 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an apparatus for controlling electric
commonly used motors, such as DC motors, synchronous motors, induction
motors and other motors.
2. Description of the Prior Art
FIG. 1 is a block diagram of a control system used in a three-phase
synchronous motor and constructed in accordance with the prior art.
The control system comprises a position sensor 8 for sensing the rotational
position P of a three-phase synchronous motor 7. The position sensor 8 is
connected to a speed sensor 5 for sensing a rotational speed VL based on
the rotational position P. The speed sensor 5 is connected to a subtracter
1 for subtracting the rotational speed VL from a speed command VLC to
determine a speed error DVL. The subtracter 1 is also connected to a PID
compensator 2 for compensating the speed error DVL to output a torque
command value T. The PID compensator 2 is connected to a current command
value setting means 3 for setting motor current commands CU, CV and CW
based on the rotational motor position P and torque command value T. The
current command value setting means 3 is also connected to a memory 6 for
storing torque ripple error information in the electric motor and a power
amplifier 4 for supplying three-phase currents IU, IV and IW corresponding
to the motor current commands CU, CV and CW to the electric motor 7.
As shown in FIG. 2, the current command value setting means 3 comprises an
address setting means 9 for preparing a read address of AD in the memory 6
based on the rotational motor position P, an adder 12 for adding torque
error information MD read out of the memory 6 to the torque command value
T to determine a current amplitude AM in the electric motor and a
three-phase current setting means 10 for setting motor current commands
CU, CV and CW from the current amplitude AM and rotational motor position
P.
On operation, the position sensor 8 senses the rotational position P of the
three-phase synchronous motor 7 while the speed sensor 5 senses the
rotational speed VL from the rotational position P. On the other hand, the
subtracter 1 receives the speed command VLC and subtracts the rotational
speed VL from the speed command VLC to determine the speed error DVL. The
speed error DVL is compensated by the PID compensator 2 which in turn
outputs the torque command value T toward the current command value
setting means 3. The current command value setting means 3 then sets the
motor current commands CU, CV and CW based on the rotational motor
position P and torque command value T. The motor current commands thus set
are provided to the power amplifier 4 which in turn supplies the three
phase motor currents IU, IV, IW to the electric motor 7.
The operation of the current command value setting means 3, that is, a
process of preparing three phase current command values with a technique
of compensating the torque ripple in the electric motor will be described
below.
The rotational motor position P is used to prepare the read address AD of
the memory 6 at which the torque error information in the electric motor
is stored. The read address AD is then used to read the torque error
information MD corresponding to the rotational motor position P. The adder
12 adds the torque error information MD to the torque command value T to
calculate the current amplitude AM in the electric motor. The three phase
current setting means 10 sets the motor current commands CU, CV and CW
from the current amplitude AM and rotational motor position P according to
the following equations:
CU=AMx sin P (1)
CV=AMx sin (P+120) (2)
CW=AMx sin (P+240) (3)
AM=T+MD (4).
The torque error information MD has been stored in the memory 6 based on
the relationship between the rotational positions P (P1, P2 . . . ) and
the corresponding torque errors (DT1, DT2 . . . ), as shown in FIG. 3. As
a result, the slot torque ripple and other factors at each of the
rotational motor positions can be compensated to realize a control by
which the torque ripple being an error component in the motor output
torque is relieved.
In general, the stator and rotator of the electric motor produce an error
torque known as torque ripple since the shapes of the winding slots,
magnetic poles and others are discontinuous in the rotational direction.
The magnitude of such an error torque can vary between about 0.5% of the
rated torque and equal to or higher than 10% from case to cases, depending
on the type of motor. Particularly, a reluctance motor tends to increase
the torque ripple since the magnetic resistance in the rotator becomes
larger depending on the rotational position. Where it is required to
control the electric motor with an increased accuracy, the torque ripple
raises an important problem if the vibrations and noises in the electric
motor may adversely affect the circumferences.
The motor control system of the prior art had a problem in that the torque
and speed could not be accurately controlled, resulting from the
dependency of the motor torque ripple on the rotational position as well
as the torque; the non-linear change of the torque ripple depending on the
magnitude of the torque; and the relation of cause and effect with the
rotational speed from the delay in the control of the entire system.
