FORM 2
THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENTS RULES, 2003
COMPLETE SPECIFICATION
[See section 10, Rule 13]
MOTOR DRIVE DEVICE, COMPRESSOR DRIVE SYSTEM, AND
REFRIGERATION CYCLE SYSTEM;
MITSUBISHI ELECTRIC CORPORATION, A CORPORATION ORGANISED
AND EXISTING UNDER THE LAWS OF JAPAN, WHOSE ADDRESS IS 7-3,
MARUNOUCHI 2-CHOME, CHIYODA-KU, TOKYO 100-8310, JAPAN
THE FOLLOWING SPECIFICATION PARTICULARLY DESCRIBES THE
INVENTION AND THE MANNER IN WHICH IT IS TO BE PERFORMED.
2
DESCRIPTION
MOTOR DRIVE DEVICE, COMPRESSOR DRIVE SYSTEM, AND
REFRIGERATION CYCLE SYSTEM
5
Field
[0001] The present invention relates to a motor drive
device that drives an alternating-current motor, a
compressor drive system, and a refrigeration cycle system.
10
Background
[0002] Currently, alternating-current motors are used as
power sources for various mechanical devices. Many of the
mechanical devices experience periodic variation in load
15 torque, that is, periodic load torque pulsation. In the
alternating-current motor, the mechanical device, or the
like, the load torque pulsation may cause vibration, noise,
or the like. Therefore, various techniques related to
vibration suppression control have been studied.
20 [0003] If information on the load torque pulsation is
known, it is not difficult to suppress the vibration by
feedforward, but normally, the amplitude, the phase, and
the like of the load torque pulsation change depending on
operating conditions of the mechanical device. It is
25 cumbersome to collect the information on the load torque
pulsation in advance, and the feedforward control does not
exert an effect when a pulsation different from the data
collected in advance occurs. Therefore, a method of
detecting and feeding back characteristic information such
30 as vibration and speed pulsation has been studied. In the
method based on feedback, a controller automatically
determines a control command value that can reduce the
vibration, speed pulsation, and the like.
3
[0004] Many vibration suppression control techniques are
based on an operation in a region where a power converter
for supplying power to the alternating-current motor
operates linearly. Patent Literature 1 discloses a
5 technique for continuously switching control such that the
vibration suppression control is performed in a linear
region of an inverter, and flux weakening control by
voltage phase control is performed in an overmodulation
region.
10
Citation List
Patent Literature
[0005] Patent Literature 1: Japanese Patent Application
Laid-open No. 2017-55466
15
Summary
Technical Problem
[0006] However, according to the above conventional
technique, when the vibration suppression control is to be
20 performed at the time of high-speed rotation, it is
necessary to make a large margin in a modulation rate by
passing a large amount of d-axis current on average. This
has caused a problem in that the efficiency of a drive
device of the alternating-current motor is impaired.
25 [0007] The present invention has been made in view of
the above, and an object of the present invention is to
provide a motor drive device capable of reducing or
preventing a reduction in efficiency while performing
vibration suppression control in an overmodulation region.
30
Solution to Problem
[0008] To solve the above problem and achieve an object,
the present invention is directed to a motor drive device
4
that controls driving of an alternating-current motor
connected to a mechanical device with periodic load torque
pulsation by using dq rotating coordinates. The motor
drive device includes: a power conversion unit to convert a
5 direct current voltage into an alternating current voltage
on the basis of a voltage command, and output the
alternating current voltage to the alternating-current
motor; a current detection unit to detect a phase current
flowing to the alternating-current motor; a position/speed
10 specifying unit to specify a magnetic pole position and a
rotational speed of the alternating-current motor; a d-axis
current pulsation generating unit to generate a d-axis
current pulsation command on the basis of periodic q-axis
current pulsation or a periodic q-axis current pulsation
15 command, the d-axis current pulsation command being in
synchronization with the q-axis current pulsation or the qaxis current pulsation command and preventing or reducing
an increase or decrease in amplitude of the voltage
command; and a dq-axis current control unit to generate the
20 voltage command that controls the phase current on the dq
rotating coordinates, which rotate in synchronization with
the magnetic pole position, by using the magnetic pole
position, the rotational speed, the phase current, the qaxis current pulsation or the q-axis current pulsation
25 command, and the d-axis current pulsation command.
Advantageous Effects of Invention
[0009] According to the present invention, the motor
drive device can reduce or prevent a reduction in
30 efficiency while performing the vibration suppression
control in the overmodulation region.
Brief Description of Drawings
5
[0010] FIG. 1 is a block diagram illustrating an example
of a configuration of a motor drive device according to a
first embodiment.
FIG. 2 is a diagram illustrating a voltage vector
5 representing a state of voltage applied to an alternatingcurrent motor when the alternating-current motor rotates in
a high-speed region, the alternating-current motor being an
interior permanent magnet synchronous motor to be
controlled by the motor drive device according to the first
10 embodiment.
