JPH0115881B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention enables smooth operation of a mobile machine having a bump in the power transmission part between an actuator (drive part) and a working end (load part), and The present invention relates to a control method that enables highly accurate control.

[Prior art and its problems]

As is well known, in mechanical devices, bumps occur to varying degrees.

This is due to the mechanical causes of movement of each part (for example, the involute tooth shape of the gear, the thread shape of the feed screw rod, etc.) and the machining accuracy, but due to the existence of this backlash, the following This is a major problem.

(1) Since the positioning accuracy cannot be lower than the backlash width, it has not been possible to realize a device with higher positioning accuracy.

(2) Batch crash is a factor that degrades stability in closed-loop control, so the loop gain can only be set to a very small value, resulting in poor response and hunting.

(3) Collisions always occur in the backlash section when starting or stopping, which significantly shortens the life of the power transmission section.

Traditionally, these problems have been dealt with by taking special anti-backlash measures into the power transmission section. In addition to taking such measures, the servo mechanism has to form a feedback loop with only the actuator to control the backlash outside the closed loop. A widely used method is to perform correction as a static state by adding up the position command value only when the operating direction changes.

However, although such a method can avoid the above (2), it is not a solution to the problems (1) and (3).

In addition, as a control method for a control system having a backlash, especially a device that drives a moving body on the load side through a gear device, a second drive system that is driven through the same type of gear device without a backlash is installed. Japanese Patent Publication No. 56-10646 discloses a control method in which the difference in drive amount between the two drive systems is detected without contact, and the difference is fed back to the control circuit of the servo motor to eliminate backlash.

However, with this control method, although it is possible to automatically correct the target value for changes in the target value in the positive and negative directions, there are Since collisions are not taken into consideration, this control method has the problem of (3) above.

[Means for solving problems]

The present invention has been made in order to solve the above problems, and in an apparatus in which there is a collision between a movable body A on the driving side and a movable body B on the load side, the respective positions of both movable bodies A and B are A device for detecting x and speed Find the boundary line that shifts the time required for By adding a function to perform gap control, it is possible to control a mobile machine having backlash without any trouble and with high precision.

For convenience of explanation, we will first consider a mobile machine in which there is no backlash in the power transmission section.

Now, when moving a load by driving an actuator from an arbitrary point P on the phase plane of position x and velocity According to theory, first accelerate with the maximum amount of operation. Then, the moving body moves along the acceleration curve f _b4 determined by the differential equation of motion.

When the position x and speed x of the moving body reach the intersection Q with the deceleration curve f _a2 passing through the origin (target point), the operation amount may be set to 0 when deceleration is started and the origin is reached.

In addition, when the load is only inertia and friction etc. are ignored, if the moment of inertia of the moving part is J, the operation amount of the actuator is u, and its maximum value is U _nax , then f _b1
The ~f _b5 curve and the f _a1 ~ f _a5 curve are parabolas expressed by the following equations, respectively.

X ₌ ^J /2u _nax X〓 ² X= _-J /2U _nax

However, if the power transmission section has a buckle bump and there is no gap due to buckle bump on the acceleration side at startup, if deceleration control is performed using such a deceleration switching curve, the actuator side will start decelerating from the point of switching, but the load side will start decelerating. tries to maintain the previous state due to inertia, so it decelerates while causing a collision within the gear due to back crash.At the same time, when the actuator reaches the home position, the amount of actuator operation is set to 0 and the load position is changed. does not exactly reach the origin.

Furthermore, if the gap caused by the backlash is on the acceleration side at the time of starting, a collision due to the gap further occurs at the time of starting.

Therefore, in a mobile machine with backlash,
Before shifting to deceleration control, the gear on the deceleration side is controlled to be set to -L/2, and if the gear is on the acceleration side at the time of starting, the gear is controlled to be set to L/2 first. By this,
This means that it can be moved to the target position accurately in the shortest possible time.