Particularly, where an electric motor having an increased reaction on the
armature is used, the torque ripple form may not be in proportion to the
output torque of the electric motor. A synchronous motor using an
electromagnet has an increased freedom of an electromagnet field. However,
even when such a type of synchronous motor is driven with an unsaturated
electromagnet field density, the torque ripple form may not be in
proportion to the output of the electric motor.
The structure of the electric motor itself has been modified to relieve the
torque ripple therein. For example, the gap between the stator and the
rotator has been increased; the structure of the rotator has been uniquely
modified; and the stator or rotator has been arranged to be skewed
relative to the rotational axis. However, such devises will reduce the
efficiency in the electric motor. Thus, the motor must correspondingly be
increased in size and complicated in structure, leading to increase of the
manufacturing cost. Furthermore, the torque error in the electric motor
depends on various parameters. If all the parameters such as torque
compensation data, motor current values and others are to be stored in the
memory means, the memory capacity will be huge, also leading to increase
of the memory cost.
SUMMARY OF THE INVENTION
In order to overcome the above problems, an object of the present invention
is to provide a control system which can compensate the torque ripple in
an electric motor to control the rotation thereof with an improved
accuracy.
To this end, the present invention provides an electric motor control
system for compensating the rotation error in an electric motor operated
by a given torque command, the control system comprising control
information storage means for storing control information corresponding to
the torques, rotational speeds and rotational positions of the electric
motor, a position sensor for sensing the rotational position of the
electric motor, a speed sensor for sensing the rotational speed of the
electric motor, control information reading means for reading the control
information corresponding to the torque command, rotational speed and
rotational position of the running motor, motor current setting means for
preparing a motor current command based on the read control information,
and a power amplifier for supplying current and voltage corresponding to
the motor current command to the electric motor.
Preferably, the control information stored in the control information
storage means includes data used to compensate the torque error
corresponding to the torque, rotational speed and rotational position of
the electric motor. The motor current setting means is adapted to prepare
the motor current command by adding the torque error compensation data to
the given torque command.
Preferably the control information stored in the control information
storage means also includes motor current information corresponding to the
torque, rotational speed and rotational position of the electric motor.
The motor current setting means is adapted to prepare the motor current
command from the read motor current information.
Preferably the control information stored in the control information
storage means further includes information represented only by a term or
terms of series developed current information in which the amplitude is
equal to or larger than a predetermined level. The motor current setting
means is adapted to determine a series from each term in the series
developed current information read out to prepare a motor current command.
The present invention also provides a method of compensating the rotation
error of an electric motor operated by a given torque command, comprising
the steps of previously storing control information corresponding to the
torques, rotational speeds and rotational positions of the electric motor,
sensing the rotational position of the electric motor, sensing the
rotational speed of the electric motor, reading the control information
corresponding to the torque command, rotational speed and rotational
position of the running motor, preparing a motor current command based on
the read control information, and supplying current and voltage
corresponding to the motor current command to the electric motor.
Preferably, the control information storing step includes a step of storing
data used to compensate the torque error corresponding to the torque,
rotational speed and rotational position of the electric motor. The motor
current command preparing step includes a step of preparing the motor
current command by adding the torque error compensation data to the given
torque command.
Preferably, the control information storing step also includes a step of
storing motor current information corresponding to the torque, rotational
speed and rotational position of the electric motor. The motor current
command preparing step includes a step of preparing the motor current
command from the read motor current information.
Preferably, the control information storing step further includes a step of
series developing the motor current information and storing information
represented only by a term or terms of series developed current
information in which the amplitude is equal to or larger than a
predetermined level. The motor current command preparing step includes a
step of determining a series from each term in the series developed
current information read out to prepare a motor current command.
In such a manner, the current and voltage applied to the electric motor can
be compensated depending on changes in the rotational speed, rotational
position and torque of the electric motor which may cause changes in the
torque error of the electric motor. Thus, the torque error in the electric
motor can be reduced without increase of the size and manufacturing cost
of the electric motor.