FIG. 3 is a first diagram illustrating a voltage
command vector for comparison when vibration suppression
control according to the first embodiment is not performed.
FIG. 4 is a second diagram illustrating a voltage
15 command vector for comparison when the vibration
suppression control according to the first embodiment is
not performed.
FIG. 5 is a third diagram illustrating a voltage
command vector for comparison when the vibration
20 suppression control according to the first embodiment is
not performed.
FIG. 6 is a diagram illustrating a voltage command
vector when the vibration suppression control according to
the first embodiment is performed.
25 FIG. 7 is a first diagram illustrating a simple method
of calculating d-axis current pulsation in a d-axis current
pulsation generating unit according to the first embodiment.
FIG. 8 is a second diagram illustrating a simple
method of calculating the d-axis current pulsation in the
30 d-axis current pulsation generating unit according to the
first embodiment.
FIG. 9 is a diagram illustrating an example of
operation waveforms as a comparative example when the
6
vibration suppression control according to the first
embodiment is not performed.
FIG. 10 is a diagram illustrating an example of
operation waveforms when the vibration suppression control
5 according to the first embodiment is performed.
FIG. 11 is a flowchart illustrating an operation of
the motor drive device according to the first embodiment.
FIG. 12 is a diagram illustrating an example of a
hardware configuration of the motor drive device according
10 to a second embodiment.
FIG. 13 is a block diagram illustrating an example of
a configuration of a motor drive device according to a
third embodiment.
FIG. 14 is a flowchart illustrating an operation of
15 the motor drive device according to the third embodiment.
FIG. 15 is a block diagram illustrating an example of
a configuration of a motor drive device according to a
fourth embodiment.
FIG. 16 is a diagram illustrating an example of a
20 voltage vector locus when voltage is not saturated in the
motor drive device according to the fourth embodiment.
FIG. 17 is a flowchart illustrating an operation of
the motor drive device according to the fourth embodiment.
FIG. 18 is a block diagram illustrating an example of
25 a configuration of a motor drive device according to a
fifth embodiment.
FIG. 19 is a first diagram for explaining the
necessity of offset correction performed by the motor drive
device according to the fifth embodiment.
30 FIG. 20 is a second diagram for explaining the
necessity of offset correction performed by the motor drive
device according to the fifth embodiment.
FIG. 21 is a diagram illustrating an example of a
7
method of calculating an offset correction amount in the
motor drive device according to the fifth embodiment.
FIG. 22 is a flowchart illustrating an operation of
the motor drive device according to the fifth embodiment.
5 FIG. 23 is a diagram illustrating an example of a
configuration of a compressor drive system according to a
sixth embodiment.
FIG. 24 is a schematic diagram illustrating an example
of a detailed configuration of a compression chamber
10 included in the compressor drive system according to the
sixth embodiment.
FIG. 25 is a graph illustrating an example of a
waveform of load torque of a refrigerant compressor
according to the sixth embodiment.
15 FIG. 26 is a diagram illustrating an example of a
configuration of a refrigeration cycle system according to
a seventh embodiment.
Description of Embodiments
20 [0011] A motor drive device, a compressor drive system,
and a refrigeration cycle system according to embodiments
of the present invention will now be described in detail
with reference to the drawings. Note that the present
invention is not limited to the embodiments.
25 [0012] First Embodiment.
FIG. 1 is a block diagram illustrating an example of a
configuration of a motor drive device 105 according to a
first embodiment of the present invention. The motor drive
device 105 is connected to an alternating-current motor 1.
30 The alternating-current motor 1 is mechanically connected
to a mechanical device 2. The alternating-current motor 1
is a power source of the mechanical device 2. The
mechanical device 2 operates when the motor drive device
8
105 outputs an alternating current voltage to the
alternating-current motor 1.
[0013] In the present embodiment, the alternatingcurrent motor 1 is assumed to be an interior permanent
5 magnet synchronous motor, but may be a surface permanent
magnet synchronous motor, a wound field synchronous motor,
an induction motor, a reluctance synchronous motor, or the
like. For convenience of description, the alternatingcurrent motor 1 will be described as a three-phase motor,
10 but may be a motor having another number of phases such as
a two-phase motor or a five-phase motor.
[0014] In the present embodiment, the mechanical device
2 has periodic load torque pulsation. Various mechanical
devices 2 having such a characteristic are conceivable,
15 where a compressor is well known as a typical example
thereof. The compressor is a device that compresses a
substance such as air or a refrigerant and discharges the
compressed substance. A large torque is required for the
motor in the step of compressing the substance, and the
20 torque required for the motor decreases in the step of
discharging the compressed substance. The compressor is
thus known to have the periodic load torque pulsation.