The present invention is based on this principle, and FIG. 2 shows a block diagram when performing backlash control of a power transmission mechanism having backlash. In the figure, 1 is a drive unit that drives the load unit 3 via the feed screw rod 2, 4 is a position and speed detector for the drive unit, 5 is a position and speed detector for the load unit 3,
Reference numeral 6 denotes an arithmetic unit that performs arithmetic and logical judgments on how to perform appropriate control based on the detection signals of the detector, 4 and 5, and 7
Reference numeral 8 indicates an actuator control unit that sends out a command signal based on the calculation result of the calculation unit 6, and 8 is a minor loop.

Here, the backlash portion H of the power supply section in FIG. 2 is schematically expressed as shown in FIG. 3.

The symbols in the figure are as follows.

X _Lr : Load part target position X _L : Load part position X _M : Drive part position e: Position difference (X _{M −} Any initial point (position,
Consider the case of moving to the target position by speed).

In the case of FIG. 3, since there is a gap e on the acceleration side (right side), first, this must be set to L/2 to integrate the moving body A on the drive side with the moving body B on the load side.

This gap control can be integrated in the shortest possible time without collision by applying the shortest time control based on the aforementioned Pontryagin's theory.

In other words, an actuator composed of a DC motor or the _like can be controlled by using an appropriate _{control circuit. J M :} Moment of inertia of the motor (For example, voltage) It is possible to perform a response expressed by a differential equation.

The shortest time control of the system expressed by equation (1) can bring the gap e to 0 in the shortest time by performing acceleration and deceleration using the limit of the manipulated variable according to the Pontryagin's logic.

The acceleration switching curve f ₇ and deceleration switching curve f ₆ in this case are expressed by the following equations.

f ₆ : e=-J _M /2u _nax e〓 ² +L/2, e〓≧0 f ₇ : e=J _M /2u _nax e〓-L/2, e〓<0 However, u _nax is the maximum operation amount , e〓 is the speed difference X〓 _M −X〓 _L This is illustrated in Fig. 4.

Therefore, control from an arbitrary state C (e ₀ , e〓 ₀ ) to the right end (L/2, 0) where the gap becomes 0 is to accelerate from C to D and decelerate when D is reached. Once started, it reaches (L/2, 0) without going too far along the deceleration switching curve _f6 . In other words, moving body A can be integrated with moving body B without any shock.

The specific content of the relative motion of the moving body A with respect to the moving body B when transitioning from acceleration to deceleration shown in FIGS. 4 and 7 will be explained.

Now, assume that the gap state of moving body B at time t=0 is ( _XL , _X〓L ) = ( _X0 , _V0 ) (e, e〓) = (L/2, 0). Since e=X _{M −XL} , the state of moving body B is (X _M , X〓 _M )=(X ₀ +L/2, V ₀ ). From this state, contact on the deceleration side (-L/2,
0), it is assumed that u=-u _nax is selected. From this moment on, -L/2<e<L/2, and moving bodies A and B separate and perform independent movements. No force acts on the moving body B, and it moves at a constant velocity. At time t ₁ , the state is (X _L , X〓 _L ) _t=t1
= ((X ₀ +V ₀ t ₁ ), V ₀ ). For moving body A, u=
_{−u nax ) acts , so ( X M ,} ^_ _{_ _ _} t ₁ )]. The state of the gap is now: (e, e〓) _t=t1 = [(L/2−u _nax /2J _M t ₁ ² ), −u _na
x / J _M t ₁ )]. As can be seen from FIGS. 4 and 7, at the time when e=0, u is switched from -u _nax to +u _nax . Therefore, at time t _s when t _s =√ _{M nax} , it switches to u=+u _nax .

