the basic system. Generally, DC motors and pulse motors are used for NC motor control.

Even in the above case, it is also possible to directly drive the pulse motor with the output pulses of each BRM. However, in automatic machines such as industrial robots to which the present invention is mainly applied, the rotation of each motor is generally linked in a linear motion. Despite that meaning, in such a system, each BR
The output pulses of M are input in parallel to the computer CPU and are constantly integrated, so that the current position of each channel is secured in absolute value form. Therefore, in order to comply with this, it is preferable to use a DC servo motor rather than a pulse motor and synchronize with a linear transducer, for example, to match the absolute value of each channel with the linear transducer using a digital-to-analog converter DA. . However, even with this method, when the linear displacement of each channel is large (this is often the case with robots and multi-axis NC automatic machines), the DA is large and the linear transducer also requires Ai accuracy. However, these drawbacks can be avoided by adopting the sine cosine transducer method.

In this method, a few digits, for example 8 bits, from the LSB of the output displacement pulse integration counter of each channel are converted into circular motion. In the basic system shown in Figure 5, 8 bits (
Indicated by hunting. ) is subjected to sine and cosine code conversion for each channel.The actual processing steps in the computer up to this point are shown in FIG.

The code-converted output of each channel is converted into a sine and cosine voltage wave by a DA. These sine and cosine voltage waves are applied to the following motors Mx, My, M2 .
Sine and cosine transducers Tx, T whose combined gear ratio is set to perform one round motion in 8 bits.
y, Tz, ..., each synchronous computing unit (SYC) χ.

(SYC)y, (SYC)z, -... and servo amplifier S A x + S A y IS A z *
・Used to perform synchronous motion via "'". This operation will be explained using the X channel as an example.

The synchronous arithmetic unit consists of multipliers ml and m2 and a differential amplifier dp. The vector angle of the command sine cosine wave of the X channel is θ, the sine cosine voltage wave is sin θ + case, the vector angle of the sine cosine transducer coupled to the follower motor is φ, and the sine cosine voltage wave is sinφ, cos
Let it be φ. Sinθ and cosφ are multiplied! - to vessel m1, ca
se and sinφ are input to the multiplier m2, and the respective output values sinθ cosφ and case sinφ are both input to the differential amplifier ctp. Here, sjnθ cosφ−case sinφ=sin(θ
-φ)...+11 Find sin(θ-φ), drive the motor Mx via the servo amplifier SA, and allow the phase lag of the tracking delay of φ-θ to be within ±180°. be done. However, in this method, both the follower motor Mx and the sine-cosine transducer Tx are coupled to the main body at a pre-designed gear ratio, so the zero position of the motor and transducer once aligned, The relative positional relationship between and the main body is always maintained.

In the case of the basic system shown in Figure 5, sine and cosine code conversion was performed by software inside the computer, but when high speed control is required, the hardware for pulse integration and code conversion may also be externally attached. is possible.

This high-speed system is shown in FIG. In this case, each BRM
The reception of interrupts to the CPU, the reception of pulses of each BRM, and the accumulation of counts are performed within the computer as in the basic system shown in Figure 5, but the sine and cosine conversion of the lower bits of the counters of each channel is performed within the computer. Don't,
m-bit counter externally connected to each channel in parallel with the output from each BRM, sine cosine conversion ROM, D
The A circuit converts it into a sine and cosine voltage wave. Synchronous motion is then performed using the sine and cosine voltage waves of the sine and cosine transducers coupled to the follow-up motors of the respective channels and the servo unit with a synchronization calculator. Although such external counters and ROM units are provided for each channel, only one channel is representatively shown in the system configuration diagram of FIG. In the figure, °C1Ch represents a channel. This system does not require the use of a computer to create and retrieve the sine and cosine conversion codes, so the entire system can run faster. In addition, in this case, the number of digits m of the m-bit counter of the external device corresponds to one rotation angle of 360° of the sine cosine transducer coupled to the corresponding follow-up motor, and the displacement accuracy is determined according to the frequency division ratio of the transducer. The number of digits m does not have to be the same for each channel in this system, although it is closely related to the following speed of the motor, etc., and there is an advantage that the number of digits m can be set adaptively for each channel.

After all, this system has the advantage of being able to improve the characteristics of the control circuit completely independently of the computer. In addition, both the high-speed system shown in Figure 7 and the basic system shown in Figure 5 use the sine and cosine conversion method, which is strong against the absolute position method. It can be said to be a fine/coarse scale system in which one rotation of the child hand is displayed as one scale and is combined with a parent (coarse) scale that displays the movement of the main body. Therefore, during operation, the movement of the main body is characterized by no zero position fluctuation. However, in order to detect fluctuations in the master scale that occur during initial settings or when a failure occurs, the displacement of the main body is displayed on the master scale using the scale S shown in FIG.
Correspondingly, the master scale in the current position register of each channel in the computer is displayed on the display T,
It is also possible to have a synchronization matching function for comparing S and T. Further, since such parent (coarse) scales and alignment devices do not require high speed detection or high precision display, the addition of these functions does not increase the complexity of the system.

Further, a circular function can be generated using the integral function of BRM. The circular function generation system shown in FIG. 8 can be expressed as follows.