[0015] In the compressor, the period of load torque
pulsation is determined by a structure of a compression
25 mechanism. For example, in a compressor of a type called a
single rotary compressor, a compression step and a
discharge step are each performed once while a compression
mechanism including one compression chamber makes one
rotation. Therefore, the angular frequency of the load
30 torque pulsation of the single rotary compressor is equal
to the rotational angular velocity of the compression
mechanism. In a compressor of a type called a twin rotary
compressor, a compression step and a discharge step are
9
each performed twice while a compression mechanism
including two compression chambers makes one rotation.
Therefore, the angular frequency of the load torque
pulsation of the twin rotary compressor is twice the
5 rotational angular velocity of the compression mechanism.
There are various other compressors such as a reciprocating
compressor, a scroll compressor, and a screw compressor.
The angular frequency of the load torque pulsation is often
N times the rotational angular velocity of the compression
10 mechanism. Note that “N” is a positive real number.
[0016] Note that although the compressor has been
described as the typical example of the mechanical device 2,
the mechanical device 2 is not limited thereto. The
control of the present embodiment can be applied to any
15 mechanical device as long as the load torque pulsation
periodically occurs in the mechanical device. Moreover, in
a case where the alternating-current motor 1 and the
mechanical device 2 are structurally integrated, the torque
ripple of the alternating-current motor 1 itself can also
20 be considered as the load torque pulsation, and thus the
control of the present embodiment can be applied to the
torque ripple of the alternating-current motor 1.
[0017] The motor drive device 105 controls driving of
the alternating-current motor 1 connected to the mechanical
25 device 2 by using dq rotating coordinates. The dq rotating
coordinates are used in a general vector control method in
controlling a motor or the like. A configuration of the
motor drive device 105 will be described. As illustrated
in FIG. 1, the motor drive device 105 includes a power
30 conversion unit 3, a current detection unit 4, a
position/speed specifying unit 5, a d-axis current
pulsation generating unit 6, and a dq-axis current control
unit 7.
10
[0018] The power conversion unit 3 converts power input
from a power source (not illustrated) into power of a
prescribed form, and outputs the power. The power
conversion unit 3 may have any configuration as long as it
5 can drive the alternating-current motor 1. Here, the power
conversion unit 3 will be described as a general-purpose
voltage source inverter. The voltage source inverter is a
device that switches and converts a direct current voltage
supplied from a direct current voltage source into a
10 desired alternating current voltage. Specifically, the
power conversion unit 3 converts a direct current voltage
into an alternating current voltage on the basis of a
voltage command output from the dq-axis current control
unit 7. The power conversion unit 3 outputs the
15 alternating current voltage obtained after conversion to
the alternating-current motor 1. Note that the power
conversion unit 3 may be another type of circuit such as a
current source inverter or a matrix converter, or may be a
multi-level converter as long as desired alternating
20 current power can be supplied to the alternating-current
motor 1.
[0019] The current detection unit 4 detects a phase
current flowing to the alternating-current motor 1. The
type, arrangement, and the like of the current detection
25 unit 4 are not particularly limited. The current detection
unit 4 may be a current sensor of a type using a
transformer called a current transformer (CT), may be a
current sensor of a type using a shunt resistor, or may use
a combination of the CT and the shunt resistor. In FIG. 1,
30 the current detection unit 4 is disposed on a wiring
between the alternating-current motor 1 and the power
conversion unit 3 to measure the phase current flowing to
the alternating-current motor 1, but may be disposed at
11
another location. For example, the current detection unit
4 may be disposed inside the power conversion unit 3.
[0020] When the current detection unit 4 is disposed
inside the power conversion unit 3, a one-shunt current
5 detection method in which a shunt resistor is disposed on
an N side of a direct-current bus of the inverter, a lowerarm shunt current detection method in which a shunt
resistor is inserted in series with a lower arm of the
inverter, or the like can be used. In the one-shunt
10 current detection method and the lower-arm shunt current
detection method, the timing at which the current can be
detected is limited as compared to the case of using the CT,
but the component cost can be reduced.
[0021] Moreover, when the power conversion unit 3 is a
15 three-phase motor in the motor drive device 105, if current
sensors are disposed in any two phases, the current of a
third phase can be calculated according to Kirchhoff's
current law so that the current sensors need not be
disposed in all the three phases.
20 [0022] In order to perform vector control on the
alternating-current motor 1, a magnetic pole position and
rotational speed of the alternating-current motor 1 needs
to be detected or estimated. The position/speed specifying
unit 5 specifies the magnetic pole position and the
25 rotational speed of the alternating-current motor 1.
Specifically, the position/speed specifying unit 5
estimates a magnetic pole position of a rotor (not
illustrated) included in the alternating-current motor 1
and rotational speed of the alternating-current motor 1 on
30 the basis of the voltage command output from the dq-axis
current control unit 7 and the phase current of the
alternating-current motor 1 detected by the current
detection unit 4. The magnetic pole position of the rotor
12
is also referred to as a rotor position.