The movement of the moving body after _this is (X _L , X〓 _L ) _t=t2 = [(X ₀ +V ₀ t ₂ ), V ₀ )] ( _{X M} , X〓 _M ) _t=t2 = [{X ₀ +V ₀ t ₂ −u _nax /J _M t _s (t ₂ −t _s
) +u _nax /2J _M (t ₂ −t _s ) ² }, {V ₀ −u _nax /J _M t _s +u _nax /J _M (t ₂ −t _s )}], and the gap is (e, e 〓) _t=t2 = [u _nax /2J _M {(t ₂ −t _s ) ² −2t _s (t ₂ −t _s )}, u _nax /J _M (t ₂ −2t _s )].

t ₂ = 2t _s = 2√ _{M nax} , (e, e〓) _t=t2 = [u _nax /2J _M (t _s ² −2t _s ² ), u _nax /J _M (2t _s −2t _s ) ]=(−u _nax /2J _M t _s ² , 0)=(−L/2, 0). As mentioned above, (e,
e〓) If the manipulated variable u is switched according to the phase plane trajectory, It is clear that transitions can be made between the two without conflict. Note that it is clear that it is possible to transition to a similar contact state for the initial state (e ₀ , e〓 ₀ ) of any internal gap surrounded by f ₆ and f ₇ , and this control method Then, exactly the same result is obtained when the load has an initial velocity.

After integrating in this way, in order to accelerate the moving body A on the drive side again, then decelerate it and move it to the target position, in a device having a backlash, the deceleration switching curve f _a2 (f ₂ ) By the time the above point is reached, the moving body A must be engaged with the moving body B at the deceleration side end of the gap.

In other words, the acceleration side gap is 0 (L/2, 0) , the gear on the deceleration side is 0 (-L/2, 0) It is necessary to perform gap control to move the

In Fig. 4, the gap control is such that if you first decelerate from the state of (L/2, 0) and then start accelerating when you reach point E (intersection with _f7 ), it will not go too far (-L). /2,0).

The maximum time required to change these gap conditions is or vice versa. It's time to move on.

The required time T ₀ at this time is T ₀ = (1+√2), √2 _{M nax} .

Note that the area surrounded by the curves f ₆ and f ₇ in FIG. 4 is the gap controllable area, and it is possible to set the area necessary to avoid any practical problems when selecting an actuator.

The above-mentioned time T ₀ is a gap control time for changing the backlash state necessary before shifting to integrated deceleration control of moving bodies A and B.

Therefore, in a mobile machine having a backlash, the time to switch to deceleration control must be advanced by this time _T0 .

Here, for any speed (X〓 _Li ) of the load, T ₀
Figure 5 shows the change in distance traveled over time.
_g It becomes like a straight line.

Therefore, the deceleration or acceleration switching curve ₁ , f ₃ in the case of a mobile machine with backlash is the same as the deceleration or acceleration switching curve f 2 , f ₃ shown in FIG. 1, as shown in FIG.
f ₄ , the time T ₀ shown in Figure 5 is earlier.
All you have to do is shift them to the left and right by an amount of _g1 . Hereinafter, f ₁ and f ₃ will be referred to as auxiliary switching curves.

If the moment of inertia when moving bodies A and B are integrated is J _LM = L _L + J _M , then the equation of motion is expressed as J _LM・X¨ _L = u, so the above switching curves f ₂ and f ₄ are f ₂ :X _L = _{J LM} /2u _nax X〓 _L ² _{X〓 L} ≧ ₀ f ₄ : X _L = _J LM /2u ^nax

When time T ₀ elapses without any external force acting on point R (X _L , X〓 _L ) on the f ₁ auxiliary switching curve, the moving body moves to point S (X' _L , X〓 _L ) on the f ₂ auxiliary switching curve. reach.

From this _, when the moving bodies of the drive section and the load section are under acceleration control as _a unit, when ( _XL , If gap control is performed to set the value to 0, and deceleration control is performed along the _f2 switching curve after _T0 time, accurate movement control to the target point will be performed. If the control is from deceleration to acceleration, the _f3 auxiliary switching curve will be used.

To facilitate understanding, the above series of controls will be explained using FIG. 7 as follows.

Figure 7a shows the movement of moving bodies A and B.
is in an expiration state. Here, moving body A... (X _M , X〓 _M ), = (X _MO , X〓 _MO ) Moving body B... (X _L , X〓 _L ) = (X _{LO ,} (X _Lr , X〓 _Lr )=(0, 0).

First, since (e, e〓) is not (L/2, 0), that is, there is a gap L on the acceleration side, the drive unit moves as described above by performing the acceleration → deceleration gap control. Shift body A to (L/2, 0), that is, set the gap to 0.