BRIEF DESCRIPTION OF THE ACCOMPANY DRAWINGS
FIG. 1 is a block diagram of a control system for a three-phase synchronous
motor according to the prior art.
FIG. 2 is a block diagram of a current command value setting means
according to the prior art.
FIG. 3 is a table illustrating the compensation of torque errors stored in
a memory according to the prior art.
FIG. 4 is a block diagram of a control system for a three-phase synchronous
motor according to the present invention.
FIG. 5 is a block diagram of a current command value setting means
according to one embodiment of the present invention.
FIG. 6 is a table illustrating the compensation of torque errors stored in
a memory according to the present invention.
FIG. 7 is a block diagram of a current command value setting means
according to another embodiment of the present invention.
FIG. 8 is a block diagram of a current command value setting means
according to still another embodiment of the present invention.
FIG. 9 is a table of series information data relating to the present
invention.
FIG. 10 is another table of series information data relating to the present
invention.
FIG. 11 is a table of series information when the erroneous torques depend
on both the output torques and rotational speeds in the embodiment of the
present invention.
FIG. 12 is a table illustrating the interpolation of torque error
compensation values determined from the torque error compensation data.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
The preferred embodiments of the invention will now be described with
reference to the drawings.
Embodiment 1
FIG. 4 is a block diagram of a control system for a three-phase synchronous
motor, constructed in accordance with the present invention.
The control system comprises a position sensor 8 for sensing the rotational
position P of a three-phase synchronous motor 7. The position sensor 8 is
connected to a speed sensor 5 for sensing a rotational speed VL based on
the rotational position P and a current command value setting means 13.
The speed sensor 5 is connected to a subtracter 1 for subtracting the
rotational speed VL from a speed command VLC to determine a speed error
DVL and the current command value setting means 13. The subtracter 1 is
connected to a PID compensator 2 for compensating the speed error DVL to
output a torque command value T. The PID compensator 2 is connected to the
current command value setting means 13 for setting motor current commands
CU, CV and CW based on the rotational motor position P, rotational speed
VL and torque command value T of the electric motor as well as torque
error information MD.sub.1 read out from a memory 61. The current command
value setting means 13 is connected to the memory 61 being compensation
storage means for storing torque error compensation data corresponding to
the torque, rotational speed and rotational position of the electric motor
and further to a power amplifier 4 for supplying three-phase currents IU,
IV and IW corresponding to the motor current commands CU, CV and CW to the
electric motor 7.
FIG. 5 is a block diagram showing the layout of the current command value
setting means 13.
As shown in FIG. 5, the current command value setting means 13 comprises an
address setting means 91 for preparing a read address AD.sub.1 in the
memory 61 based on the torque command T, speed information VL and
rotational motor position P, an adder 12 operative as a motor current
preparing means for preparing a motor current amplitude AM based on the
torque command T and also a torque error compensation data MD.sub.1 read
out from the memory 61, and a three-phase current setting means 10
operative as a first motor current setting means for preparing motor
current commands CU, CV and CW based on the prepared motor current
amplitude AM and the rotational motor position P.
The error compensation data MD.sub.1 stored in the memory 61 depends on
three parameters, torque, speed VL and rotational position P and may be
stored in the memory 61 as shown in FIG. 6.
On operation, the position sensor 8 senses the rotational position P of the
three-phase synchronous motor 7 while the speed sensor 5 senses the
rotational speed VL by differentiating the rotational position P thus
sensed. On the other hand, the subtracter 1 receives the speed command VLC
and subtracts the rotational speed VL from the speed command VLC to
determine the speed error DVL. The speed error DVL is compensated by the
PID compensator 2 which in turn outputs the torque command value T toward
the current command value setting means 13. The current command value
setting means 13 then sets the motor current commands CU, CV and CW based
on the rotational motor position P, rotational speed VL and torque command
value T. The motor current commands thus set are provided to the power
amplifier 4 which in turn supplies the three phase motor currents IU, IV,
IW to the electric motor 7. In such a manner, the rotation of the electric
motor can more accurately be controlled by compensating the torque error.