[0023] Note that in the motor drive device 105, a
position sensor 5a may be used as the position/speed
specifying unit 5 if the position sensor 5a can be attached
5 to the alternating-current motor 1. The position sensor 5a
may be a rotary encoder or a resolver. Alternatively,
instead of the position sensor 5a, a speed sensor called a
tachogenerator may be used. However, the position sensor
5a, the speed sensor, or the like may not be usable
10 depending on restrictions such as use environment and cost.
Therefore, the present embodiment will be described on the
assumption that motor drive device 105 performs position
sensorless control. This, however, is not intended to
limit the invention, and it is apparent that the motor
15 drive device of the present application may be configured
using the position sensor 5a or the speed sensor.
[0024] Various methods have been proposed for the
position sensorless control of the alternating-current
motor 1, but basically any method may be used. For example,
20 a speed estimation method is known in which a state
quantity of the alternating-current motor 1 is estimated by
a state observing device, and rotational speed is
adaptively identified using an estimation error of the
state quantity. This method is a method called an adaptive
25 observer, and has an advantage that speed estimation robust
against a change in an induced voltage constant can be
performed. When the adaptive observer is not used, the
magnetic pole position may be estimated simply from the
arctangent of a speed electromotive force. This method is
30 called an arctangent method. The arctangent method has a
disadvantage that an error occurs in speed estimation when
the induced voltage constant has an error, but the
calculation is simpler than the adaptive observer. Many
13
other position sensorless control methods have been
proposed, but any method may be used as long as the
rotational speed and the magnetic pole position of the
alternating-current motor 1 can be estimated.
5 [0025] The d-axis current pulsation generating unit 6
determines, that is, generates a d-axis current pulsation
command idAC
* from a q-axis current pulsation iqAC. The
details will be described later because the d-axis current
pulsation generating unit 6 is the most important point for
10 performing a characteristic operation in the motor drive
device 105. Here, the q-axis current pulsation iqAC may be
a command value or a current value actually flowing in the
alternating-current motor 1. That is, on the basis of the
periodic q-axis current pulsation iqAC or a periodic q-axis
current pulsation command iqAC
* 15 described later in a third
embodiment, the d-axis current pulsation generating unit 6
generates the d-axis current pulsation command idAC
*,which
is in synchronization with the q-axis current pulsation iqAC
or the q-axis current pulsation command iqAC
* and prevents
20 or reduces the increase or decrease in the amplitude of the
voltage command. The first embodiment assumes that the qaxis current pulsation iqAC is given from a speed pulsation
suppression control unit or a vibration suppression control
unit (not illustrated). In order to reduce the vibration
25 of the mechanical device 2, it is necessary to cause the
motor torque of the alternating-current motor 1 to pulsate
in synchronization with the periodic load torque pulsation,
and thus it is not uncommon to have q-axis current
pulsation in this type of motor drive device 105. Note
30 that the configuration of the speed pulsation suppression
control unit, the vibration suppression control unit, or
the like is not particularly limited. For example, a
control unit that performs feedback control as described in
14
Japanese Patent Application Laid-open No. H01-308184 may be
used, or a control unit that performs feedforward
compensation by checking the amplitude, phase, and the like
of the load torque pulsation in advance may be used.
5 Alternatively, the control unit may be a control unit that
observes speed pulsation of the alternating-current motor 1
and performs control to cancel the speed pulsation, or may
be a control unit that performs control to cancel
acceleration pulsation observed by an acceleration sensor
10 (not illustrated) that is attached to the mechanical device
2. Yet alternatively, the control unit may be a control
unit that performs control to reduce pulsation of repeated
stress applied to the mechanical device 2 using a force
sensor such as a strain gauge (not illustrated).
15 [0026] The dq-axis current control unit 7 controls the
phase current flowing to the alternating-current motor 1.
As the dq-axis current control unit 7, it is preferable to
use a vector control unit on the dq rotating coordinates.
A typical vector control unit performs current control on
20 the dq rotating coordinates with respect to the magnetic
pole of the rotor. This is because when the phase current
is converted into a value on the dq rotating coordinates,
the alternating current value becomes a direct current
value and the control becomes easier. Because magnetic
25 pole position information is required for the coordinate
transformation, the position/speed specifying unit 5
estimates the magnetic pole position. The dq-axis current
control unit 7 calculates a dq-axis current command using
at least two pieces of information being the q-axis current
pulsation iqAC and the d-axis current pulsation command idAC
* 30 .
