JPH04117508A - Position feedback controller - Google Patents

Abstract

PURPOSE:To improve the control gain by defining the feedback interval as a specific interval for the secondary vibration characteristic of a driving system. CONSTITUTION:The vibration frequency detector 110 of a movable part 14 is provided together with a feedback interval setter 112 which sets the sampling cycle of a counter 102. The setter 112 calculates an optimum feedback interval, i.e., an optimum sampling cycle based on the detecting result of the detector 110 and then sets a cycle where a counting signal is supplied to a computing element 104 from the counter 102. In such a condition, the position is compensated through the compensation of the velocity or the current. Then the position feedback control can be effectively carried out with a time error of about pi/3 - pi vibration frequency in regard of the time error. Thus, the control gain is improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a position feedback control device, and particularly to an improvement of the device in consideration of vibrations of a controlled system.

[Prior Art] In recent years, there has been a strong demand for control devices that accurately control the movement of movable parts in various machine tools, X-Y tables, measuring machines, etc. for extremely high-precision machining or measurement.

Conventionally, such a control device includes a movable part (for example, a table, a measuring probe, an arm of a processing robot, etc.) connected to a driving force transmitting part such as a ball screw from a driving part such as a motor through the driving force transmitting part such as a ball screw. Many of them perform predetermined movement control. In this case, in order to determine the movement trajectory or stop position of the movable part, the position of the movable part is detected,
Generally, the position signal is fed back to the drive section to control the position of the movable section.

FIG. 16 shows a conceptual diagram of a general drive device, and each symbol indicates the following content.

Jm, Dm: Moment of inertia of the motor (driver), coefficient of viscous friction Jd, Dd: Moment of inertia of the movable part (including the drive transmission part), coefficient of viscous friction Kd: Rigidity τm of the shaft connecting the motor and the movable part : Torque applied to the rotor of the motor θ-(0 m): Rotational displacement (angular velocity) of the rotor of the motor θd: Rotational displacement of the movable part In the positioning servo system of the low-rigidity mechanism shown in this way, the positioning servo system of the low-rigidity mechanism is shown in FIG. When the target value P0 is inputted as shown in FIG.

In addition, 03w, ω7. is the vibration coefficient of each prefecture, and ζv1ζd is the damping constant of each prefecture.

Then, the actual position p is fed back, and the target value p0
The control is performed so that the actual position p1 and the real position p1 coincide with each other.

On the other hand, FIG. 18 shows a normal positioning servo system for a single motor, and each symbol indicates the following contents.

Gp: Position compensation characteristic Gv: Speed compensation characteristic Gi: Current compensation characteristic L・Inductance of armature winding of motor S: Differential operator R: Resistance of armature winding Kt: Torque constant In the same figure, motor rotational displacement target value θ. is supplied to the motor via the position compensation characteristic Gp, the speed compensation characteristic Gv, and the current compensation characteristic Gi.

In the motor, the armature characteristic 1/(Ls+
The actual current value 11 via R) is fed back and controlled to match the target current value 10.

Further, the actual rotational speed v1 of the motor is outputted via the motor torque constant Kt and moment of inertia Jm, and the actual rotational speed v1 is controlled by speed feedback to match the target speed v0.

Further, the actual rotational displacement θ is outputted based on the integral (1/s) value of the actual rotational speed V of the motor, and the actual rotational displacement θ is set to the target displacement θ by position feedback. controlled to match.

By the way, it is assumed that among the current, speed, and position feedback control systems shown in FIG. Then, each compensation characteristic can be expressed as a constant, and the secondary vibration characteristic (2) can be expressed as follows. Therefore, the basic model for analyzing the position feedback system can be rewritten as shown in FIG.

[Problems to be Solved by the Invention] By the way, in the conventional positioning servo system, since calculation is required to perform position feedback, a control delay corresponding to the calculation time occurs. This control delay is generally replaced by a sampling interval of position feedback information.

However, the method for determining the sampling interval is largely based on empirical rules, and the established theory is that the shorter the better.

However, if the sampling interval is shortened, the calculation speed must be increased accordingly, which requires a high-speed computer.

On the other hand, when considering a realistic positioning servo system, for example, as shown in Fig. 20, the actual trajectory P (solid line) has secondary vibration characteristics with respect to the reference trajectory Pc (dotted line). Suppression is also important for accurate positioning.