The operation of the current command value setting means 13 will be
described below. FIG. 6 is a table illustrating the compensation of torque
errors stored in the memory.
The memory 61 has previously stored, as torque error data, compensated
current amplitude values corresponding to the torques, speeds and
rotational positions of the electric motor. The address setting means 91
prepares a read address AD.sub.1 in the memory 61 based on the torque
command T, speed information VL and rotational motor position P. The
prepared read address AD.sub.1 is used to read a torque error compensation
data MD.sub.1 from the memory 61. For example, if the rotational position
is P3 and the rotational speed is VL2 as shown in FIG. 6, a torque error
compensation data DT32 is selected and read out from the memory 61. The
torque error compensation data DT32 contains a plurality of data
corresponding to the magnitude of the respective torques. When the torque
error compensation data is actually used, it will contain values more or
less different from the actual values relating to the torque, speed and
position of the electric motor. Therefore, the torque error compensation
data containing values approximate to the actual values is selected.
Alternatively, the torque error compensation data is interpolated and
calculated based on a plurality of data approximate to the actual data.
The adder 12 adds the torque error compensation data MD.sub.1 thus prepared
to the torque command T to calculate a current amplitude AM in the
electric motor. The three-phase current setting means 10 then sets
three-phase AC current commands CU, CV and CW from the current amplitude
AM and the rotational motor position P according to the following
equations:
CU=AMx sin P (1)
CV=AMx sin (P+120) (2)
CW=AMx sin (P+240) (3)
AM=T+MD (4).
Electric power is supplied to the electric motor through a three-phase AC
inverter or the like according to the three-phase AC current commands CU,
CV and CW. Thus, the torque error in the electric motor can be compensated
to realize a control for controlling the rotation of the electric motor
with an improved accuracy.
Embodiment 2
FIG. 7 is a block diagram showing another current command value setting
means 14 usable in the present invention.
As shown in FIG. 7, the cur:cent command value setting means 14 comprises
an address setting means 92 for preparing a read address AD.sub.2 in a
memory 62 from the torque command T, speed information VL and rotational
motor position P and a three-phase current setting means 10 operative as a
motor current setting means for preparing motor current commands CU, CV
and CW based on information read out from the memory 62 (motor current
amplitude value AM.sub.1). Therefore, the adder 12 used in the embodiment
1 is not required.
On operation, the memory 62 has previously stored motor current amplitude
values AM corresponding to the torques, speeds and rotational positions of
the electric motor. The address setting means 92 prepares a read address
AD.sub.2 in the memory 62 based on the torque command T, speed information
VL and rotational motor position P. The prepared read address AD.sub.2 is
used to read a motor current amplitude value AM from the memory 62. The
motor current amplitude value AM contains a plurality of data
corresponding to the magnitude of the respective torques. When the motor
current amplitude value is actually used, it will contain values more or
less different from the actual values relating to the torque, speed and
position of the electric motor. Therefore, the torque error compensation
data containing values approximate to the actual values is selected.
Alternatively, the torque error compensation data is interpolated and
calculated from a plurality of data approximate to the actual data. The
three-phase current setting means 10 sets three-phase AC current commands
CU, CV and CW from the motor current amplitude AM and the rotational motor
position P in the same manner as described above. The three-phase AC
current commands CU, CV and CW may be stored in the memory 62 and read out
therefrom according to the read address AD.
Embodiment 3
FIG. 8 is a block diagram showing still another current command value
setting means 15 usable in the present invention.
A memory 63 functions as a series information storage means for storing
motor current information corresponding to the torque commands T, speed
information VL and rotational motor positions P as series information
decomposed into a Fourier series. The current command value setting means
15 comprises an address setting means 93 for preparing a read address AD
in the memory 63 based on the torque command T or a combination of the
torque command T with the speed information VL, a motor current setting
means 11 operative as a series reproducing means for reproducing the
series based on the series information read out from the memory 63 as well
as the rotational motor position P and a three-phase motor current setting
means for preparing current commands CU, CV and CW in the respective
phases of the motor from the output AM of the motor current setting means
11.
