U.S. patent application number 11/636593 was filed with the patent office on 2007-06-14 for position detecting device and synchronous motor driving device using the same.
This patent application is currently assigned to Hitachi, LTD.. Invention is credited to Toshiyuki Ajima, Bunji Furuyama, Hideki Miyazaki, Masataka Sasaki, Tokihito Suwa.
Application Number | 20070132423 11/636593 |
Document ID | / |
Family ID | 37745871 |
Filed Date | 2007-06-14 |
United States Patent
Application |
20070132423 |
Kind Code |
A1 |
Ajima; Toshiyuki ; et
al. |
June 14, 2007 |
Position detecting device and synchronous motor driving device
using the same
Abstract
A position detecting device which can increase accuracy in
detecting the pole position of a motor used to perform quick
acceleration and deceleration over the range from a zero speed to a
high rotation speed, and a synchronous motor driving device using
the position detecting device. A position detector detects
basic-wave component signals in sensor signals and executes
position calculation. A correcting unit calculates signal
information representing at least one of a gain, an offset and a
phase by a phase detector from the basic-wave component signals
detected by an error calculator, and makes correction based on the
calculated signal information such that a position detection error
is zero.
Inventors: |
Ajima; Toshiyuki; (Tokai,
JP) ; Miyazaki; Hideki; (Hitachi, JP) ;
Sasaki; Masataka; (Hitachi, JP) ; Furuyama;
Bunji; (Hitachi, JP) ; Suwa; Tokihito;
(Hitachinaka, JP) |
Correspondence
Address: |
CROWELL & MORING LLP;INTELLECTUAL PROPERTY GROUP
P.O. BOX 14300
WASHINGTON
DC
20044-4300
US
|
Assignee: |
Hitachi, LTD.
Tokyo
JP
|
Family ID: |
37745871 |
Appl. No.: |
11/636593 |
Filed: |
December 11, 2006 |
Current U.S.
Class: |
318/719 |
Current CPC
Class: |
H02P 2209/07 20130101;
H02P 6/10 20130101; H02P 6/16 20130101; G01D 5/2046 20130101 |
Class at
Publication: |
318/719 |
International
Class: |
H02P 1/46 20060101
H02P001/46; H02P 27/00 20060101 H02P027/00 |
Foreign Application Data
Date |
Code |
Application Number |
Dec 12, 2005 |
JP |
2005-357695 |
Claims
1. A position detecting device for detecting a motor pole position
by using two or more sensor signals, said position detecting device
comprising: a position detector for detecting basic-wave component
signals in said sensor signals and executing position calculation;
and correcting means for calculating signal information
representing at least one of a gain, an offset and a phase from the
detected basic-wave component signals, and making correction based
on the calculated signal information such that a position detection
error is zero.
2. The position detecting device according to claim 1, wherein said
position detector extracts said sensor signals in a range between
predetermined first and second determination values, thereby
detecting said basic-wave component signals.
3. The position detecting device according to claim 1, wherein said
position detector extracts sensor signals representing middle
potentials of said sensor signals, thereby detecting said
basic-wave component signals.
4. The position detecting device according to claim 1, wherein said
sensor signal has a distorted waveform including a predetermined
zone in which a sensor signal value is hardly changed.
5. The position detecting device according to claim 1, wherein said
sensor signals are three-phase signals of U-, V- and W-phases, and
said position detector extracts a signal from a cross point between
the U-phase signal and the W-phase signal to a cross point between
the V-phase signal and the W-phase, a signal from a cross point
between the V-phase signal and the W-phase signal to a cross point
between the U-phase signal and the V-phase, and a signal from a
cross point between the U-phase signal and the V-phase signal to a
cross point between the U-phase signal and the W-phase, thereby
detecting said basic-wave component signals.
6. The position detecting device according to claim 1, wherein said
correcting means comprises: an error calculator for calculating a
position error amount from a position calculated value obtained by
said position detector; and a position corrector for correcting the
position calculated value by using the position error amount
calculated by said error calculator.
7. The position detecting device according to claim 1, wherein said
correcting means includes a signal corrector for determining an
error amount of said sensor signal from a difference between said
signal information and a predetermined reference value, and for
correcting the sensor signal such that the determined error amount
is zero.
8. The position detecting device according to claim 1, further
comprising a sensor unit for generating said sensor signal, wherein
said sensor unit includes a magnetic adjusting member for adjusting
an amount of effective magnetic flux crossing a sensor.
9. The position detecting device according to claim 1, further
comprising a sensor unit for generating said sensor signal, wherein
said sensor unit includes a magnetic shielding member oriented
perpendicularly to a direction in which a sensor senses magnetic
flux.
10. The position detecting device according to claim 1, further
comprising a sensor unit for generating said sensor signal, wherein
said sensor unit includes a plurality of sensors attachable to a
motor housing from the outside.
11. A position detecting device for detecting a motor pole position
by using two or more sensor signals, said position detecting device
comprising: a position detector for detecting basic-wave component
signals in said sensor signals and executing position calculation;
and correcting means for calculating signal information
representing at least one of a gain, an offset and a phase from the
detected basic-wave component signals, and making correction based
on the calculated signal information such that a position detection
error is zero, wherein said position detector extracts said sensor
signals in a range between predetermined first and second
determination values, thereby detecting said basic-wave component
signals.
12. A permanent-magnet synchronous motor driving device comprising:
a position detecting device for detecting a motor pole position by
using two or more sensor signals; and a motor controller for
producing a PWM drive signal based on a position detected value
detected by said position detecting device, and outputting the PWM
drive signal to an inverter, thereby driving a permanent-magnet
synchronous motor, wherein said position detecting device
comprises: a position detector for detecting basic-wave component
signals in said sensor signals and executing position calculation;
and correcting means for calculating signal information
representing at least one of a gain, an offset and a phase from the
detected basic-wave component signals, and making correction based
on the calculated signal information such that a position detection
error is zero.
13. The permanent-magnet synchronous motor driving device according
to claim 12, wherein said motor controller includes a rotor
position settling unit for driving said permanent-magnet
synchronous motor through a predetermined electrical angle in a
stepping manner under PWM control and settling a rotor stop
position of said motor when the signal information is calculated by
said correcting means, and said correcting means of said position
detecting device calculates, at a rotor position settled by said
rotor position settling unit, a rotor position detection error from
a difference between a value of the settled rotor position and the
position detected value detected by said position detecting device,
said rotor position detection error being corrected.
14. The permanent-magnet synchronous motor driving device according
to claim 12, wherein said motor controller receives a torque
command, and said permanent-magnet synchronous motor has a motive
power load over a wide speed range from a stopped state to a state
of large acceleration and deceleration.
15. The permanent-magnet synchronous motor driving device according
to claim 12, wherein said motor controller receives a torque
command from a vehicular control unit, and said permanent-magnet
synchronous motor has a vehicular motive power load over a wide
speed range from a stopped state to a high-speed travel state.
Description
BACKGROUND OF THE INVENTION
[0001] 1. Field of the Invention
[0002] The present invention relates to a position detecting device
for detecting a motor pole position and a synchronous motor driving
device using the position detecting device. More particularly, the
present invention relates to a position detecting device and a
synchronous motor driving device using the same, which are suitably
used to perform acceleration and deceleration over the range from a
zero speed (stopped state) to a high rotation speed.
[0003] 2. Description of the Related Art
[0004] As one of devices for driving a synchronous motor
(hereinafter referred to simply as a "motor"), there is known a
type of driving the motor by using a position detecting device
which detects a rotor pole position. Here, the term "position
detecting device" means a device, such as a resolver or an absolute
rotary encoder, which can not only detect a motor rotational angle
during rotation, but also determine an absolute position in the
stopped state. In order to achieve driving of the motor with high
efficiency, however, it is essential to detect the phase of a
counter electromotive voltage (i.e., the pole position of a motor
rotor) by the position detecting device with high accuracy. Several
techniques adapted for that purpose are also known.
[0005] For example, Patent Document 1 (Japanese Patent No. 3315847)
discloses a technique of employing a detection unit for executing
position calculation using a linear zone of a sensor signal, and
correcting sensitivity and an offset based on maximum and minimum
values of the sensor signal.
[0006] Also, Patent Document 2 (Japanese Patent No. 3397013)
discloses a technique of employing a deviation detecting unit which
detects a voltage and a current supplied to an armature coil to
estimate a rotor pole position, and which detects a position
deviation from an error between the estimated rotor pole position
and a detected rotor rotational position, and correcting the
detected position deviation.
SUMMARY OF THE INVENTION
[0007] As disclosed in Japanese Patent No. 3315847, in the field of
the synchronous motor, there is known a position detecting method
using an inexpensive Hall device as a position sensor which can not
only detect a motor rotational position during rotation, but also
determine an absolute position in the stopped state. Further, the
technique for making correction based on the maximum and minimum
values of the sensor signal is proposed to eliminate the influence
of temperature change of the Hall device itself.
[0008] However, when the position sensor is used in a vehicle
driving motor which is exposed to an abrupt temperature change, it
is required to set a voltage range of the sensor signal in
consideration of such a temperature change as well. For that
reason, detection resolution is deteriorated and the accuracy in
position detection becomes eventually insufficient, thus leading to
a problem that specifications required for the synchronous motor
cannot be satisfied. Another problem is that, because a deviation
of the mount position is not corrected, convenience in fabrication
and maintenance is not satisfactory.
[0009] Also, in the technique disclosed in Japanese Patent No.
3397013, the position detected value is corrected by calculating
(storing) a position error between the motor rotor (magnet
position) and the position sensor based on a mechanical mount
error, which is caused in motor assembly, through pole-position
estimating calculation.