However, if the sampling period of the position feedback information is shortened, vibrations may be accelerated, or the control gain cannot be increased, resulting in problems such as a decrease in the responsiveness of the position feedback.

The present invention has been made in view of the problems of the prior art described above, and an object thereof is to provide a position feedback control device that can set an optimal interval for position feedback control and significantly improve control gain.

[Means for Solving the Problems] In order to achieve the above object, a position feedback control device according to the present invention includes a feedback interval setting section that sets a position feedback interval, and the feedback interval is determined by the secondary vibration characteristics of the drive system. It is characterized by a vibration period of π/3 to π.

The present invention also includes a vibration detection section that detects vibrations of the drive system, and based on the detection result of the vibration frequency detection section, the feedback interval setting section sets the feedback interval to be performed at π/3 to π of the vibration period. It is preferable to do so.

Furthermore, in the present invention, the calculation section corrects the sample time delay in the control operation, and the feedback interval setting section
Set the feedback interval to π/2 of the secondary vibration characteristics of the drive system.
It is preferable to set it to 2π.

[Operation] As described above, in the position feedback control device according to the present invention, the feedback interval is π/3 to π of the vibration period, so the control gain can be set extremely large by this feedback interval.

Furthermore, since the feedback interval itself is rather a long period, high-speed calculation is not required, and the control device itself is relatively inexpensive.

Note that by setting the feedback interval using the frequency detection unit so that the feedback interval is set at π/3 to π of the vibration period, it is possible to always properly correct fluctuations in the optimal feedback interval due to load fluctuations, etc. can.

Furthermore, in the present invention, the arithmetic unit corrects the sample time delay in the control operation, thereby making it possible to set the feedback interval even longer.

[Example] Hereinafter, preferred embodiments of the present invention will be described based on the drawings.

FIG. 1 is a block diagram showing the configuration of a position feedback control device according to a first embodiment of the present invention.

The drive system shown in the figure includes a servo motor 1 that constitutes a drive section.
0, a driving force transmitting section 42 that transmits the driving force of the motor 10 via a ball screw, etc., and a movable section 14 that receives the driving force from the driving force transmitting section 12 and performs a predetermined movement.

The motor 10 is connected to a position feedback control device 10.
The drive is controlled from 0 through the summing point 18, the speed compensator 20, the summing point 22, the current compensator 24, and the driver 26.

On the other hand, a current detector 28 is provided at the output end of the driver 26, and a negative signal is applied to the summing point 22 for current feedback.

Further, the rotation of the motor 10 is detected by a tachometer 80, and a negative signal is supplied to the summing point 18 for speed feedback.

Furthermore, a position signal of the movable part 14 based on the drive of the motor 10 is detected by a scale 32 (position detection section), and is inputted to the position feedback control device 100 as a scale signal.

This position feedback control device 100 performs position control processing using software using a microcomputer or the like.
Addition point 18, speed compensator 20, addition point 22
, the current compensator 24 and the driver 26 perform hardware speed control and current control, and these software and hardware processes perform drive control.

Here, the position feedback control device 100 has a counter 1
02, arithmetic unit 104, holder 106, D/A converter 1
08, the scale signal from the scale 32 is counted by the counter 102, and after the desired calculation is performed by the arithmetic unit 104, the position control data is temporarily held in the holder 106, and the D/A converter 108, the signal is fed back to the drive system as an analog signal.

Note that when new position control data is output from the computing unit 104, the holder 106 sequentially updates its stored contents.

The characteristic feature of the present invention is that the control period is set to an optimal value according to the vibration period of the drive system, and the control gain is greatly improved. number detector 110 and counter 102
A feedback interval setter 112 is provided to set the sampling period of the .

Then, the feedback interval setter 112 calculates the optimum feedback interval, that is, the sampling period in this embodiment, based on the detection result of the vibration frequency detector 110, and sets the period for supplying the count signal from the counter 102 to the calculator 104. .

Here, the physical system considering the sampling period is the 19th
As shown in Fig. 2, the one sample delay characteristic is: e -aT.

A system in which zero-order holder characteristic: (1-e-“”)/S is added is considered. Note that TS is a sampling interval.

If we consider a dimensionless model that normalizes this, it will be as shown in Figure 3.

Note that the dimensionless parameters are as follows.