In addition to these two pieces of information, the dq-axis
current control unit 7 may use information given from
another control unit such as a speed control unit or a flux
15
weakening control unit (not illustrated) to determine the
dq-axis current command. The dq-axis current control unit
7 performs control such that the dq-axis current command
including the d-axis current pulsation command idAC
* matches
5 a dq-axis current, and determines a voltage command for the
power conversion unit 3. Specifically, the dq-axis current
control unit 7 uses the magnetic pole position and the
rotational speed specified by the position/speed specifying
unit 5, the phase current detected by the current detection
10 unit 4, the q-axis current pulsation iqAC or the q-axis
current pulsation command iqAC*, and the d-axis current
pulsation command idAC
* generated by the d-axis current
pulsation generating unit 6, to generate the voltage
command for controlling the phase current on the dq
15 rotating coordinates rotating in synchronization with the
magnetic pole position. As a current control method in the
dq rotating coordinates, it is common to employ a method of
disposing a proportional integral (PI) control unit on each
of the d-axis and the q-axis and using in combination a
20 decoupling control unit that compensates an interference
component of the dq axis by feedforward. However, any
other method may be used as long as the dq-axis current
properly follows the dq-axis current command. The dq-axis
current control unit 7 determines the voltage command on
25 the dq rotating coordinates, performs coordinate
transformation to convert the voltage command on the dq
rotating coordinates into a value of a three-phase
stationary coordinate by using the magnetic pole position
information, and outputs the value to the power conversion
30 unit 3.
[0027] Next, the necessity and operation of the d-axis
current pulsation generating unit 6 will be described. FIG.
2 is a diagram illustrating a voltage vector representing a
16
state of voltage applied to the alternating-current motor 1
when the alternating-current motor 1 rotates in a highspeed region. The alternating-current motor 1 is the
interior permanent magnet synchronous motor to be
5 controlled by the motor drive device 105 according to the
first embodiment. In the high-speed region, a voltage drop
due to coil resistance of the alternating-current motor 1
is often negligible, so that the voltage drop due to the
coil resistance is omitted in FIG. 2. Also, FIG. 2 is the
10 diagram illustrating the voltage vector in a steady state
and omits a transient term. In the alternating-current
motor 1, as electric angular velocity ωe increases, a speed
electromotive force ωeΦa increases. Here, “Φa” represents a
dq-axis flux linkage and is a value unique to the motor.
15 The speed electromotive force ωeΦa is generated in the
direction of the q-axis. In the permanent magnet
synchronous motor, the q-axis current and the magnet torque
of the motor are proportional to each other. Normally, the
mechanical device 2 performs some mechanical work, and thus
20 the alternating-current motor 1 needs to output some torque.
A q-axis current iq flows through the alternating-current
motor 1, and a voltage ωeLqiq is generated in the direction
of the d-axis by armature reaction of the q-axis current iq.
Here, “Lq” represents a q-axis inductance. On the other
25 hand, a d-axis current id contributes to a small extent to
the torque, and thus is controlled to a smaller value in a
low and middle speed region, in which the rotational speed
of the alternating-current motor 1 is lower than that in
the high-speed region, than in the high-speed region. As a
30 method of determining a d-axis current command in the low
and middle speed region, id=0 control, maximum torque per
ampere control (MTPA), and the like are known. On the
other hand, in the high-speed region, a vector sum of the
17
speed electromotive force ωeΦa and the voltage ωeLqiq may
exceed the maximum output voltage of the power conversion
unit 3, and a method called flux weakening control needs to
be used.
5 [0028] In general, the maximum voltage of the
alternating current voltage that can be output from the
power conversion unit 3 to the alternating-current motor 1
is limited, so that when the limit value of the dq-axis
voltage is “Vom”, a relationship of an approximate
10 expression of expression (1) is established with respect to
the limit value Vom in the high-speed region. Note that
strictly speaking, the output limit range of the power
conversion unit 3 has a hexagonal shape, but is
approximated by a circle here. Although the discussion in
15 the present embodiment assumes the approximation with a
circle, it is needless to say that the discussion may be
made by strictly assuming a hexagon.
[0029]
[Expression 1]
20 ... (1)
[0030] In the present embodiment, a circle whose radius
centered on the origin is the limit value Vom, is referred
to as a voltage limit circle 21. Note that the limit value
Vom is known to vary depending on the value of a direct
25 current bus voltage in a case where the power conversion
unit 3 is a pulse width modulation (PWM) inverter.
[0031] Because the speed electromotive force ωeΦa is
very large in the high-speed region, in order to increase
the q-axis current iq, it is necessary to pass the d-axis
30 current id in a negative direction and to keep the
amplitude of a voltage command vector ν* within the range
   
2
e
2 om
q q
2
a d d
V
L i L i 








   
18
of the voltage limit circle 21. The method of control for
reducing the voltage amplitude by generating a d-axis
stator flux Ldid in the direction opposite to the dq-axis
flux linkage Φa as just described, is generally called flux
5 weakening control. Here, “Ld” represents a d-axis
inductance. The voltage phase control described in Patent
Literature 1 is also one kind of flux weakening control.