Rearranging for X, x = 2x Solving this, and similarly solving for y, let Δθ, and if the number of bits of BRM is n1, and the clock frequency is f, then it can be expressed as follows.

Therefore, in FIG. 8, x = Asin θ + 3' = A cos θ is given as an initial value to the n-bit BRM of the X channel and the n-bit BRM of the Y channel, and each block pulse is input to the CPU interrupt IN with the clock being f.
Each time the B RM output pulses of the X and Y channels are received from The cosine side (Y coordinate value) is fed to the BRM setting value for the Y channel, and the cosine side (Y coordinate value) is fed to the X channel BRM setting value. The outputs of each BRM operated in this manner are introduced in parallel to each output servo unit, and the X and Y axes perform circular motion. Note that the number of interrupts is expressed as Δθ according to equation (3), so in this case, the number of interrupts is Δθ+
If 1 is set, the circular motion can be stopped at a predetermined end point position.

Moreover, in this case, the speed of the circular motion can be kept constant.

In the explanation so far, the external input terminal is n
We have described the case of using bi,,l-BRM, but this arithmetic unit can be used as an n-bit DDA! Exactly the same thing will hold true even if you replace it. The replacement of both will be explained by comparing the BRM symbolization representations in Figures 2 (a) and 2 (b) with the DDA symbolization representations in Figures 4 (a) and 4 (b). ,
The clock f of BRM is the calculation clock f of DDA, and the clock f of BR
The rate setting of M can be replaced with the integral value setting of DDA, and the pulse train of BRM can be replaced with the overflow pulse train of DDA. In addition, in the multi-channel basic system shown in FIG. 5 and the multi-channel high-speed system shown in FIG. CPU interrupt IN
Each time the pulse train from each BRM is received by T, the pulse train from each BRM is taken in from the I10 boat, the current position is integrated into each channel inside the CP and U, and the number of interrupts is set and used as the end point reference. Multiple n-pi 1-D
When setting the DA integral value, set the integral value from the MSB of the DDA using the maximum digit position as a reference, and accept the DDA calculation clock using the CPU interrupt INT.
The current position is taken in from the port and integrated inside the CPU.
This can be directly replaced by setting the number of interrupts and using it as the end point reference. Also, in the circular function generation method shown in FIG.
It can be replaced with DDA.

[Brief explanation of the drawing]

Figure 1 is a diagram showing the principle of BRM, Figure 2 is a symbolization diagram of BRM, Figure 3 is a diagram showing the principle of DDA, and Figure 4 is a diagram showing the principle of DDA.
DA symbol representation diagram, FIG. 5 is a block diagram showing the basic embodiment of the present invention, FIG. 6 is a diagram showing the processing procedure in the computer of the basic embodiment, and FIG. 7 is the same as shown in FIG. 5. FIG. 8 is a block diagram showing a circular function generation system. Figure 6 Figure 7

Claims (1)

[Claims] 1. In the pulse distribution type multi-axis numerical control system, a pulse distribution computing unit (BRM or DDA) is added to each axis externally through a computer interface, and each axis is controlled in response to a block signal. The rate or integrand value of each BRM
Or set it to DDA, drive the arithmetic unit with the same arithmetic clock to generate a pulse train for each axis, use the arithmetic clock as an interrupt trigger of the computer, and send the generated pulse train for each axis via the I10 port of the interface. Each pulse train is input into a computer, each pulse train is counted and accumulated in a current position register for each channel, and the end point position of each block signal is determined from the number of interruptions or the integrated value of feedback pulses for each axis position. Multi-axis motor control method used. 2. Interrupt acceptance (A
According to claim 1, the other one or more corresponding external devices programmed in advance by 6+ are outputted to the setting section of the arithmetic unit through an interface so as to update the setting values of the arithmetic unit BRM or DDA. Multi-axis motor control method. 3. The lower digit part of the counter-type register that adds and integrates the pulse train of each axis taken into the computer to each axis position is converted into one rotation angle of the sine cosine transducer that connects to each axis follow-up motor of the output power unit for each axis outside the computer. Correspondingly, the outputs of the low-order counting units for each axis are converted into sine and cosine codes and output to each axis through the interface,
2. The multi-axis motor control system according to claim 1, wherein the voltage wave is converted into a sine-cosine voltage wave and synchronized with the stopper sinusoidal wave of the sine-cosine transducer on each axis to operate the heavy machine on each axis. 4. The output signals of each axis arithmetic unit taken into the computer are inputted in parallel to each axis output power unit,
The sine-cosine type transducer connected to the follow-up motor of each axis is guided to the needle number part having the necessary number of digits corresponding to one rotation angle, and the output code converted via the sine-cosine conversion memory is DA. 2. The multi-axis motor control system according to claim 1, wherein a sine-cosine voltage wave is generated through the sine-cosine transducer, and each axis of the heavy machine is operated in synchronization with the rotation of the corresponding sine-cosine transducer. 5. The display within one rotation angle of the sine cosine transducer connected to the electric motor of each axis is used as a child (ellipse) scale, and the amount of movement of the main body is expressed as W.
It is displayed on the L (coarse) scale, and the register inside the computer is also divided into fine and coarse image parts accordingly, and the value of the coarse scale part is output and compared with the parent (coarse) scale without indicating movement of the main body. A multi-axis motor control system according to claim 4, which monitors out-of-step of the scale portion.