Normalized proportional gain: ni=ni/ωn Normalized sampling interval: τ5=Ts·ωn Normalized Laplace operator: s'=s/ωn As a result, the following is replaced.

Based on the model shown in Figure 3 above, the relationship between the normalized sampling interval τS and the normalized proportional gain is expressed by each attenuation constant ζ
The results of the simulation are shown in FIG.

As is clear from the figure, the damping constant ζ = 0-0゜25
There is a sampling interval that greatly improves the gain of the stability limit specified for continuous systems within a range of It is understood that determining the sampling interval to achieve that maximum gain.

Although it is difficult to conceptually explain the relationship between the sampling period and the gain, it can be understood as follows, for example.

In FIG. 5, the horizontal axis Pc is a target reference locus, and it is assumed that an actual locus P is drawn by generating a sinusoidal waveform vibration as shown by the solid line in the figure. In addition, in the figure, speed changes are shown by dotted lines, and acceleration changes are shown by dashed-dotted lines.

Then, at time t, a comparison is made between the reference trajectory and the actual trajectory, and the positional deviation d is fed back.

Conventionally, this positional deviation d has been fed back as quickly as possible.

Here, as is clear from FIG. 1, compensation for the positional shift is ultimately performed at the current level and realized as a change in the acceleration of the motor 10.

That is, in the conventional case, the acceleration a in the downward direction in FIG. 5 is applied at time t, or at time t, which is a short time later.

However, at time t, the acceleration change corresponding to the upward position change in the figure is downward in the figure, and at time t,
Adding the acceleration as in the downward direction in the figure is
This acts to further increase the change in acceleration. For this reason, it is possible that vibrations will be encouraged, and that accurate positioning will be hindered.

Therefore, in a conventional system in which the feedback interval is controlled to be as short as possible, it has been impossible to obtain a large control gain.

On the other hand, when the feedback interval is set to π/2 as in the present invention, the positional deviation d at time t1
, at time 1. This will have to be adjusted.

This time 1. Hereinafter, the acceleration change is upward in the figure, and applying the acceleration al downward rather acts in the direction of suppressing the change in acceleration.

In reality, the position compensation system is performed via speed compensation or current compensation, and if the time lag is taken into consideration, it is extremely effective to perform position feedback control with a time lag of about π/3 to π of the vibration value. One thing is understood.

FIG. 6 shows the relationship between the normalized proportional gain and the normalized sampling interval τS when the attenuation constant ζ=0.02.

FIGS. 7 to 13 show the results when position feedback control is performed at each sampling period and gain indicated by ■ to ■ in FIG. 6.

As shown in Figure 6 (■), when position feedback control is performed with a gain and feedback interval outside the normalization diagram, as shown in Figure 7, even though the sampling interval is short and the gain is not very large. Vibration increases and accurate positioning cannot be achieved.

On the other hand, when sampling is performed at a gain and a feedback interval within the frame of the normalization diagram as shown in Fig. 6 (■ to ■), even though the gain is the same as in (■) above,
As shown in FIGS. 8 to 11, vibrations are suppressed and accurate positioning is achieved.

Furthermore, if the normalized sampling interval is set to about 1.7 as shown in Figure 6, Vibration suppression and positioning control are performed.

In this way, in this example, by setting the sampling interval to π/3 to π of the vibration period of the secondary vibration characteristic,
The gain is significantly improved, making it possible to quickly control positioning, improve steady-state characteristics, and improve disturbance suppression characteristics.

Further, since the sampling interval is relatively long, the time required for calculation can be obtained with sufficient time, and it becomes possible to perform highly accurate position feedback control using a relatively inexpensive computer.

In addition, since the frequency detector 110 is used in this embodiment, when it is used for position control of a robot arm, for example, it is possible to accurately correct changes in secondary vibration characteristics based on changes in the load. can. It is preferable to perform this frequency detection at regular time intervals or at the time of starting the drive system.

Furthermore, if the load does not fluctuate frequently, there is no need to intentionally provide the frequency detector 110.

Next, the inventors investigated the sample delay and the optimal feedback interval.

In the first embodiment, the one-sample delay characteristic remained, but it was found that by compensating for this one-sample delay characteristic, the feedback interval could be further lengthened.