[0010] However, because the position error between the motor rotor
(magnet position) and the position sensor is corrected by executing
the pole-position estimating calculation based on the motor applied
voltage, the motor current, and the motor constants, complicated
pole-position estimating calculation is required. Another problem
is that high accuracy of the position sensor cannot be achieved
because calculation errors caused by variations in the motor
constants due to mass-production and variations in accuracy of
current detection impose a limitation in correction of the position
detected value.
[0011] An object of the present invention is to provide a position
detecting device which can increase accuracy in detecting the pole
position of a motor used to perform quick acceleration and
deceleration over the range from a zero speed to a high rotation
speed, and a synchronous motor driving device using the position
detecting device. [0012] (1) To achieve the above object, the
present invention provides a position detecting device for
detecting a motor pole position by using two or more sensor
signals, the position detecting device comprising a position
detector for detecting basic-wave component signals in the sensor
signals and executing position calculation; and a correcting unit
for calculating signal information representing at least one of a
gain, an offset and a phase from the detected basic-wave component
signals, and making correction based on the calculated signal
information such that a position detection error is zero.
[0013] With those features, it is possible to increase the accuracy
in detecting the pole position of a motor which is used to perform
quick acceleration and deceleration over the range from a zero
speed to a high rotation speed. [0014] (2) In above (1),
preferably, the position detector extracts the sensor signals in a
range between predetermined first and second determination values,
thereby detecting the basic-wave component signals. [0015] (3) In
above (1), preferably, the position detector extracts sensor
signals representing middle potentials of the sensor signals,
thereby detecting the basic-wave component signals. [0016] (4) In
above (1), preferably, the sensor signal has a distorted waveform
including a predetermined zone in which a sensor signal value is
hardly changed. [0017] (5) In above (1), preferably, the sensor
signals are three-phase signals of U-, V- and W-phases, and the
position detector extracts a signal from a cross point between the
U-phase signal and the W-phase signal to a cross point between the
V-phase signal and the W-phase, a signal from a cross point between
the V-phase signal and the W-phase signal to a cross point between
the U-phase signal and the V-phase, and a signal from a cross point
between the U-phase signal and the V-phase signal to a cross point
between the U-phase signal and the W-phase, thereby detecting the
basic-wave component signals. [0018] (6) In above (1), preferably,
the correcting unit comprises an error calculator for calculating a
position error amount from a position calculated value obtained by
the position detector; and a position corrector for correcting the
position calculated value by using the position error amount
calculated by the error calculator. [0019] (7) In above (1),
preferably, the correcting unit includes a signal corrector for
determining an error amount of the sensor signal from a difference
between the signal information and a predetermined reference value,
and for correcting the sensor signal such that the determined error
amount is zero. [0020] (8) In above (1), preferably, the position
detecting device further comprises a sensor unit for generating the
sensor signal, and the sensor unit includes a magnetic adjusting
member for adjusting an amount of effective magnetic flux crossing
a sensor. [0021] (9) In above (1), preferably, the position
detecting device further comprises a sensor unit for generating the
sensor signal, and the sensor unit includes a magnetic shielding
member oriented perpendicularly to a direction in which a sensor
senses magnetic flux. [0022] (10) In above (1), preferably, the
position detecting device further comprises a sensor unit for
generating the sensor signal, and the sensor unit includes a
plurality of sensors attachable to a motor housing from the
outside. [0023] (11) To achieve the above object, the present
invention also provides a position detecting device for detecting a
motor pole position by using two or more sensor signals, the
position detecting device comprising a position detector for
detecting basic-wave component signals in the sensor signals and
executing position calculation; and a correcting unit for
calculating signal information representing at least one of a gain,
an offset and a phase from the detected basic-wave component
signals, and making correction based on the calculated signal
information such that a position detection error is zero, wherein
the position detector extracts the sensor signals in a range
between predetermined first and second determination values,
thereby detecting the basic-wave component signals.
[0024] With those features, it is possible to increase the accuracy
in detecting the pole position of a motor which is used to perform
quick acceleration and deceleration over the range from a zero
speed to a high rotation speed. [0025] (12) To achieve the above
object, the present invention further provides a permanent-magnet
synchronous motor driving device comprising a position detecting
device for detecting a motor pole position by using two or more
sensor signals; and a motor controller for producing a PWM drive
signal based on a position detected value detected by the position
detecting device, and outputting the PWM drive signal to an
inverter, thereby driving a permanent-magnet synchronous motor,
wherein the position detecting device comprises a position detector
for detecting basic-wave component signals in the sensor signals
and executing position calculation; and a correcting unit for
calculating signal information representing at least one of a gain,
an offset and a phase from the detected basic-wave component
signals, and making correction based on the calculated signal
information such that a position detection error is zero.
[0026] With those features, it is possible to increase the accuracy
in detecting the pole position of the motor which is used to
perform quick acceleration and deceleration over the range from a
zero speed to a high rotation speed. [0027] (13) In above (12),
preferably, the motor controller includes a rotor position settling
unit for driving the permanent-magnet synchronous motor through a
predetermined electrical angle in a stepping manner under PWM
control and settling a rotor stop position of the motor when the
signal information is calculated by the correcting unit, and the
correcting unit of the position detecting device calculates, at a
rotor position settled by the rotor position settling unit, a rotor
position detection error from a difference between a value of the
settled rotor position and the position detected value detected by
the position detecting device, the rotor position detection error
being corrected. [0028] (14) In above (12), preferably, the motor
controller receives a torque command, and the permanent-magnet
synchronous motor has a motive power load over a wide speed range
from a stopped state to a state of large acceleration and
deceleration. [0029] (15) In above (12), preferably, the motor
controller receives a torque command from a vehicular control unit,
and the permanent-magnet synchronous motor has a vehicular motive
power load over a wide speed range from a stopped state to a
high-speed travel state.
[0030] According to the present invention, higher accuracy can be
obtained in detecting the pole position of the motor which is used
to perform quick acceleration and deceleration over the range from
a zero speed to a high rotation speed.
BRIEF DESCRIPTION OF THE DRAWINGS
[0031] FIG. 1 is a block diagram of a synchronous motor driving
device using a position detecting device according to a first
embodiment of the present invention;
[0032] FIG. 2 is a sectional view showing the structure of a
synchronous motor and a position sensor for use with the position
detecting device according to the first embodiment of the present
invention, the view being taken along the axial direction of the
motor;
[0033] FIG. 3 is a sectional view taken along the line A-A' in FIG.
2;
[0034] FIG. 4 is a sectional view taken along the line B-B' in FIG.
2;
[0035] FIG. 5 shows waveform charts of operation signals for
explaining the operation of the position detecting device according
to the first embodiment of the present invention;
[0036] FIG. 6 shows waveform charts of the operation signals for
explaining the operation of the position detecting device according
to the first embodiment of the present invention;
[0037] FIG. 7 shows waveform charts of the operation signals for
explaining the operation of the position detecting device according
to the first embodiment of the present invention;
[0038] FIG. 8 is a graph for explaining the relationship between a
voltage vector and the pole position (phase) of a rotor in the
position detecting device according to the first embodiment of the
present invention;
[0039] FIG. 9 is a flowchart showing error detection and correction
operation executed in the position detecting device according to
the first embodiment of the present invention;
[0040] FIG. 10 is a block diagram of a position detecting device
according to a second embodiment of the present invention;
[0041] FIG. 11 shows waveform charts of operation signals for
explaining the operation of the position detecting device according
to the second embodiment of the present invention;
[0042] FIG. 12 is a sectional view showing the structure of a
synchronous motor and a position sensor for use with the position
detecting device according to a third embodiment of the present
invention, the view being taken along the axial direction of the
motor;
[0043] FIG. 13 is a sectional view taken along the line B-B' in
FIG. 12;
[0044] FIG. 14 shows waveform charts of operation signals for
explaining the operation of the position detecting device according
to the third embodiment of the present invention;
[0045] FIG. 15 is a block diagram of a synchronous motor driving
device using a position detecting device according to a fourth
embodiment of the present invention;
[0046] FIG. 16 is a schematic view showing the construction of a
hybrid vehicle to which is applied the motor driving device shown
as any of the embodiments of the present invention; and
[0047] FIG. 17 is a schematic view showing the construction of an
electric power steering system to which is applied the motor
driving device shown as any of the embodiments of the present
invention.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0048] The construction of a synchronous motor driving device using
a position detecting device according to a first embodiment of the
present invention will be described below with reference to FIGS.
1-9.
[0049] A description is first made of the construction of the
synchronous motor driving device using the position detecting
device according to this first embodiment with reference to FIG.
1.
[0050] FIG. 1 is a block diagram of the synchronous motor driving
device using the position detecting device according to the first
embodiment of the present invention.
[0051] A battery BATT serves as a DC voltage source for an inverter
801 in a motor driving device 800. A DC voltage is converted by the
inverter 801 to a three-phase AC current with a variable voltage
and a variable frequency, and the three-phase AC current is applied
to a synchronous motor 300. A position sensor unit 330 is mounted
to the synchronous motor 300 for controlling the phase of an
applied three-phase AC voltage and the phase of an induced voltage
of the synchronous motor to any desired phases. In order to control
the voltage applied to the synchronous motor 300, the DC voltage is
detected by a DC voltage sensor (not shown) and is used in a
voltage controller 805.
[0052] By using a position detected value .theta.s from a position
detecting device 700 and a detected value of the DC voltage, the
voltage controller 805 produces a drive signal, which is subjected
to pulse width modulation (PWM) in a well-known manner, in match
with a voltage command (for example, three-phase voltage commands
in the case of AC control or d- and q-axis voltage commands in the
case of vector control). The produced drive signal controls
turning-on/off of a semiconductor switching device (not shown) of
the inverter 801 through a driver, i.e., a drive circuit 804.