[0032] The simplest method of flux weakening control is
a method of determining a d-axis current command on the
10 basis of a voltage equation. Expression (2) is obtained by
solving expression (1) for the d-axis current id.
[0033]
[Expression 2]
... (2)
15 [0034] However, the flux weakening control for obtaining
the d-axis current id expressed by expression (2) has a
disadvantage that it is sensitive to a change, variation,
or the like of the motor constant, and is not used much in
the industry.
20 [0035] Integral flux weakening control is used instead
of the flux weakening control for obtaining the d-axis
current id expressed by expression (2). For example, a
known method determines the d-axis current command id
* by
performing integral control on a difference between the
amplitude of the voltage command vector |ν* 25 | and the limit
value Vom. In this method, when the amplitude of the
voltage command vector |ν*| is larger than the limit value
Vom, the d-axis current command id
* is increased in the
negative direction, or conversely, when the amplitude of
the voltage command vector |ν* 30 | is smaller than the limit
 
d
2
q q
2
e
om
a
d L
L i
V
i
 








  

19
value Vom, the d-axis current command id
* is decreased. In
general, a limiter is appropriately inserted into the daxis current command id
*. This is to prevent the d-axis
current command id
* from becoming excessive and the
5 alternating-current motor 1 from being demagnetized.
Moreover, a limiter in a positive direction may be inserted
in order to prevent the passage of the positive d-axis
current id in the low and middle speed region of the
rotational speed of the alternating-current motor 1. The
10 limit value in the positive direction is usually equal to
zero or a “maximum torque/current command value of current
control”.
[0036] As in Patent Literature 1, there is also known a
method of equivalently performing flux weakening control by
15 adding an output of integral control to a control phase and
advancing a voltage phase.
[0037] However, the integral flux weakening control is
not suitable when the q-axis current iq changes at a high
frequency, although it is robust against fluctuations in
20 the constant of the alternating-current motor 1. This is
because a control response of the integral control cannot
be increased. A current control response in general
position sensorless control is about 3000 to 4000 rad/s.
Because the flux weakening control is configured as an
25 outer loop of current control, a flux weakening control
response can only be designed to be about one-tenth of the
current control response considering control stability.
Therefore, the flux weakening control response is limited
to 300 to 400 rad/s. On the other hand, the performance
30 required for the speed pulsation suppression control and
the vibration suppression control is increasing year by
year, and it is required to suppress the vibration of 1000
to 2000 rad/s in terms of a disturbance angular frequency.
20
That is, because the upper limit value of the flux
weakening control response is too low for the required
specifications, it can be said that the high-frequency
vibration suppression control cannot be achieved by the
5 conventional integral flux weakening control. Even in the
case where the motor drive device is configured using the
position sensor 5a, a similar problem can occur in control
response design although there is a difference in number.
For example, when vibration having a disturbance frequency
10 of 2000 rad/s in the high-speed region is to be suppressed
by control with a position sensor, it is necessary to set
the current control response to a high gain of 20,000 rad/s
or more and the flux weakening control response to a high
gain of 2000 rad/s or more. Considering the balance
15 between the device cost and the control performance, there
are many cases where the control response cannot be set
this high even in the control with the position sensor.
Therefore, regardless of the presence or absence of the
position sensor 5a, it can be said that the high-frequency
20 vibration suppression control cannot be achieved by the
conventional integral flux weakening control.
[0038] In addition to the fact that it is said to be
difficult to achieve highly responsive torque control in
the high-speed region where voltage saturation occurs with
25 the alternating current voltage output from the power
conversion unit 3 to the alternating-current motor 1, when
the flux weakening control has the problem of control
response, it has to be said that it is difficult to perform
the vibration suppression control in the high-speed region.
30 For example, Patent Literature 1 discusses a method of
smoothly stopping the vibration suppression control without
performing the vibration suppression control in the highspeed region.
21
[0039] Note that as mentioned in Patent Literature 1,
when motor efficiency is ignored, the vibration suppression
control should be possible even in the high-speed region by
constantly passing the d-axis current id more than
5 necessary. However, this is not realistic. For example,
an air-conditioning compressor is subject to strict energy
saving regulations, so that it is not possible to
constantly pass the d-axis current id more than necessary.
From a viewpoint other than energy saving as well, it is
10 easy to imagine that an increase in copper loss due to the
excessive d-axis current id may cause various problems such
as adversely affecting the heat dissipation design of the
mechanical device 2.
[0040] On the other hand, the present embodiment
15 achieves the vibration suppression control in the highspeed region and high-frequency region that has been
extremely difficult in the related art. Specifically, the
vibration suppression control in the high-speed region and
high-frequency region is achieved by decomposing the q-axis
20 current into “a low frequency component including a direct
current component” and “a high frequency component
including a disturbance angular frequency”, and performing
the flux weakening control to a necessary small extent as
far as possible on the high-frequency q-axis current
25 pulsation iqAC.