That is, FIG. 14 shows a diagram corresponding to FIG. 3 of the physical system of the position feedback control device according to the second embodiment of the present invention, and FIG. 15 shows each attenuation constant ζ when sample delay is eliminated. An explanatory diagram showing the relationship between the normalized proportional gain and the normalized sampling interval is shown.

As is clear from FIG. 14, the physical characteristics of the device according to this embodiment eliminate the following shown in FIG.

As a result, as shown in FIG. 15, as in FIG. 4, when the damping multiplier ζ is small, the optimal feedback interval tends to become longer. By increasing the length to 2 to 2π, the gain of the servo system can be improved. Also, ζ=0
Even in the case of (no damping), the control system can be stabilized by setting =π to 2π.

Among these, the sampling period that maintains a significantly large gain is τ,=nπ/m.

Also, the value of τ that maintains the largest gain margin is approximately 3
．． 14, it is understood that even if one sample delay is compensated for, a large gain margin cannot be obtained unless the feedback period is doubled or more.

As a means for compensating for the one-sample delay, a method of feeding back the output of the controller, a method of compensating by detecting the instantaneous speed by extrapolation, or a method of using other known software or hardware processing can be used.

Here, the method of compensation by detecting the instantaneous speed by extrapolation is
This is particularly effective when the signal detection method itself includes dead time.

In other words, when detecting the rotational speed using an encoder (by counting the number of pulses during one sampling period, etc.), it is the average speed of one sampling period, which is approximately 1/1 of the instantaneous speed at the sampling time. It is delayed by two sampling periods. To compensate for this, 2
It is sufficient to extrapolate from the average speed ω1, ω, by ω(k+1)=ω, +(ω, −ω, )/2.

Furthermore, a method of extrapolating from three average speeds is also suitable.

[Effects of the Invention] As explained above, according to the position feedback control device according to the present invention, since the feedback period is set to be π/3 to π of the vibration period of the secondary vibration characteristic, the gain will be significantly improved.

Further, by correcting the sample time delay in the control operation using the arithmetic unit, it is possible to further lengthen the feedback interval.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a position feedback control device according to an embodiment of the present invention, FIGS. 2 and 3 are explanatory diagrams of the physical system of a position feedback control device having a predetermined sampling period, and FIG. 4 is an explanatory diagram showing the relationship between the normalized proportional gain and the normalized sampling interval at each damping constant ζ, FIG. 5 is a detailed explanatory diagram of the present invention, and FIG. Fig. 7 to Fig. 13 are explanatory diagrams of the vibration suppression states in ■ to ■ shown in Fig. 6, Fig. 14 is an explanatory diagram of the physical system when one sample time delay is compensated, Fig. 15 is an explanatory diagram showing the relationship between normalized proportional gain and normalized sampling interval at each attenuation constant ζ when one sample time delay is compensated, Fig. 16 is an explanatory diagram of a general drive system, and Fig. 17 The figure is an explanatory diagram of the dynamic characteristics of the system shown in Fig. 16, Fig. 18 is an explanatory diagram of the positioning servo system, Fig. 19 is an explanatory diagram of the basic model for analysis, and Fig. 20 is an explanatory diagram of the conventional position feedback control device. It is an explanatory diagram of the problem. 10... Servo motor 12... Driving force transmission section 14... Movable section 100... Position feedback control device 102...
Counter 104... Arithmetic unit 106... Holder 8... D/A converter 0... Frequency detector 2... Feedback interval setter

Claims (3)

[Claims] (1) A position feedback control device that includes: a position detection section that detects the position of a movable part; and a calculation section that computes a position correction amount based on the detection result of the position detection section; A position feedback control device comprising an interval setting section, wherein the feedback interval is π/3 to π of a vibration period of a secondary vibration characteristic of a drive system. (2) The position feedback control device according to claim 1, further comprising a frequency detection unit that detects the frequency of the drive system, and based on the detection result of the frequency detection unit, the feedback interval setting unit adjusts the feedback interval to the vibration period. A position feedback control device characterized in that the position feedback control device is set so that the feedback is performed at π/3 to π. (3) In the position feedback control device according to claim 1 or 2, the calculation section corrects a sample time delay in the control operation, and the feedback interval setting section adjusts the feedback interval to π// of the secondary vibration characteristic of the drive system. A position feedback control device characterized in that the angle is 2 to 2π.