Herein, the position detected value .theta.s represents rotor
position information and is obtained by calculating a position
.theta. in a position detector 710 from sensor signals (Hu, Hv, Hw)
and correcting a position error contained in the calculated
position .theta. by a correcting unit 720. A correction amount CS1
is determined by an error calculator 721. The detailed operation of
the position detecting device 700 will be described later with
reference to FIGS. 5-9.
[0053] When the motor driving device 800 controls a motor rotation
speed, it calculates the motor rotation speed based on the position
detected value .theta.s supplied from the position detecting device
700 and produces a voltage command such that the calculated motor
rotation speed is matched with a speed command from a higher-level
controller. Also, when the motor driving device 800 controls motor
output torque, it detects a motor current, calculates the motor
output torque based on the detected motor current, and produces a
voltage command such that the calculated motor output torque is
matched with a torque command from the higher-level controller.
[0054] The structure of the synchronous motor and a position sensor
(rotation sensor of the position sensor unit) for use with the
position detecting device according to this first embodiment will
be described below with reference to FIGS. 2-4.
[0055] FIGS. 2-4 show the structure of the synchronous motor and
the position sensor for use with the position detecting device
according to the first embodiment of the present invention. More
specifically, FIG. 2 is a sectional view showing the structure of
the synchronous motor and the position sensor for use with the
position detecting device according to the first embodiment of the
present invention, the view being taken along the axial direction
of the motor. FIG. 3 is a sectional view taken along the radial
direction of the motor, specifically taken along the line A-A' in
FIG. 2, and FIG. 4 is a sectional view taken along the radial
direction of the motor, specifically taken along the line B-B' in
FIG. 2.
[0056] The motor shown in FIGS. 3 and 4 is a permanent-magnet
synchronous motor in which a permanent magnet is used to generate a
magnetic field, more specifically a permanent-magnet synchronous
motor of the magnet embedded type that a permanent magnet is
embedded in a rotor core 321. Further, the motor is an inner-rotor
type motor in which a rotor (motor rotor) 320 is arranged inside
the stator core 311 with a gap left between them, and it is a
concentrated-winding motor in which a winding 312 is wound over
each single tooth 311T of a stator core. The rotor 320 comprises
the rotor core 321, a permanent magnet 322, and a motor shaft 360.
The motor shaft 360 constituting the rotor 320 is rotatably
supported by a bearing 350, and the position sensor unit 330
including rotation sensors is fixed to a motor housing. A stator
310 is also fixed to the motor housing by press-fitting or by using
a key way. The winding 312 of the stator 310 is constituted by
three-phase windings 312U, 312V and 312W of U, V and W phases which
are arranged in sequence. The permanent magnet 322 of the rotor 320
is constituted by permanent magnets 322N each having an N polarity
on the side facing the gap and permanent magnets 322S each having
an S polarity on the side facing the gap, those permanent magnets
322N and 322S being alternately arranged in the circumferential
direction.
[0057] The permanent magnet 322 is preferably made of a rare-earth
magnet of Nd--Fe--B, for example, from the viewpoints of coercive
force and cost, but it may also be made of another kind of
rare-earth magnet or a ferrite magnet. In the latter case, only a
motor output characteristic is changed. Also, the motor in this
first embodiment is illustrated as constituting a 2-to-3 system
which has 16 poles and 24 slots (i.e., a motor having poles and
slots which are respectively integer times the pole number=2 and
the slot number=3), but it may be constituted as any of other
4-to-3, 8-to-9 and 10-to-12 systems. Further, while this first
embodiment is described in connection with the magnet-embedded and
concentrated-winding motor, another surface-magnet or
distributed-winding motor can also be used without problems.
[0058] As shown in FIG. 4, the position sensor unit 330 is made up
of rotation sensors H1, H2 and H3 which are arranged at an angular
interval of 120.degree. in terms of motor electrical angle between
the two sensors. The rotation sensors H1, H2 and H3 detect changes
of magnetic flux from the same permanent magnet with rotation of
the motor rotor in the sequence of the sensor
H1.fwdarw.H2.fwdarw.H3 (or the reversed sequence when the direction
of the rotation is reversed). While the rotation sensors H1, H2 and
H3 are shown as being arranged at an angular interval of
120.degree. in terms of electrical angle between the two sensors,
they may be arranged at an angular interval of 120.degree. in terms
of mechanical angle between the two sensors. In the latter case,
however, because the rotation sensors H1, H2 and H3 detect magnetic
fluxes from different magnets and influences of errors caused in
mounting (bonding) the permanent magnets appear, it is required to
prepare memories in number corresponding to the number of the
permanent magnets and to make correction through a learning process
in a repeated manner.
[0059] The rotation sensors H1, H2 and H3 are usually packaged into
one unit including sensor elements, an amplification circuit, a
temperature correction circuit, etc. As an alternative, only the
sensor elements may be disposed in the position sensor unit 330,
whereas the amplification circuit, the temperature correction
circuit, etc. may be disposed on a printed board constituting the
motor driving device 800. Further, a plurality of rotation sensors
may be packaged into one member and mounted in the position sensor
unit 330. Conversely, the position detecting device 700 may be
manufactured in the form of a dedicated IC, and the sensor elements
and the position detecting device 700 may be both mounted in the
position sensor unit 330. In the last-mentioned case, the number of
outputs from the position sensor unit 330 can be reduced from three
(Hu, Hv and Hw) to one (.theta.s) by causing one output of the
position detected value .theta.s to be issued as analog output or a
digital output after being subjected to the pulse width
modulation.
[0060] The operation of the position detecting device 700 according
to this first embodiment will be described below with reference to
FIGS. 5-7.
[0061] FIGS. 5-7 show waveform charts of operation signals for
explaining the operation of the position detecting device according
to the first embodiment of the present invention. FIG. 5 shows
examples of the operation signals in an ideal state, FIG. 6 shows
examples of the operation signals including a gain error and an
offset error, and FIG. 7 shows examples of the operation signals
including a phase error.
[0062] First, the operation signals being in the ideal state and
containing no errors are described below with reference to FIG. 5.
The vertical axis in (a) of FIG. 5 represents a sensor signal,
i.e., an output signal of the position sensor unit 330. The
vertical axis in (b) of FIG. 5 represents an extracted signal
extracted by the position detector 710. The vertical axis in (c) of
FIG. 5 represents a phase signal VO obtained by the position
detector 710. The vertical axis in (d) of FIG. 5 represents signal
information SI1 obtained by the correcting unit 720. The vertical
axis in (e) of FIG. 5 represents an angle signal indicating the
detected position (absolute position) .theta.s, which is outputted
from the position detecting device 700. The horizontal axis in (a)
to (e) of FIG. 5 represents an angle, namely an ideal phase
(angle).
[0063] Further, in (a) to (d) of FIG. 5, a solid line represents a
U-phase signal, a dotted line represents a V-phase signal, and a
one-dot chain line represents a W-phase signal.
[0064] In the position detecting device 700, the position detector
710 receives, as input signals, the sensor signals Hu, Hv and Hw
from the position sensor unit 330 and calculates the position
.theta. based on the detected input sensor signals (also called the
detected sensor signals), and the correcting unit 720 calculates
the error correction amount CS1 and executes feed-forward
correction for correcting the position detected value .theta.s that
indicates the rotational angle of the synchronous motor 300.
[0065] As shown in FIG. 4, three Hall devices are mounted to the
position sensor unit 330 at the angular interval of 120.degree. in
terms of electrical angle. Therefore, the position sensor unit 330
outputs the sensor signals Hu, Hv and Hw corresponding to the U-,
V- and W-phases shifted from each other by 120.degree. in terms of
electrical angle as shown at (a) in FIG. 5, which are inputted to
the position detector 710.
[0066] The position detector 710 has a maximum value VP1 and a
minimum value VP2 in its input range. The maximum value VP1 and the
minimum value VP2 of the input range are values limited due to an
input range of a power supply voltage for an analog circuit or an
input range of an A/D converter. For example, the signals inputted
to the position detector 710 are signals (detected sensor signals)
ranging from VP1.apprxeq.5 V to VP2.apprxeq.0 V. The position
detector 710 compares the inputted sensor signals with a first
determination value VL1 larger than an average value (central
value) VP0 of the sensor signals and with a second determination
value VL2 smaller than the average value, and extracts the sensor
signals in the range between the first and second determination
values VL1, VL2, thereby detecting the extracted signals shown at
(b) in FIG. 5. The extracted signals are thus given as signals
obtained by clamping the sensor signals Hu, Hv and Hw based on the
first and second determination values VL1 and VL2. Stated another
way, the average value shown at (a) or (b) in FIG. 5 represents a
middle potential of the sensor signals, and signals near the middle
potential are extracted as the extracted signals.
[0067] Herein, the first and second determination values VL1, VL2
are preferably set as follows: VP1.gtoreq.VL1.gtoreq.(1/2 of crest
value of sensor signal)+VP0 VP2.ltoreq.VL2.gtoreq.VP0-(1/2 of crest
value of sensor signal) Thus, they are values that can be set when
a maximum correction amount of each sensor signal is designed.
[0068] The extracted signals are signals which are used in position
calculation and constitute basic wave components of the sensor
signals. In other words, each of the extracted signals has a
waveform obtained by cutting the vicinity of a peak value of the
sensor signal and can be utilized in the range of a signal level
with high sensitivity (although sensitivity becomes non-linear
depending on the extracted zone, no problems arise in position
detection).