[0041] Effects of the present embodiment will be
specifically described with reference to the drawings. FIG.
3 is a first diagram illustrating a voltage command vector
for comparison when the vibration suppression control
30 according to the first embodiment is not performed. FIG. 4
is a second diagram illustrating a voltage command vector
for comparison when the vibration suppression control
according to the first embodiment is not performed. FIG. 5
22
is a third diagram illustrating a voltage command vector
for comparison when the vibration suppression control
according to the first embodiment is not performed. FIG. 6
is a diagram illustrating a voltage command vector when the
5 vibration suppression control according to the first
embodiment is performed. Here, for convenience of
description, it is assumed that an integral flux weakening
control unit and a feedback vibration suppression control
unit are used. In FIG. 3, in order to perform the
10 vibration suppression control, it is assumed that highfrequency q-axis current pulsation iqAC1 is to be generated.
However, in order not to change the average speed, it is
assumed that a low-frequency q-axis current iqDC1 including
a direct current component needs to be passed constantly.
15 At this time, if there is no large change in the d-axis
current id, the locus of a tip of the voltage command
vector ν* moves in a direction parallel to the d axis.
There is no problem as long as the voltage command vector
ν
* is always within the range of the voltage limit circle
20 21, but in the high-speed region or high-load region, the
voltage command vector ν* goes beyond the range of the
voltage limit circle 21 at the moment when a sum of the qaxis currents iqDC1+iqAC1 increases to be large. A desired qaxis current cannot be passed when the voltage is saturated.
25 When the q-axis current pulsation iqAC1 cannot be passed,
the feedback vibration suppression control unit tries to
pass q-axis current pulsation iqAC2 having a larger
amplitude as illustrated in FIG. 4. Because the q-axis
current pulsation iqAC2 is larger than the q-axis current
30 pulsation iqAC1, a larger voltage is required to pass the qaxis current pulsation iqAC2. Here, in order to reduce the
amplitude of the voltage command vector |ν*|, the integral
flux weakening control unit increases the d-axis current id
23
in the negative direction. As a result of such an
operation, q-axis current pulsation iqAC3 passes eventually.
Even if the q-axis current pulsation iqAC1 is a sinusoidal
signal, the q-axis current pulsation iqAC3 is distorted due
5 to the influence of voltage saturation and becomes a nonsinusoidal signal.
[0042] When the load torque pulsation is small, the
vibration can be reduced to a certain extent without using
the vibration suppression control according to the present
10 embodiment. However, when the operation is performed as
illustrated in FIG. 4, the motor torque is distorted to
generate anomaly noise, and the dq-axis current command is
made larger than necessary to cause an increase in the
copper loss. Moreover, when the load torque pulsation is
15 large, the q-axis current pulsation iqAC may not be able to
be generated as intended, and the vibration may be
increased instead.
[0043] FIG. 5 illustrates a state in which the highfrequency q-axis current pulsation iqAC1 is to be generated
20 with the d-axis current id being sufficiently large so as
not to cause voltage saturation. If the d-axis current id
is sufficiently large, the problem as illustrated in FIG. 4
does not occur, but the method of FIG. 5 has a problem in
terms of energy saving as described above.
25 [0044] On the other hand, FIG. 6 illustrates a state in
which the vibration suppression control according to the
present embodiment is performed, and represents a locus of
the voltage command vector suitable for passing the highfrequency q-axis current pulsation iqAC. Here, it is
30 assumed that the high-frequency q-axis current pulsation
iqAC can be expressed as in expression (3).
[0045]
[Expression 3]
24
... (3)
[0046] Here, “fd” represents the frequency of the
disturbance applied to the alternating-current motor 1 by
the mechanical device 2, “IqAMP” represents the amplitude of
5 the high-frequency q-axis current pulsation, and “δ”
represents a phase correction amount. It is assumed that
the disturbance frequency fd is sufficiently higher than a
design response of the speed control unit or feedback flux
weakening control unit (not illustrated). The amplitude of
10 the high-frequency q-axis current pulsation IqAMP and the
phase correction amount δ are parameters that may be
determined by the designer of the motor drive device 105.
Normally, the amplitude of the high-frequency q-axis
current pulsation IqAMP and the phase correction amount δ
15 are determined so as to suppress the vibration of the motor
drive device 105, but may be determined using another
criterion. The amplitude of the high-frequency q-axis
current pulsation IqAMP and the phase correction amount δ
may be determined by trial and error evaluation of an
20 actual device, or may be determined using the vibration
suppression control described in the aforementioned
Japanese Patent Application Laid-open No. H01-308184 or the
like.