The third embodiment is characterized by that the enormous data containing
motor torque error data or compensated motor current values is compressed
to greatly reduce the memory capacity for realizing a reduced
manufacturing cost in the system. In the compensation of torque error
according to the prior art, the amount of data to be compensated will be
enormous. Where the motor erroneous torque does not depend on the
rotational speed, for example, the amount of data to be compensated is
represented by:
[1000 point data per one revolution].times.[500 point data depending on the
magnitude of torque].times.[2 byte data per one point]=1,000,000 bytes of
data per one revolution.
Therefore, the present invention decomposes proper motor current
information used to compensate the motor torque error into a series such
as a Fourier series, depending on the torque or a combination of torque
with speed. Among them, only the representative series components highly
affecting the system are stored in the memory 63 in a number sufficient
for the required accuracy. The series information may contain amplitude or
phase values in the respective orders of the series.
An example of series information where the erroneous torque of the electric
motor depends on the magnitude of output torque, but not the rotational
speed, is shown in FIGS. 9 and 10. FIG. 9 shows the relationship between
the series information FT0, FT1, FT2 . . . FTN which correspond to torque
commands T, respectively. The series information FTN is a group of data
which are obtained by series developing current amplitudes per one
revolution of an electric motor having its output torque being TN and
selecting series terms having relatively large amplitudes by the number
required by the desired accuracy. As shown in FIG. 11, each of the series
terms comprises a combination of amplitude FA1, FA2, FA3 . . . or FAN With
phase FP1, FP2, FP3 . . . or FPN. These series information have been
stored in the memory 63. When this is to be represent by an equation, the
ideal amplitude in the motor current corresponding to a torque command TX
can be represented by AX (P) being a function of rotational motor position
P. The function AX (P) is generally represented by a Fourier series as
shown by an equation 5 and can be modified as shown by an equation 6:
AX(P)=A.sub.0 /2+(A.sub.1 x cos P)+(A.sub.2 x cos 2P)+(A.sub.3 x cos 3P)+ .
. . +(A.sub.N x cos NP)+(B.sub.1 x sin P)+(B.sub.2 x sin 2P)+(B.sub.3 x
sin 3P)+ . . . +(B.sub.N x sin NP) (5)
AX(P)=FA0+FA1x sin (P+FP1)+FA2x sin (2P+FP2)+FA3x sin (3P+FP3)+ . . . +FANx
sin (NP+FPN) (6)
FIGS. 9 and 10 show the example of series data table obtained according to
the equation 6, but the present invention is not limited to such a case.
The present invention may be applied to any other form of series or
modifications.
As described, the prior art had to handle data amounting to 1,000,000
bytes. However, the present invention only requires the storing of
information of orders having higher amplitudes. If it is now assumed that
series information belonging to ten sets of orders is to be stored, the
amount of data to be compensated will be equal to
[amplitudes and phases in ten sets of orders].times.[500 point data
depending on the magnitude of torque]=data of 20,000 bytes/one revolution
of the electric motor.
This is equal to 1/50 of the amount of data to be handled in the prior art.
The reduction of data according to the present invention can qualitatively
be explained by the fact that since an electrical motor is inherently
constructed into a geometrically accurate form, the great portion of
causes resulting in torque ripple errors resides in the shapes of slots,
magnetic poles and others to provide periodical errors, rather than fully
random errors.
FIG. 11 shows series information in a case where the erroneous torque of an
electrical motor depends on both the output torque and rotational speed of
the same. The series information two-dimensionally exist relating to
torque commands T1, T2, T3 . . . and revolutions VEL1, VEL2, VEL3 . . .
However, the concrete contents of the respective series information are of
the same type as in FIG. 10 and show amplitudes and phases in the
respective orders.
A process of preparing current command values AM in an electric motor will
be described below.
Address setting means 93 prepares a read address AD.sub.3 based on a torque
command T and motor speed information VL. The prepared read address
AD.sub.3 is used to read series information MD.sub.3 from the memory 63.