[0069] Further, higher harmonic components superimposed on the
sensor signal near its peak can be removed. In particular, as
described later with reference to FIG. 6, when the sensor signal
exceeds the input range, the detected sensor signal is given as a
signal having a trapezoidal waveform, which has been subjected to
peak cutting based on the maximum and minimum values (VP1, VP2) of
the input range and includes many higher harmonics. However, the
position calculation logic used in this first embodiment is not
affected by those higher harmonics. The peak-cut signal has a
distorted waveform including a zone where the sensor signal value
is not changed. Stated another way, even when the sensor signal has
a distorted waveform resulting from the peak cutting (i.e., even
when a function of tan.sup.-1 cannot be used to determine a phase),
position detection and position correction can be performed.
[0070] In addition, the position detector 710 calculates, from
signal values of the extracted signals, corresponding phases by
using the inverse trigonometric function or by executing a table
search, thereby obtaining signal position calculated values FS1.
The signal position calculated values FS1 are each given as a phase
signal VO (at (c) in FIG. 5) near.+-.30.degree. while the phase (0,
60, 120, 180, 240 and 300 in terms of angle) of a cross point
between each ideal sensor signal and the central value is used as a
reference phase value.
Phases VO1 and VO2 corresponding to the first and second
determination values have the relationship of:
VO2.ltoreq.VO.ltoreq.VO1 The phase signal VO can be given by
selecting one of values near the cross point between the signals of
different phases, or by using one of values including the cross
point itself as well. The signals of the respective phases in the
ideal state cross each other at points of .+-.30.degree.. Thus, the
position .theta. can be obtained as an absolute angle in the range
of 0-360.degree. by repeatedly summing the phase signal VO which
has been detected with a reference phase value being a base.
[0071] The correcting unit 720 calculates the signal information
SI1, shown at (d) in FIG. 5, from the phase signal VO obtained by
the position detector 710. Further, the correcting unit 720 detects
an error from the difference between the signal information SI1 and
a reference value PR, to thereby correct the position .theta.. For
example, the correcting unit 720 can obtain the signal information
SI1, shown at (d) in FIG. 5, through the steps of differentiating
the phase signal VO, dividing the differentiated value by an output
frequency f' of the inverter, and taking an absolute value of the
divided value. Assuming here that a basic wave component of the
phase signal VO is sin(.omega.t+.alpha.) (.alpha. is each phase of
three-phase signals), its differentiated value is given by
.omega.cos(.omega.t+.alpha.). Because .omega.=2.pi.f is held and a
frequency f of the electrical angle of the synchronous motor is
substantially equal to the output frequency f' of the inverter, the
reference value PR (2.pi. in this example) of a signal gradient is
obtained by dividing .omega. by the output frequency f' of the
inverter.
[0072] Stated another way, the signal information SI1 is signal
information containing information of both gain and offset, and the
position detected value .theta.s after correction is obtained
through the steps of executing correction calculation on the signal
information SI1 to calculate the error correction amount CS1, and
adding the error correction amount CS1 to the position .theta..
Note that since the signal waveforms shown in FIG. 5 represent
ideal signals, the signal information SI1 is a constant value in
match with 2.pi. of the reference value PR in the range of
effective position information. Therefore, gain=1 and offset=0 are
detected and the error correction amount CS1 resulting from the
correction calculation is zero.
[0073] Thus, a signal having been subjected to peak cutting and
having a trapezoidal waveform is avoided from becoming the detected
sensor signal used in the position calculation, and stable position
calculation can be realized. Also, by setting the first and second
determination values (VL1, VL2) to fall within the range of the
maximum and minimum values (VP1, VP2) of the input range, the
position detection can be performed while making a pass/fail check
of the calculation for the position detection. For example, the
position calculation is determined as being properly executed if
the corresponding phase signal VO satisfies the relationship of:
VO2<VO and VO<VO1 It is therefore possible to check
abnormality in the position detection and to prevent runaway of the
motor.
[0074] The operation signals in the state where the sensor signals
include a gain error and an offset error will be described below
with reference to FIG. 6. The vertical axes in (a) to (e) of FIG. 6
represent the same items as those in (a) to (e) of FIG. 5. The
horizontal axis represents an angle, namely an ideal phase (angle),
as in FIG. 5.
[0075] FIG. 6 shows, as seen from (a) of FIG. 6, an example in
which a gain of the U-phase sensor signal Hu is increased and an
offset (+) is caused in the V-phase sensor signal Hv. The W-phase
sensor signal Hw is identical to that having the ideal values. A
zone of the U-phase sensor signal Hu exceeding the maximum and
minimum values (VP1, VP2) of the input range is indicated by a
dotted line, and the detected sensor signal is indicated by a solid
line after being clamped based on the maximum and minimum values
(VP1, VP2).
[0076] Because the position calculation is executed using the
levels (determination values) VL1, VL2 of the extracted signals,
influences of the gain error and the offset error also appear in
the phase signal VO as shown at (c) in FIG. 6.
[0077] Accordingly, as shown at (d) in FIG. 6, the influences of
both the errors are detected in the U and V phases of the signal
information SI1 obtained through the signal position calculation.
The detected position is shown such that a solid line represents
the detected value not yet subjected to the error correction, and a
broken line represents the ideal detected value. Further, the gain
error appears such that, in two detection zones located within
0-360.degree., error amounts (i.e., amounts of position error with
respect to the ideal detected value) at the start and the end of
each zone take positive and negative values. The offset error
appears such that error amounts (i.e., amounts of position error
with respect to the ideal detected value) in two detection zones
located within 0-360.degree. take positive and negative values.
Thus, by detecting those error amounts, the gain error and the
offset error can be corrected.
[0078] A practical detection method will be described below. A gain
error Gain1 detected on the U-phase signal is detected as a gain
error coefficient Ge and is expressed by: (gain error coefficient
Ge)=(signal information SI1)/(reference value PR) However, the gain
error coefficient Ge has an effective value only in a zone of
effective position information (U-phase signal information
SI1.noteq.0). Also, the timing of detecting the gain error is
preferably near a central value of the sensor signal. In one
example, that timing corresponds to the timing at which the change
of the U-phase signal information SI1.apprxeq.0 is obtained. In
another example, that timing corresponds to the timing of
zero-crossing that is obtained by differentiating the extracted
signal twice.
[0079] On the other hand, an offset error Off1 detected on the
V-phase signal is detected as an offset error coefficient Oe given
by the difference between two detection points (Off11, Off12) as
follows: (offset error coefficient Oe)=Off11-Off12 However, the
offset error coefficient Oe has an effective value only in a zone
of effective position information (V-phase signal information
SI1.noteq.0). Also, the timing of detecting the offset error is
preferably a point in time near each of the start and the end of
the effective position information. In one example, that timing
corresponds to the timing at which the V-phase signal information
SI1.noteq.0 is obtained. In another example, that timing
corresponds to the timing at which the extracted signal takes a
value almost equal to each of the first and second determination
values (VL1, VL2).
[0080] The correcting unit 720 calculates a phase to be corrected
by using the gain error coefficient Ge and the offset error
coefficient Oe, and corrects the error correction amount CS1 with
respect to the position .theta. (i.e., the detected value not yet
subjected to the error correction (indicated by the solid line)),
thereby obtaining the position detected value .theta.s that is made
asymptotic to the ideal detected value (indicated by the broken
line).
[0081] Alternatively, the gain error and the offset error may be
detected as follows. Assuming the detected sensor signal including
errors to be expressed by Asin(.omega.t+.alpha.)+D,
2.pi.Acos(.omega.t+.alpha.) is obtained, as described above, by
differentiating it and dividing the differentiated value by the
output frequency f' of the inverter. Therefore, 2.pi.A is obtained
by measuring a maximum value of the latter signal, and an offset D
is obtained from the detected sensor signal at the timing of
(.omega.t+.alpha.=0) at which the maximum value is obtained (the
timing at which the maximum value is obtained may be a
zero-crossing point resulting from differentiating the detected
sensor signal twice, which corresponds to the zone of the extracted
signal). Thus, the gain error can be determined as A (i.e., a ratio
of 2.pi.A to the reference value 2.pi.), and the offset error can
be determined as D.
[0082] The operation signals in the state where the sensor signal
includes a phase error will be described below with reference to
FIG. 7. The vertical axes in (a) to (e) of FIG. 7 represent the
same items as those in (a) to (e) of FIG. 5. The vertical axis in
(f) of FIG. 7 represents signal information SI2 obtained by the
correcting unit 720. The horizontal axis represents an angle,
namely an ideal phase (angle), as in FIG. 5.
[0083] FIG. 7 shows, as seen from (a) of FIG. 7, an example in
which the phase of only the V-phase sensor signal Hv is advanced.
The U- and W-phase sensor signals Hu, Hw are identical to those
having the ideal values.
[0084] Because the position calculation is executed using the
levels (determination values) VL1, VL2 of the extracted signals, an
influence of the phase error also appear in the phase signal VO
(i.e., a state where the V phase is advanced), as shown at (c) in
FIG. 7. The levels of the V-phase extracted signal are
substantially equal to the ideal values and therefore an error is
not caused in the position calculation. Hence the influence of the
phase error does not appear in the signal information SI1 obtained
through the signal position calculation.
[0085] A phase shift in the V phase appears as a phase error
relative to each reference phase value (120.degree., 300.degree.)
in the V phase, and the position .theta. is given by the detected
value (indicated by the solid line) not yet subjected to the error
correction. The phase error differs from the offset error in that
error amounts (i.e., amounts of position error relative to the
ideal detected value) in two detection zones located within
0-360.degree. appear excessive or deficient in the same direction.
Thus, by detecting those error amounts, the phase error can be
corrected.
