Electricity: motive power systems – Positional servo systems
Reexamination Certificate
2003-02-04
2004-11-30
Duda, Rina (Department: 2837)
Electricity: motive power systems
Positional servo systems
C318S565000, C318S567000, C318S568170, C318S626000, C700S028000, C700S031000, C700S041000, C700S044000, C700S045000, C700S046000
Reexamination Certificate
active
06825631
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a controlling device of a machine tool, and a robot, etc.
BACKGROUND OF THE INVENTION
In the prior arts, there is a device disclosed in the xi-th embodiment of International Unexamined Patent Publication No. W093/20489, which is proposed by the present applicant, as this type of controlling device.
FIG. 6
is a block diagram showing a configuration of a device according to the xi-th embodiment of the International Unexamined Patent Publication No. W093/20489.
In
FIG. 6
, i indicates the present time, i+X (X is a positive integer) indicates a future time coming in X sampling from the present sampling time i at a sampling time Ts, and i−X indicates a past time elapsed by X sampling from the time i at a sampling time Ts. Also, r(Y) (Y is an integer) indicates a target command of time Y, y(Y) indicates an output from a controlled object (not illustrated) at time Y, and u(Y) indicates an input (hereinafter called “control input”) to a controlled object (not illustrated) at time Y. A prior art controlling device
80
shown in
FIG. 6
is a controlling device for reconciling the output of a controlled object (not illustrated) with a target command given.
Referring to
FIG. 6
, the controlling device
80
inputs a control input u(i) into a controlled object (not illustrated), so that an output y(i) from the object is reconciled with a target command r(i), using a future target command r(i+M) given and an output y(i−K) from the object as inputs. The controlling device
80
includes memories
81
,
82
,
83
,
84
, an operational unit
85
, and a subtracter
86
.
The memory
81
stores a target command from K sampling past to M sampling future, memory
82
stores controlling constants, memory
83
stores outputs from a controlled object from K+Na (Na is a natural number) sampling past to K sampling past, and memory
84
stores control inputs from K+Nb (Nb is a natural number) past to 1 sampling past. The subtracter
86
acquires an error between the target command r(i−K) and an output y(i−K) from the controlled object.
The operational unit
85
is a computing element for determining a control input u(i) as shown below, so that an performance function regarding both a future error prediction value obtained by using a transfer function model from a control input u(i) to an output y(i) of a controlled object and the control input become minimal:
u
(
i
)
=
∑
m
=
1
M
q
m
r
(
i
+
m
)
-
∑
n
=
0
Na
p
n
y
(
i
-
K
-
n
)
+
Ee
(
i
-
K
)
-
∑
n
=
1
Nb
+
K
g
n
u
(
i
-
n
)
According to the prior art controlling device
80
, since the control input is determined so that the future error prediction value is minimized, control of high follow-up accuracy can be applied to the controlled object.
However, in the prior art controlling device
80
shown in
FIG. 6
, if any feedforward control is applied to the controlled object, the controlling device
80
predicts a future error with the feedforward signal (hereinafter called “FF signal”) not taken into consideration, wherein a prediction error is produced in the future error prediction value, and resultantly there is a problem of worsening the follow-up accuracy.
DISCLOSURE OF THE INVENTION
It is therefore an object of the invention to provide a controlling device of high follow-up accuracy, the prediction accuracy of which does not deteriorate by an FF signal where feedforward control is applied to a controlled object.
In order to solve the above-described problem, a prediction controlling device according to the invention outputs a control input and feedforward signal to the controlled object, so that the output of the controlled object is reconciled with a target command, and the prediction controlling device has a feedforward signal generation command filter that receives a target command signal which is information of the future target command as an input and outputs a future command increment which is an increment from one sampling period to the next sampling period of a target command signal from the present sampling time to a multiple-sampling future and a feedforward signal from the future target command signal, and the prediction controlling device further has a prediction controller that receives the future command increment, the feedforward signal and a controlled object output at the past sampling time over zero sampling as inputs, acquires the future error prediction value by using a transfer function model from the feedforward signal and the control input to the controlled object output, determines the control input so that the performance function of the error prediction value and the control input becomes minimal, and applies the control input to the controlled object.
In operations in the prediction controller, since a future error prediction value is acquired by a transfer function model with the feedforward signal taken into consideration, and the control input is determined so that the performance function of the future prediction value and control input is minimized, the prediction accuracy is not degraded by adding the feedforward control thereto.
According to an embodiment of the invention, the above-described feedforward signal generation command filter receives the above-described target command signal at the present sampling time as an input and outputs the corresponding target command signal or increments between the respective sampling periods of signals obtained by sampling the corresponding target command signal as the above-described command increments.
According to the present embodiment of the invention, the above-described feedforward signal generation command filter calculates and outputs the above-described feedforward signals, which are:
FF
1
(
i
)=Gain
1
·
&Dgr;r
(
i+m
1
)
FF
2
(
i
)=Gain
2
·
{&Dgr;r
(
i+m
2
)−
&Dgr;r
(
i+m
2
−1)}
where i is the present sampling time, Gain
1
and Gain
2
are constants, m
1
and m
2
are integers that meet 0≦m
1
≦m
2
, &Dgr;r(i+m
1
) is the above-described command increment of the m
1
sampling future, and FF
1
(i) and FF
2
(i) are the feedforward signals.
According to the embodiment of the invention, the above-described controlled objects are a motor and its speed controller, the control input is a speed command, the controlled object output is a motor position, and the feedforward signals are a feedforward signal for speed control, and a feedforward signal for torque control.
According to the present embodiment of the invention, the above-described prediction controller includes: an integrator that receives the future command increment as an input and calculates the target command from the present sampling time to a multiple-sampling future; a memory section that stores constants for prediction control in advance and receives as inputs the target command calculated by the integrator, two feedforward signals, controlled object output, and control input, and stores the past target command, past feedforward signals, past controlled object output and past control input; a subtracter that subtracts the controlled object output from the past target command and acquires the past error; and an operational unit that receives as inputs the target command acquired by the above-described integrator, the past feedforward signals, the past controlled object output, the past control input and constants for prediction control, which are stored by the memory section, and the error obtained by the subtracter, obtains a future error prediction value by using a discrete-time transfer function model, which is:
Y
(
z
)={(
b
1
z
−1
+ . . . +b
Nb
z
−Nb
)
U
(
z
)+(
d
1
z
−1
+ . . . +d
Nd
z
−Nd
)
FF
1
(
z
)+(
c
1
z
−1
+ . . . +c
Nc
z
−Nc
)
FF
2
(
z
)}/(1
−a
1
z
−1
31
. . . −a
Na
z
−Na
),
from the feedforward signal and the control input to the controlled o
Duda Rina
Kabushiki Kaisha Yaskawa Denki
Knobbe Martens Olson & Bear LLP
LandOfFree
Prediction controlling device does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Prediction controlling device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Prediction controlling device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3284325