This situation is →→ in FIG. 7b.

In this state, the drive section and the load section work together to control the movement of the load, and the shift progresses from →' as shown in FIG. 7c.

When the f ₁ auxiliary switching curve is reached, control is then performed to change the state of the gap, that is, control to set the gap on the deceleration side to 0 is performed, and the shift shifts to '→.

In this way, when we reach 'c' in Figure 7,
This time, deceleration control is started, and first, the gap on the deceleration side is brought to 0 by the deceleration → acceleration control of '→→ in FIG. 0) is reached.

The manner in which the manipulated variables are given in the series of controls described above is shown in FIG.

As is clear from the above description, what is necessary to implement the present invention is to determine the switching curve and to determine the deceleration point (acceleration point in the opposite case) of the actuator based on the load condition. The conditions for the switching point are summarized and listed as follows.

f ₆ : e=-J _M /2u _nax・e〓 ² +L/2e〓＞0 f ₇ : e=J _M /2u _nax・e〓 ² −L/2e〓<0 f ₂ :XL= _{-J LM} ／2u _{nax ・ X〓 L} ² _{X〓 L} ≧ _{0 f 4} ^： X _L ＝J _{LM ／ 2u nax} ² _{−ToX L} ^X〓 _{L ≧ 0 f 3} :XL = J _LM /2u _nax As a general-purpose control device, several types of inertia moment and operation amount can be calculated in advance and prepared, and during actual use, they can be selected manually, for example with a setting pin, or automatically based on changes in position and speed. It is desirable to configure the system so that it can be selectively selected.

In addition, each switching curve may be replaced with a broken line approximate curve in practice.

Whether or not the position and speed of the moving body have reached the switching curve may be determined when the difference from the switching curve becomes small.

FIG. 9 shows an embodiment in which the present invention is applied when a load 12 is driven by a motor M via a power transmission mechanism 11 with bumps, and 13 is a speed detector connected to the motor M. TG ₁
and the detection signal of the position detector PG ₁ using a pulse generator, the detection signal of the speed detector TG ₂ connected to the load 12 and the position detector PG ₂ using a pulse generator, and the preset or detected signals. selected by signal
A judgment device consisting of a minicomputer or the like sends switching signals for switching devices 14 and 15 from the f ₁ , f ₂ , f ₃ , f ₄ , f ₆ , f ₇ cut pipe curves, and 16 performs gap control on the acceleration side. 17 is a gap width setting device; 18
1 is a command device that sends a command signal d to the target position after the load side position/velocity reaches the f ₂ or f ₄ switching curve; 19 is a target position setting device; This is a drive device that applies a quantity u) to a motor M to drive it.

A determination flowchart for obtaining switching signals of the switching devices 14 and 15 in the determination device 13 is shown in FIG.

Now, when the gap due to buck crush is on the acceleration side at the time of starting, first, the switching device 13 is switched to the 14b, 15b side to control the acceleration of the motor M, and then the switching device 15 is switched to the 14b, 15b side.
It is switched to 5a and the acceleration is controlled and the gap is 0.
become.

Next, the switching device 15 is switched to 15b, and the motor M is accelerated. When the position/velocity reaches the value on the _f1 auxiliary switching curve mentioned above, the switching device 15 switches to 1.
When the switching device 15 switches to the 15b side and enters deceleration control, and the position and speed of the drive side input to the judgment device 13 match the values on the _f7 switching curve, the switching device 15 switches to the 15b side and shifts to acceleration control. The gap on the deceleration side is set to 0.

At this time of gap 0, the position and speed of the load side
Since X _L and X〓 _L reach the f ₂ switching curve, the switching device 14 is now switched to 14a, and deceleration control is performed until the target value is reached.

When there is no backlash and there is a limit to the speed of the load, the acceleration/deceleration switching curve will be as shown in FIG. 11, different from that shown in FIG.

In such a case, the strategy for reaching the origin from an arbitrary point P (x, x) is to set the manipulated variable to 0 midway and add control to maintain a constant speed V _nax .