[0086] A practical correction method will be described below. The
correcting unit 720 detects a phase shift by using the signal
information SI2 shown at (f) in FIG. 7. The signal information SI2
is produced from the signal obtained through the signal position
calculation (or obtained from the extracted signal as an
alternative) and has a signal waveform produced from values at
cross-points between the respective signals or middle values of the
respective signals. Incidentally, the ideal signals used in the
signal position calculation cross each other at points of
.+-.30.degree..
[0087] To detect the phase shift, the U-phase signal is used as a
reference phase. More specifically, an amount of phase shift of the
V-phase signal relative to the U-phase signal (hereinafter referred
to simply as a "V-phase signal phase error UV"), and an amount of
phase shift of the W-phase signal relative to the U-phase signal
(hereinafter referred to simply as a "W-phase signal phase error
UW") are determined. In (f) of FIG. 7, ideal phase differences at
cross points between the U- and V-phase signals (two points within
0.degree.-360.degree.) are assumed to be Phas11 and Phas12. The
V-phase signal phase error UV appears in the case of
Phas11.noteq.Phas12 and is detected by taking one of those two
values or an average value thereof as an error amount. The detected
error amount is used as the error correction amount CS1 in the
phase detection zone for the V phase to correct the phase
(position) .theta. (i.e., the detected value (indicated by the
solid line) not yet subjected to the error correction), thereby
obtaining the position detected value .theta.s that is made
asymptotic to the ideal detected value (indicated by the broken
line).
[0088] As described above, the position detection can be performed
even when the inputted sensor signal has a signal waveform in
excess of the input range or it has a distorted waveform resulting
from clamping made in match with the input range. Also, even when
an error is superimposed on the sensor signal, the position
detected signal .theta.s can be obtained with accuracy as high as
when the ideal sensor signal is inputted, by correcting the
superimposed error.
[0089] On the other hand, assuming the case where the U-phase
signal as the reference signal also has a phase shift (hereinafter
referred to simply as a "U-phase signal phase error UU"), the
accuracy of the position detection in the range of
0.degree.-360.degree. can be increased by the above-described phase
error correction, but an error of the absolute position (i.e., the
phase shift of the U-phase signal) cannot be corrected by only the
above-described phase error correction. Even when such an error of
the absolute position remains, that error affects neither a ripple
of the voltage applied to the motor nor a ripple of the motor
current (equivalent to a torque ripple), and therefore the motor
can be smoothly driven with high accuracy. However, that error
appears as a drop of the motor efficiency when the motor is rotated
at high speed. One preferable method for reducing the error of the
absolute position comprises the steps of obtaining the U-phase
signal phase error UU from the V-phase signal phase error UV, the
W-phase signal phase error UW, and a phase difference VW between
the V- and W-phase signals by calculation, and averaging those
phase errors to reduce an overall amount of the phase error.
Another preferable method comprises the steps of applying a
predetermined voltage vector to drive the motor in a stepping way,
and detecting the U-phase signal phase error UU.
[0090] The relationship between a voltage vector and the pole
position (phase) of the rotor in the position detecting device 700
according to this first embodiment will be described below with
reference to FIG. 8.
[0091] FIG. 8 is a graph for explaining the relationship between a
voltage vector and the pole position (phase) of the rotor in the
position detecting device according to the first embodiment of the
present invention.
[0092] The sensor signals from the position sensor unit 330 are
mathematically expressed by the following formulae (1), (2) and
(3): HU.sub.(n)=A.sub.1sin(.theta.a.sub.(n)+.alpha.)+C.sub.1 (1)
HV.sub.(n)=A.sub.2sin(.theta.a.sub.(n)-120+.beta.)+C.sub.2 (2)
HW.sub.(n)=A.sub.3sin(.theta.a.sub.(n)+120+.gamma.)+C.sub.3 (3)
While the sensor signals are mathematically expressed here as
sine-wave signals for the sake of simplicity, the following
discussion is similarly applied to trapezoidal-wave signals
containing higher harmonics. In the formulae, a suffix (n)
represents a value corresponding to arbitrary sampling, and
.theta.a represents the pole position (phase) of the rotor.
Further, (A1, A2, A3), (C1, C2, C3) and (.alpha., .beta., .gamma.)
represent respectively gains, offsets and phase errors of the
sensor signals (HU, HV, HW). The gain error, the offset error, the
V-phase signal phase error UV, and the W-phase signal phase error
UW of the sensor signals are detected as described above.
Accordingly, if the U-phase signal phase error UU (=.alpha.) is
obtained, the absolute position can also be corrected by setting
.beta.=V-phase signal phase error UV-.alpha.) and .gamma.=W-phase
signal phase error UW-.alpha..
[0093] On the other hand, in the three-phase inverter 801, because
the output voltage takes 0 or 1 for each of the three phases, there
are eight (=third power of 2) combinations in total. The pole
position (phase) of the rotor is decided by voltage vectors
expressed using those eight combinations of the output voltages.
Generally, the pole position (phase) is not definitely decided by
two voltage vectors which are called zero voltage vectors, i.e., a
voltage vector V0 and a voltage vector V7. However, the pole
position (phase) can be definitely decided by applying the
remaining six voltage vectors for a time sufficient for the rotor
to rotate. For example, in the case of a voltage vector V1 in which
the U-phase voltage is 1 and the voltages in the other phases are
0, the relationship among respective currents (IU, IV, IW) in U, V
and W phases is such that a current of IU=-(IV+IW) flows and the
rotor is stopped in a predetermined electrical angle position
(phase). That description is similarly applied to the other voltage
vectors. Thus, the rotor can be held in an electrical angle
position (phase) at intervals of 60.degree. by applying the six
voltage vectors in total.
[0094] In such a case, the phase error detection logic is required
to be executed at the start or the end of the motor. However, since
the phase error is attributable to a mounting error, the phase
error detection logic is just required to be executed once when the
sensor is mounted or replaced, by storing the detection result. In
other words, the phase error detection logic is not always required
to be executed at the ordinary start of the motor.
[0095] The electrical angle position (phase) representing the rotor
pole position, which corresponds to the applied voltage vector, is
compared with the phase of the sensor signal obtained by the
position detecting device 700 and expressed by each of the formulae
(1) to (3) (which may be any of the signal position calculated
value FS1, the phase (position) .theta., and the position detected
value .theta.s). The thus-detected error is employed as the phase
shift (error amount) in the U phase to correct the phase errors in
the V and W phases. Alternatively, the position detected value
.theta.s may be corrected. As a result, even when the phase shift
is caused in the U-phase signal and the phase errors of the V- and
W-phase signals are detected with the U-phase signal employed as
the reference phase, the phase correction can be realized with high
accuracy.
[0096] Preferably, a plurality of voltage vectors are applied so as
to rotate the rotor forward and backward, and the electrical angle
position (phase) .theta.a of the rotor and the phase obtained by
the position detecting device 700 are compared with each other
successively plural times. Such a process enables the phase
correction to be performed with higher accuracy even when load
torque or friction torque is caused. Also, by adjusting the time of
application of the voltage vector at intervals of 60.degree.
(through PWM (pulse width modulation)), it is also possible to
detect the phase error even at an intermediate angle (phase) during
the intervals of 60.degree..
[0097] The above description is made on assumption that the gains
(A1, A2, A3) and the offsets (C1, C2, C3) of the sensor signals
(HU, HV, HW) are known. However, by applying a plurality of voltage
vectors and detecting the sensor signals (HU, HV, HW) at each
electrical angle position (phase) .theta.a of the rotor, the
above-mentioned simultaneous equations (1), (2) and (3) can be
solved to obtain the respective errors of the sensor signals. Such
a method is effective, for example, in making pass/fail
determination of a product in an offline mode (i.e., in a
correction value detection mode separate from actual operation)
during production of products, etc., or setting initial values in a
shorter time.
[0098] The error detection and the correction operation executed in
the position detecting device 700 according to this first
embodiment will be described below with reference to FIG. 9.
[0099] FIG. 9 is a flowchart showing the error detection and the
correction operation executed in the position detecting device
according to the first embodiment of the present invention.
[0100] When source power is turned on, the position detecting
device 700 first calls, in step F1, preceding gain, offset, and
phase correction values from, e.g., a flash memory and sets them as
correction initial values.
[0101] Then, in step F2, the voltage controller 805 starts to
rotate the motor by using the position detected value .theta.s from
the position detecting device 700. Here, the position detected
value .theta.s is obtained by correcting the output .theta. of the
position detector 710 with correction initial values CS1 set in
step F1.
[0102] In step F3, the error calculator 721 calculates the phase
.theta. during the motor operation as described above and detects
respective sensor errors.
[0103] In step FS1, the position detecting device 700 checks the
magnitude and variation of each of the detected errors. If the
detected error is larger than a predetermined allowable error
range, the position detecting device 700 proceeds to step F5 in
which the presence of a large detection error is informed
(displayed) to a higher-level ECU (engine control unit) and a
control panel, followed by bringing the control flow to an end. If
the detected error is within the predetermined allowable error
range, the position detecting device 700 proceeds to step F4 in
which the gain error correction value is updated.
[0104] In step FS2, based on the magnitude of the offset error, the
position detecting device 700 determines whether or not the offset
error has the magnitude or it reaches the timing to update the
correction value. If the update is to be performed, the position
detecting device 700 updates the offset correction value in step
F6. Otherwise, the position detecting device 700 determines in step
FS3, based on the magnitude of the phase error, whether or not the
phase error has the magnitude or it reaches the timing to update
the correction value. If the update is to be performed, the
position detecting device 700 updates the phase correction value in
step F7.
[0105] In step F8, the error calculator 721 applies the correction
amount CS1 to the phase .theta., to thereby obtain the position
detected value .theta.s after the correction.