The motor current setting means 11 then substitutes the series
information, motor speed information VL and rotational motor position P
for the equation (6) to realize a Fourier series which is in turn used to
determine motor current amplitudes AM. As can be seen from FIGS. 9 to 11,
the series information stored are almost all the finite number of
numerical approximations relating to torque commands T and revolutions VEL
for predetermined time intervals, rather than accurate series information
relating to torque commands T and revolutions VEL in the actually Sunning
motor. This can be overcome the following manner.
As shown in FIG. 12, it is now assumed that the torque command T and
revolution VEL being now controlled are on plane coordinates FT (X, Y)
while the series information stored in the memory are ones relating to FT
(N, M), FT (N+1, M), FT (N, M+1) and FT (N+1, M+1). The simplest process
utilizes the series information of the nearest FT (N, M+1) as a
representative series information of FT (X, Y). Another process determines
the series information corresponding to FT (X, Y) by interpolating four
approximate sets of series information. More particularly, the series
information of coordinates FT (X, M) is equally distributed from the
series information of FT (N, M) and FT (N+1, M), followed by a linear
interpolation. Similarly, the series information of coordinates FT (N, Y)
is equally distributed from the series information of FT (N, M+1) and FT
(N+1, M+1), followed by a linear interpolation. The series information of
FT (X, Y) to be determined is then linearly interpolated from the series
information of FT (X, M) and FT (N, Y).
The motor current amplitudes AM and rotational motor position P thus
determined are used to set three-phase AC current commands CU, CV and CW
in the same manner as described. According to the three-phase AC current
commands CU, CV and CW, electric power is supplied from the power
amplifier 4 such as a three-phase AC inverter, to the electric motor 7. In
such a manner, the torque errors in the electric motor can be compensated
to realize an accurate control of motor rotation.
Embodiment 4
The embodiment 3 can be modified into an embodiment 4 which will be
described below.
In the embodiment 3, the current amplitudes AM in the three-phase AC
current commands CU, CV and CW are assumed to be a function AX (P) of the
rotational motor positions P. These current amplitudes AM are developed
into a Fourier series representing series information which will be stored
in the memory. When the electric motor is to be controlled, the series
information are inversely calculated to prepare the three-phase AC current
commands CU, CV and CW. However, the embodiment 4 does not directly handle
the concept of current amplitude. More particularly, it is considered that
the three-phase AC current commands CU, CV and CW depend on torque T,
rotational speed VL and rotational position p. The three-phase AC current
commands CU, CV and CW are respectively represented by a Fourier series as
a function of rotational position P and then stored in the memory. On
reproduction, the three-phase AC current commands CU, CV and CW are
determined by reproducing the respective series in the same manner as in
the embodiment 3. Thus, the three-phase currents are respectively treated
directly as a Fourier series without the use of the concept that the
current amplitudes are DC values converted directly from the three-phase
currents. This will increase the memory capacity and calculation, but may
be aided in the following manner.
Since two commands CV and CW in the three-phase AC current commands are
CV(P)=CU(P+120)and
CW(P)=CU(P+240),
the series information relating to the values CV (P) and CW (P) are
substituted by the series information of CU (P). Thus, the necessary
memory capacity can be reduced. Even in the embodiment 3, therefore,
several modified techniques can be used.
The present invention may be applied to a control system for a three-phase
AC motor performing two-to-three phase conversion and further to any other
control system for multi-phase and multi-pole motors, DC motors and
induction motors.
Although the description and claims are expressed by the dependency of
torque error compensation data and current information on three
parameters, that is, torque, speed and rotational position, the present
invention may naturally be applied to such a case that the control depends
less on one of these three parameters, e.g., speed. In such a case, the
torque error compensation data and current information will be controlled
as values which only depend on the remaining parameters, that is, torque
and rotational position. The present invention also covers such a control
system.
Data may be partially stored to reduce the memory capacity for motor torque
error information and others. To realize a more strict control, it is of
course possible that a delayed time in the detection of position, a
delayed time on control and an average delayed time until the current
substantially acts on the electric motor may be presumed without direct
use of the rotational position information. The rotational position may be
controlled by additionally using changes in rotational position
corresponding to these delayed times.
* * * * *
|
|
 |
|
 |
Description |
|