The same thing holds true for the control of a system with backlash, so f ₁ ,
By setting a function only within a certain speed limit as the switching curve of f ₂ , f ₃ , and f ₄ , the control method of the present invention can be used as is by setting the manipulated variable to 0 when the speed limit value is reached.ゝCan be applied.

The control method according to the present invention is most effective when no frictional force acts on the moving body, but if there is a frictional force or other uncertain factors, the control method of the drive device 20 in the embodiment The system of the present invention can be effectively utilized by changing the characteristics to those shown in FIG. 12 (linear control).

In other words, by providing a certain degree of linear range, control can be performed effectively by giving a manipulated variable proportional to the difference between positive and negative values of "d" without adding the maximum manipulated variable. become.

As described above, according to the present invention, even if there is a backlash, a collision does not occur at the backlash portion, and
Therefore, the life of the machine can be greatly extended, and there is also a manufacturing advantage in that there is no need to require unreasonable precision in machine assembly, and the control system is completely stable even if the backlash is in a closed loop. This is an industrially excellent invention, as it is guaranteed that the

[Brief explanation of drawings]

Fig. 1 is a diagram showing acceleration and deceleration switching curves in a moving device without backlash, Fig. 2 is a block diagram of the control system of a power transmission mechanism with backlash, Fig. 3 is a model of the backlash section, and Fig. 4 is a diagram showing a control system of a power transmission mechanism with backlash. The figure is an explanatory diagram of the gap control method using backlash, and Figure 5 shows the speed of the load
Diagram showing the relationship with the distance traveled in T ₀ time, 6th
The figure shows the relationship between the acceleration/deceleration switching curve and the auxiliary switching curve, Figure 7 is an explanatory diagram showing a method of controlling a moving device with backlash, and Figure 8 is a diagram showing an example of how to give a manipulated variable. , FIG. 9 is a block diagram of an embodiment of the present invention, FIG. 10 is a flowchart of the judgment device, FIG. 11 is a diagram showing a control method when there is a limit on the load speed, and FIG. 12 is a diagram showing a control method when the load speed is limited. FIG. 2 is a block diagram showing an embodiment. M... Drive motor, 11... Power transmission mechanism with bumps, TG ₁ , TG ₂ ... Speed detector,
PG ₁ , PG ₂ ... Position detector, 12 ... Load, 13
...Determination device, 14, 15...Switching device, 16...
...Function generator, 17...Gap width setting device, 1
8... Command device, 19... Target position setting device, 2
0...Drive device.

Claims (1)

[Scope of Claims] 1. In a mobile machine in which a gap due to backlash exists between a movable body A on the drive side and a movable body B on the load side, the positions and speeds of the movable bodies A, B and the gap are respectively expressed as (X _M , X〓 _M ), (X _L ,
When X〓 _L ), (e, e〓) and the width of the backlash is L, a determine whether to accelerate or decelerate the load (X _L ,
X〓 _L ) Judging from the phase plane, in case of acceleration, acceleration side contact (e, e〓) = (L/2, 0), in case of deceleration, deceleration side contact (-L/2, 0) command. b Determine whether the state of the gap (e, e〓) is in the specified contact state c Gap control that determines the amount of operation so that the state of the gap (e, e〓) becomes the commanded state d Gap control Load position control that determines the amount of operation so that the load state (X _L , X〓 _L ) becomes the commanded state when the state of Equipped with a function to switch between control and positioning in the positive (negative) direction without collision (e, e〓) = (L/2,
0) (Negative direction contact (-L/2, 0)) Accelerate the load in the positive (negative) direction Constant speed (optional) Command deceleration based on acceleration/deceleration judgment Negative direction contact (-L/2) without collision , 0), (positive direction) contact (L/2, 0)) Decelerates the load in the negative (positive) direction, and achieves high accuracy and A method for controlling a mobile machine having a backlash that enables highly responsive position control. 2. A control system for a mobile machine according to claim 1, characterized in that a movement control device with nonlinear control is used, and the gap control is performed by nonlinear control.