[0106] If it is determined in step FS4 that the motor is not to be
stopped, the position detecting device 700 repeats the
above-described control flow. If the motor is to be stopped, the
position detecting device 700 stores, in step F9, final gain,
offset, and phase correction values in, e.g., the flash memory.
Because the temperature generally differs between the start and the
stop of the motor operation, the correction values after the lapse
of a certain time from the start of the operation subsequent to the
power-on may be stored after a repeated learning process instead of
storing the final gain and offset correction values. Since the
phase correction-value is a value to correct a mechanical mounting
error and is not so changed, it can be stored as a correction value
obtained through a repeated learning process or by taking a moving
average. In addition, the phase correction value is preferably
reset after maintenance work, for example, when the position sensor
unit 330 is replaced.
[0107] Thus, even when the sensor suffers from change of the design
temperature, deterioration, etc. after shipment of the product,
sensor correction or failure detection can be performed online
(during actual operation).
[0108] While the above description is made of the method of
extracting the basic wave component from the detected sensor signal
through the determination process, the position detection and the
position correction can also be similarly performed by extracting
the basic wave component through a frequency analysis, such as FFT,
of the peak-cut sensor signal or the detected sensor signal. In
that case, the position .theta. can be easily obtained, for
example, by using a signal value resulting from inverse transform
of only the basic wave, and the position correction can also be
easily performed by using the amplitude, the offset and the phase
which have been obtained by the waveform analysis.
[0109] According to this first embodiment, as described above,
since the gain error, the offset error, and the phase error can be
corrected on the sensor signals detected by the position sensor
unit 330, the detection accuracy of the pole position can be
increased. It is hence possible to increase the accuracy in
detecting the pole position of the motor which is used to perform
quick acceleration and deceleration over the range from a zero
speed to a high rotation speed.
[0110] Further, by correcting the position error after the
power-on, a deviation of the mount position can also be corrected,
thus resulting in higher convenience in both fabrication and
maintenance.
[0111] Moreover, the correction of the position error also makes it
possible to correct a calculation error that is caused by a
variation in motor constants attributable to mass production of
motors and a variation in accuracy of current detection.
[0112] The construction of a synchronous motor driving device using
a position detecting device according to a second embodiment of the
present invention will be described below with reference to FIGS.
10 and 11. Note that the basic construction of the synchronous
motor driving device using the position detecting device according
to the second embodiment is the same as that shown in FIG. 1.
[0113] FIG. 10 is a block diagram of the position detecting device
according to the second embodiment of the present invention. FIG.
11 shows waveform charts of operation signals for explaining the
operation of the position detecting device according to the second
embodiment of the present invention.
[0114] The vertical axes in (a), (b), (c) and (f) of FIG. 11
represent the same items as those in (a), (b), (c) and (f) of FIG.
7. The vertical axis in (a2) of FIG. 11 represents the sensor
signal after correction, and the vertical axis in (d2) of FIG. 11
represents signal information SI3. The horizontal axis represents
an angle, namely an ideal phase (angle), as in FIG. 5.
[0115] FIG. 10 differs from FIG. 1 in that a position detecting
device 700A corrects the inputted sensor signals Hu, Hv and Hw in a
correcting unit 720A by using a signal corrector 722A. Stated
another way, while the signal error is corrected by feed forward
control in the first embodiment of FIG. 1, the signal error is
corrected by feedback control in this second embodiment.
[0116] Also, the extracted signal shown at (b) in FIG. 11 greatly
differs from the extracted signal in FIG. 7 in that the former has
a signal waveform produced from values at cross-points between the
respective sensor signals after correction ((a2) in FIG. 11) or
middle values of those signals. Operations indicated by other
identical characters are the same as those in FIG. 7.
[0117] The sensor signals Hu, Hv and Hw inputted to the position
detecting device 700A in FIG. 10 are given as the detected sensor
signals (indicated by respective lines at (a) in FIG. 11 after
clipping) and are corrected by the correcting unit 720A in
accordance with a correction amount CS2 outputted from the signal
corrector 722A, followed by being outputted as the corrected sensor
signals Su, Sv and Sw after the correction. The position detector
710A calculates the phase from the corrected sensor signals Su, Sv
and Sw and outputs the phase signal VO. The correcting unit 720A
detects the gain error, the offset error, and the phase error,
which are contained in the sensor signals, from the phase signal
outputted from the position detector 710A, and outputs the
correction amount CS2. The corrected sensor signals Su, Sv and Sw
are obtained by executing signal calculation using the correction
amount CS2.
[0118] In an example shown at (a) in FIG. 11, the V-phase sensor
signal Hv contains the gain error and the offset error, and the
W-phase sensor signal Hw contains the phase error. The U-phase
sensor signal Hu takes the same value as the ideal one.
[0119] The corrected sensor signals Su, Sv and Sw having been
subjected to the correction by the signal corrector 722A, shown at
(a2) in FIG. 11, are signals resulting from executing the
correction calculation on the detected sensor signals (i.e., the
sensor signals having waveforms indicated by respective lines at
(a) in FIG. 11 after clipping) at the timing of each cross point
between those sensor signals.
[0120] The extracted signal shown at (b) in FIG. 11 is a signal
produced by the position detector 710A based on values at the
cross-points between the corrected sensor signals or middle values
of those signals, and it is given as one signal resulting from
combining the three signals with each other. The position detector
710A executes the signal position calculation based on the
extracted signal and obtains the phase signal VO shown at (c) in
FIG. 11.
[0121] The signal corrector 722A executes signal calculation (e.g.,
differentiation) on the phase signal VO and obtains the signal
information SI3 shown at (d2) in FIG. 11. The signal information
SI3 is detected as a signal superimposed with information of the
gain error and the offset error contained in the three sensor
signals. Also, the signal corrector 722A detects information of the
phase error from the signal information SI2. The phase error is
given as a difference, indicated by Phas1(n), Phas2(n) and Phas3(n)
(where n is the sampling number), relative to the ideal values
(.+-.30.degree.) at the timing at which the sensor signals cross
each other.
[0122] Based on the signal information SI3, a gain error Gain2 and
an offset error Off21 can be detected in a first detection zone
(about 100.degree.-150.degree.) in the V phase. The detected gain
error Gain2 is multiplied by a calculation coefficient for the gain
correction to correct the V-phase detected sensor signal at the
timing at which the U- and W-phase signals cross each other. In
other words, the gain correction is made at the timing near
210.degree. along the horizontal axis. At that timing, the V-phase
signal is not used in the position calculation. Accordingly, even
if abrupt correction is applied to the V-phase detected sensor
signal, the position detecting device 700A can always calculate the
position .theta. in a stable way.
[0123] The offset error is also corrected in a similar manner. More
specifically, the offset error Off2 is detected as follows:
Off2=k*{(Off21-Gain2)-(Off22-Gain2)} The detected offset error is
multiplied by a calculation coefficient for the offset correction
to correct the V-phase detected sensor signal at the timing at
which the U- and W-phase signals cross each other. The offset
correction is preferably performed using an average value for one
cycle (i.e., an average of the values detected in two detection
zones per cycle). Stated another way, the offset correction is made
at the timing near 380.degree. along the horizontal axis. In the
above formula, k is the calculation coefficient and is 0.5, for
example.
[0124] A W-phase signal phase error Phas32 is detected from the
signal information SI2, and the phase correction is also performed
at the same timing as the offset correction. The W-phase signal
phase error Phas32 is detected as follows:
Phas32=k*{Phas3(1)-Phas3(2)} The detected phase error is multiplied
by a calculation coefficient (generally nearly 1 (.apprxeq.1)) for
the phase correction, to thereby correct the sensor signal through
the phase correction calculation of the detected sensor signal.
Moreover, the phase error is detected using an average value for
one cycle (i.e., an average of the values detected in two detection
zones per cycle). Preferably, the phase error is not detected
during a period in which the offset correction is performed, and
the detection of the phase error and the phase correction is
performed after the offset correction result has been converged
into a predetermined allowable error range.
[0125] Thus, the detected sensor signal can be corrected in a
feedback manner such that each sensor error is converged into the
predetermined allowable error range. Therefore, the position
detection can be stably performed with high accuracy.
[0126] Additionally, the gain error, the offset error, and the
phase error can also be detected using only the signal information
SI2. For example, the gain error of the U-phase signal can be
detected as Phas2 (1).apprxeq.Phas3 (1), and the offset error of
the V-phase signal can be detected as Phas3 (2).apprxeq.-Phas1 (1).
By multiplying the thus-detected error by the correction
coefficient, the error can be corrected in a feedback manner. This
method is effective in simplifying processing required for the
signal calculation.
[0127] Further, the gain error and the offset error may be
corrected in a feedback manner, while the phase error may be
corrected in a feed forward manner as described in the first
embodiment.
[0128] According to this second embodiment, as described above,
since the gain error, the offset error, and the phase error can be
corrected on the sensor signals with higher accuracy, the detection
accuracy of the pole position can be further increased. It is hence
possible to increase the accuracy in detecting the pole position of
the motor which is used to perform quick acceleration and
deceleration over the range from a zero speed to a high rotation
speed.
[0129] The construction of a synchronous motor driving device using
a position detecting device according to a third embodiment of the
present invention will be described below with reference to FIGS.
12-14. Note that the basic construction of the position detecting
device according to the third embodiment is the same as that shown
in FIG. 10. Also, the basic construction of the synchronous motor
driving device using the position detecting device according to the
third embodiment is the same as that shown in FIG. 1.
[0130] FIG. 12 is a sectional view showing the structure of a
synchronous motor and a position sensor for use with the position
detecting device according to the third embodiment of the present
invention, the view being taken along the axial direction of the
motor. FIG. 13 is a sectional view taken along the radial direction
of the motor, i.e., taken along the line B-B' in FIG. 12. FIG. 14
shows waveform charts of operation signals for explaining the
operation of the position detecting device according to the third
embodiment of the present invention.
[0131] The vertical axes in (a), (b), (c), (d2) and (f) of FIG. 14
represent the same items as those in (a), (b), (c), (d2) and (f) of
FIG. 11. The horizontal axis represents an angle, namely an ideal
phase (angle), as in FIG. 5.
[0132] FIGS. 12 and 13 show the structure of the synchronous motor
and the position sensor. FIGS. 12 and 13 differ from FIG. 2 in that
a sensor rotor is used in the position sensor unit. The operations
of the other components indicated by the same characters are the
same as those in FIG. 2. FIG. 14 shows the operation signals and
differs from FIG. 2 in using two sensors.
[0133] The position sensor unit shown in FIGS. 12 and 13 comprises
a sensor rotor 340 made of a magnetic material and sensor units
330U and 330V (suffixes U and V are added to identify two sensor
units individually). Gear teeth (in number 32) are formed along an
outer circumference of the sensor rotor 340 (in facing relation to
the sensor units).
[0134] As shown in FIG. 13, the sensor units 330U and 330V are
constituted respectively by sensors 331U and 331V, biasing
permanent magnets 332U and 332V, and yokes 333U and 333V for
forming magnetic paths of the biasing permanent magnets 332U and
332V. Further, the sensor units 330U and 330V are packaged in a
resin molding (not shown) for ensuring resistance against
environments and convenience in use. The sensor units 330U and 330V
are mounted to the motor housing in an easily detachable (or
attachable) structure using, e.g., screws (not shown). Preferably,
magnetic shields 334U and 334V are disposed to shield magnetic
disturbances generated when currents are supplied to the stator
windings. The magnetic yokes 333U and 333V are disposed to be able
to adjust the amounts of magnetic flux detected by the sensors 331U
and 331V.
[0135] The two sensors 331U and 331V are arranged at an interval of
90.degree. on an assumption- that the pitch of the gear teeth of
the sensor rotor 340 corresponds to one cycle (360.degree.). Sensor
outputs of the sensors 331U and 331V are each outputted as an
analog voltage level in proportion to the amount of magnetic flux
passing through a gap between the sensor rotor 340 and the sensor,
which is changed with rotation of the sensor rotor 340.
[0136] While the number of the gear teeth of the sensor rotor 340
is preferably the integer time the number of the magnet pole pairs,
it can be set to an appropriate value depending on the resolution
of a rotational angle, manufacturability of the sensor rotor 340,
and the output frequency (frequency response) of the sensor. The
shape of each gear tooth is preferably selected to be able to
provide an analog voltage level corresponding to the rotational
angle of the sensor rotor 340. The sensors 331U and 331V are each
generally constituted by a Hall device. In the case of employing
the biasing magnet, however, a magnetic resistance device can also
be used without problems because the amount (magnitude) of the
magnet is detected. Further, the rotation sensor can be of any
suitable type capable of measuring the change of the gap with the
rotation of the sensor rotor 340. Another type of sensor of passing
excitation magnetic flux through the sensor rotor and detecting the
amount of the excitation magnetic flux, e.g., a resolver, is also
applicable. In the latter case, the excitation magnetic flux of
high frequency is generally employed. When the amount of the
magnetic flux is detected in the form of a voltage output, for
example, signal processing using, e.g., a low-pass filter or other
processing such as to make the sampling timing in sync with
excitation frequency is additionally performed.
[0137] From the sensor signals Hu and Hv shown at (a) in FIG. 14,
the position detector 710A extracts signals shown at (b) in FIG.
14. The phase signal VO is obtained from the extracted signals. The
signal corrector 722A obtains the signal information SI2 and SI3
from the phase signal VO.
[0138] In an example shown in FIG. 14, the sensor signals are ideal
signals free of the gain error, the offset error, and the phase
error. If the sensor signals contain the gain error, the offset
error, and the phase error, a Gain2 signal is detected due to the
gain error, and Off21 and Off22 signals are detected due to the
offset error, as shown at (d2) in FIG. 11. Also, Phas4l and Phas42
signals are detected as zero due to the no phase error as shown at
(f) in FIG. 14. Based on those error signals, the signal corrector
722A can detect the respective errors by the methods described
above with reference to FIG. 11.
[0139] When two sensor signals are used like this embodiment, the
timing of detecting the signal information SI2 is set to a point in
time at which the sensor signal level matches with each of the
first and second determination values (VL1, VL2). When cross points
between the sensor signals can be utilized, the timing of each
cross point is also used in a combined manner. As an alternative,
absolute values of the extracted signals may be taken to perform
the detection at the cross points without problems. Further,
smaller one of the absolute values may be extracted. Thus, even in
the case of employing the two sensor signals, it is possible to
correct the gain error, the offset error, and the phase error, and
to perform the position detection at a lower cost with high
accuracy.
[0140] While the above embodiments have been described as employing
two or three sensors, the detection accuracy or reliability can be
increased by detecting two or more sensor signals. For example, if
one of the three sensors mounted at intervals of 120.degree. is
failed, failure diagnosis can be made based on the magnitude of an
error contained in the sensor signal. Stated another way, if the
two sensors are normally operated, this is regarded as being
equivalent to the arrangement that they are mounted with the phase
error shifted at 30.degree. in comparison with the case where only
the two sensors are mounted from the beginning. Thus, by correcting
the phase error of 30.degree. in accordance with the error
detection and the error correction of the present invention, the
position detected value .theta.s can be obtained in the range of
0.degree.-360.degree.. Meanwhile, when a maximum error correction
amount is set to 20.degree. in the stage of design, the driving of
the motor can be continued within the range of a position detection
error being 10.degree. at maximum by setting the error diagnosis so
as to allow a phase error of remaining 10.degree. that exceeds a
limit of the correction. As a result, reliability can be greatly
increased.
[0141] When the sensor signals contain the gain error, the offset
error, and the position error as discussed above in the
embodiments, the position error is superimposed in the range of
about .+-.30.degree. with the reference phase value being a base
from the principle of the position detection. In other words,
whenever the reference phase value is updated for each cycle of
60.degree., a large position detection error appears. Eventually,
the output voltage of the inverter causes a voltage phase shift
(pulsation) at a cycle of 60.degree. and the motor current varies
(namely, the current phase abruptly changes). Particularly, if the
error is superimposed on the waveform of the sensor signal during
the driving of the motor, the motor current may vary, thus causing
noisy sounds. By performing the error detection and the error
correction according to the present invention, since the error of
the sensor signal can be corrected within about several cycles
(theoretically within one cycle at minimum), the motor voltage or
the motor current varies just temporarily after superimposition of
the error and can be quickly stabilized.
[0142] Even when the method of detecting the position from two
signals having a phase difference of 90.degree. between them by
using arc tangent is employed instead of the above-described
position calculation method using the reference phase value, the
error of the sensor signal can be detected and corrected by the
position correction method according to the present invention.
[0143] While the above embodiments have been described as using the
Hall device, the sensor may be of any suitable type capable of
outputting an analog signal without problems. For example, the
sensor may be an MR device or a resolver. In the case of requiring
the signal detection in sync with excitation frequency as in a
resolver, the driving of the motor can be realized with high
accuracy and high efficiency by sampling the sensor signal in sync
with the excitation frequency and performing the correction method
according to the present invention.
[0144] Further, when the motor driving is to be continued even in
the case of a sensor failure, reliability can be improved by
correcting the position detected value through the known
calculation for estimating the pole position while using the
applied voltage (which may be replaced with a command value), the
motor current, and motor parameters (such as winding resistance,
winding inductance, and an induced voltage coefficient).
[0145] According to this third embodiment, as described above,
since the gain error, the offset error, and the phase error can be
corrected, the detection accuracy of the pole position can be
increased. It is hence possible to increase the accuracy in
detecting the pole position of the motor which is used to perform
quick acceleration and deceleration over the range from a zero
speed to a high rotation speed.
[0146] In addition, the temperature and the amount of the magnetic
flux can also be detected using the amounts in changes of the gain
error and the offset error. For example, the amount of change in
the offset error can be determined as a temperature change and
utilized for detecting the temperature. Also, the difference
obtained by subtracting the gain error from a value corresponding
to the temperature change can be determined as a change in the
amount of the magnetic flux and utilized for detecting the amount
of the magnetic flux. Therefore, temperature monitoring can be
performed together with the motor driving while the motor output
torque is corrected. Moreover, since the amount of change in the
phase error can be used to detect the amount of torsion (torque) of
a shaft, e.g., the sensor rotor, the motor driving can also be
performed while the motor output torque is measured.
[0147] The construction of a synchronous motor driving device using
a position detecting device according to a fourth embodiment of the
present invention will be described below with reference to FIG.
15.
[0148] FIG. 15 is a block diagram of the synchronous motor driving
device using the position detecting device according to the fourth
embodiment of the present invention. Note that, in FIG. 15, the
same characters as those in FIG. 1 denote the same components.
[0149] A motor driving device 800B includes a rotor position
settling unit 807 in addition to the construction of the motor
driving device 800 shown in FIG. 1. The rotor position settling
unit 807 outputs a command to the voltage controller 805 so that
the permanent-magnet synchronous motor 300 is driven through a
predetermined electrical angle in a stepping manner under PWM
control and the rotor stop position of the motor 300 is settled.
Also, the rotor position settling unit 807 outputs angle
information, which represents the settled rotor stop position, to
an error calculator 721B.
[0150] The error calculator 721B of a correcting unit 720B in a
position detecting device 700B calculates a position detection
error based on the difference between the position detected value
inputted from the position detector 710 and the angle information
inputted from the rotor position settling unit 807. The rotor
position detection error is calculated by solving the simultaneous
equations given as the above-mentioned formulae (1), (2) and (3).
The correcting unit 720B corrects the rotor position detection
error based on the correction amount CSI obtained by the error
calculator 721B.
[0151] According to this fourth embodiment, as described above,
since the gain error, the offset error, and the phase error can be
corrected, the detection accuracy of the pole position can be
increased. It is hence possible to increase the accuracy in
detecting the pole position of the motor which is used to perform
quick acceleration and deceleration over the range from a zero
speed to a high rotation speed.
[0152] The construction of a hybrid vehicle system to which is
applied the motor driving device shown as any of the embodiments of
the present invention will be described below with reference to
FIG. 16.
[0153] FIG. 16 is a schematic view showing the construction of the
hybrid vehicle system to which is applied the motor driving device
shown as any of the embodiments of the present invention.
[0154] The operation of an engine 1 is started by a starter 9. In
the engine 1, an amount of intake air is controlled by an
electronic control throttle 10 disposed in an intake pipe (not
shown), and fuel is injected from a fuel injector (not shown) in
amount corresponding to the amount of intake air. Also, the
ignition timing is decided in accordance with signals indicating an
air/fuel ratio decided from the amount of intake air and the amount
of fuel, an engine revolution speed, etc., and an air-fuel mixture
is ignited by an ignition device (not shown).
[0155] A gear 17 provided with a meshing gear 31, a gear 39
provided with a meshing gear 32, a hub sleeve 27, and gears 15, 16,
37 and 38 are fitted over an input shaft 12 of a transmission
mechanism 100. The gears 15, 16, 37 and 38 are fixedly fitted over
the input shaft 12, and the gears 17 and 39 are fitted in a
structure not allowing those gears to move in the axial direction
of the input shaft 12. The hub sleeve 27 is coupled to the input
shaft 12 through a meshing mechanism (not shown) which allows the
hub sleeve to move in the axial direction of the input shaft 12,
but restricts it in the rotating direction thereof.
[0156] A gear 20 provided with a meshing gear 29, a gear 21
provided with a meshing gear 30, a hub sleeve 26, a gear 24
provided with a meshing gear 33, a gear 25 provided with a meshing
gear 34, a hub sleeve 28, and gears 22 and 23 are fitted over an
output shaft 13 of the transmission mechanism 100. The gears 22 and
23 are fixedly fitted over the output shaft 13, and the gears 20,
21, 24 and 25 are fitted in a structure not allowing those gears to
move in the axial direction of the output shaft 13. The hub sleeves
26 and 28 are coupled to the output shaft 13 through a meshing
mechanism (not shown) which allows the hub sleeves to move in the
axial direction of the output shaft 13, but restricts them in the
rotating direction thereof.
[0157] The gears 15 and 20, the gears 16 and 21, the gears 17 and
22, the gears 39 and 23, and the gears 37 and 24 are meshed with
each other to constitute gear combinations having different gear
ratios when torque is transmitted from the input shaft 12 to the
output shaft 13. Further, the gear 25 can be coupled to the gear 38
through a reverse gear 35 and serves to selectively reverse the
relationship in rotating direction between the input shaft 12 and
the output shaft 13.
[0158] A gear train made up of the gears 15 and 20 corresponds to a
first speed, a gear train made up of the gears 16 and 21
corresponds to a second speed, and a gear train made up of the
gears 17 and 22 corresponds to a third speed. A gear train made up
of the gears 39 and 23 corresponds to a fourth speed, a gear train
made up of the gears 37 and 24 corresponds to a fifth speed, and a
gear train made up of the gears 25, 35 and 38 corresponds to
reverse.
[0159] When the torque of the input shaft 12 is transmitted to both
the gears 17 and 39, the meshing gears 31 and 32 are directly
coupled to the hub sleeve 27 through meshing grooves (not
shown).
[0160] A clutch mechanism constituted by the hub sleeve 27, the
meshing gear 31, and the meshing gear 32 is called a meshing (dog)
clutch which can transmit the torque of the input shaft 12 to the
output shaft 13 with high efficiency, thus resulting in reduced
fuel consumption.
[0161] Further, the hub sleeve 26, the meshing gear 29, and the
meshing gear 30 constitute a dog clutch, while the hub sleeve 28,
the meshing gear 33, and the meshing gear 34 constitute a dog
clutch. These dog clutches operate similarly to the above-mentioned
dog clutch so that the torque is transmitted to the output shaft 13
through the gears 20 and 21 and the gears 24 and 25,
respectively.
[0162] A clutch 4 is interposed between a crankshaft 11 of the
engine 1 and the input shaft 12. When the clutch 4 is engaged,
motive power can be transmitted from the engine 1 to the input
shaft 12. When the clutch 4 is disengaged, the transmission of
motive power from the engine 1 to the input shaft 12 can be cut
off. Generally, the clutch 4 is constituted by a dry and
single-plate friction clutch, and the torque transmitted from the
engine 1 to the input shaft 12 can be adjusted by controlling a
pressing force applied to the clutch 4. As an alternative, the
clutch 4 may be constituted by a wet and multi-plate friction
clutch or an electromagnetic clutch so long as it is capable of
adjusting the transmitted torque. Stated another way, the clutch 4
is also used in an ordinary gasoline engine vehicle, and the
vehicle can be started by gradually pressing the clutch 4.
[0163] A final gear 14 is fitted over the output shaft 13 of the
transmission mechanism 100, and the final gear 14 is coupled to a
tire 36 through a vehicle driving shaft 2.
[0164] A gear 19 is fixed to a motor output shaft 18, and the gear
19 is meshed with the gear 37. Torque of a motor 300 can be thus
transmitted to the input shaft 12.
[0165] The engine 1 and the motor 300 are controlled by an engine
C/U (control unit; this is equally applied to the following
description) 6 and a motor driving device 800, respectively. The
clutch 4 and the transmission mechanism 100 are controlled by a
transmission C/U 7. A power train C/U 5 receives not only various
signals from an accelerator-pedal opening sensor, a vehicle speed
sensor, etc. (not shown), but also operating states (such as the
rotation speed, the torque, and the gear ratio) of the engine 1,
the motor 300, the clutch 4, and the transmission mechanism 100.
Based on those input signals and states, the power train C/U 5
controls the engine C/U 6, the motor driving device 800, and the
transmission C/U 7 in a supervising manner.
[0166] At startup of the vehicle with a driving force of the motor
300, the vehicle can be smoothly started by meshing the hub sleeve
26 with the meshing gear 29, disengaging the clutch 4, and
generating the torque of the motor 300 in the positive side
(forward direction). Thereafter, the clutch 4 is engaged for
switchover to a driving force of the engine such that the vehicle
continues to travel by the engine driving force. Thus, since the
motor driving is utilized in a low load range (e.g., at the
startup) where engine efficiency is poor, fuel economy can be
improved.
[0167] Further, sensor units 330U and 330V are mounted to a
transmission case in such a manner that they can be replaced from
the outside of the transmission case. Therefore, even when sensors
are replaced in the manufacturing stage of a product or by a
dealer, errors of the sensors themselves and mounting errors can be
corrected on the running basis in the manner described above, and
workability can be increased to a large extent.
[0168] The construction of an electric power steering system to
which is applied the motor driving device shown as any of the
embodiments of the present invention will be described below with
reference to FIG. 17.
[0169] FIG. 17 is a schematic view showing the construction of the
electric power steering system to which is applied the motor
driving device shown as any of the embodiments of the present
invention.
[0170] An electric actuator comprises a torque transmission
mechanism 902, the motor 300, and the motor driving device 800. The
electric power steering system mainly comprises the electric
actuator, a steering wheel 900, a steering detector 901, and a
control input command unit 903. When an operating force is applied
from a driver to the steering wheel, the electric power steering
system assists the steering torque with the aid of the electric
actuator.
[0171] A torque command .tau.* for the electric actuator is issued,
as a steering assist torque command (produced by the control input
command unit 903), with the steering operation in order to reduce
the steering force to be applied from the driver by utilizing an
output of the electric actuator. The motor driving device 800
receives, as an input command, the torque command .tau.* and
controls a motor current based on the torque constant of the
driving motor and the torque command .tau.* so as to follow a
torque command value.
[0172] A motor output .tau.m outputted from the output shaft 360
(see FIG. 2) directly coupled to the rotor of the motor 300
transmits torque to a rack 910 of the electric power steering
system through a speed reducing mechanism such as a worm, a wheel
and a planetary gear, or the torque transmission mechanism 902
using a hydraulic mechanism, thereby steering wheels 920 and 921
while relieving (assisting) the steering force (operating force)
applied from the driver to the steering wheel 900 with the aid of
an electrical power force. An amount of the assisted electrical
power force is decided as the torque command .tau.* by the control
input command input 903 through the steps of detecting operation
variables, such as a steering angle and steering torque, by the
steering detector 901 which is built in a steering shaft and
detects the steering state, and taking into account status
variables, such as the vehicle speed and the road surface
state.
[0173] Since the motor driving device of the present invention is
able to correct a phase difference caused in the rotation sensor
and to drive the motor with high efficiency even in response to the
torque command .tau.* required for the electric actuator which is
subjected to quick acceleration and deceleration, the motor can be
driven without reducing motor efficiency even in a field weakening
range where the electric actuator is operated at high speed and
high torque. Further, at low speed, it is also possible to reduce
the sensor mounting error and to stably drive the motor with a
small torque variation. Accordingly, the electric power steering
system employing the motor driving device of the present invention
can realize a high-torque and a fast-response electric power
steering system without impairing a satisfactory feel of the driver
in the steering operation.
* * * * *