[0047] When the disturbance frequency fd is sufficiently
25 higher than the design response of the speed control unit
(not illustrated) and the alternating-current motor 1 is to
be rotated at a constant speed, the low-frequency q-axis
current iqDC including the direct current component can be
regarded as substantially constant as expressed by
30 expression (4).
[0048]
[Expression 4]
i  I sin2f t   qAC qAMP d
25
... (4)
[0049] Similarly, when the disturbance frequency fd is
sufficiently higher than the design response of the
feedback flux weakening control unit (not illustrated) and
5 the alternating-current motor 1 is to be rotated at a
constant speed, the low-frequency d-axis current idDC
including the direct current component can also be regarded
as substantially constant as expressed by expression (5).
Note that because the d-axis current idDC may be adjusted
10 manually or may be calculated by substituting the d-axis
current idDC into expression (2), the feedback flux
weakening control unit need not necessarily be included.
[0050]
[Expression 5]
15 ... (5)
[0051] In FIG. 6, the d-axis current pulsation idAC is
given such that the locus of the tip of the voltage command
vector coincides with the voltage limit circle 21 in a
state in which the low-frequency d-axis current idDC
20 including the direct current component is passed. The
integral flux weakening control described above is known as
a method of passing the d-axis current idDC. When the sum
of the q-axis currents iqDC+iqAC is large, a larger negative
d-axis current is required, and when the sum of the q-axis
25 currents iqDC+iqAC is small, the negative d-axis current
should be reduced to reduce the copper loss. Therefore, in
the present embodiment, the d-axis current pulsation idAC is
changed in synchronization with the q-axis current
pulsation iqAC. As a result, the torque distortion is also
30 reduced, and the problem of the response of the integral
flux weakening control unit can also be solved. Although
details will be described later, there is an advantage of
i Constant qDC 
i Constant dDC 
26
being more robust against fluctuations in the motor
constant compared to the flux weakening control of
expression (2).
[0052] For this reason, in the present embodiment, the
5 d-axis current pulsation generating unit 6 determines the
d-axis current pulsation command idAC
* from the q-axis
current pulsation iqAC.
[0053] A specific method of calculating the d-axis
current pulsation idAC in the d-axis current pulsation
10 generating unit 6 will be described. Although a
calculation formula that causes the locus of the tip of the
voltage command vector to completely coincide with the
voltage limit circle 21 may be established, a simpler and
practical method will be considered here. FIG. 7 is a
15 first diagram illustrating a simple method of calculating
the d-axis current pulsation idAC in the d-axis current
pulsation generating unit 6 according to the first
embodiment. FIG. 8 is a second diagram illustrating the
simple method of calculating the d-axis current pulsation
20 idAC in the d-axis current pulsation generating unit 6
according to the first embodiment. When a vector sum of a
voltage ωeLqiqDC, a voltage ωeLdidDC, and the speed
electromotive force ωeΦa is defined as an average voltage
command vector, a tangent is drawn to the voltage limit
25 circle 21 from an intersection of the average voltage
command vector and the voltage limit circle 21. Here, an
angle formed by the average voltage command vector and the
q axis is referred to as an average voltage phase angle
θνave. When the locus of the voltage command vector is to
30 be matched with the tangent, the d-axis current pulsation
command idAC
* can be easily determined by focusing on two
right triangles illustrated in FIG. 7. FIG. 8 is an
enlarged view of the two right triangles in FIG. 7. It is
27
apparent by an elementary geometric calculation that one
angle of each of these right triangles is equal to θνave.
Then, from FIG. 8, it can be seen that when the q-axis
current pulsation iqAC and the average voltage phase angle
5 θνave are known, a desired voltage command vector locus can
be obtained if the d-axis current pulsation idAC expressed
in expression (6) can be passed. Note that when the
average voltage phase angle θνave can be regarded as being
substantially constant, expression (6) can be considered as
10 a linear function of the q-axis current pulsation iqAC.
[0054]
[Expression 6]
... (6)
[0055] That is, it can also be said that the d-axis
15 current pulsation generating unit 6 generates the d-axis
current pulsation command idAC
* on the basis of a result of
multiplication of a tangent of the average value of the
voltage advance angle and the q-axis current pulsation iqAC
or the q-axis current pulsation command iqAC
*. As
20 illustrated in FIG. 7, under the condition that the q-axis
current pulsation iqAC is relatively small, a difference
between the circular locus and the tangential locus is
small. Therefore, the tangential locus seems to be
sufficient in practical use, but an approximation error is
25 not easily ignored when the vibration is very large. In a
case where it is desired to allow the locus to further
approach the circular locus, it is only required to add a
square term, a cubic term, a quartic term, or the like of
the q-axis current pulsation iqAC to expression (6) to
30 increase the order of approximation and slightly distort
the d-axis current pulsation idAC. In the vicinity of the
  ave
d
q
dAC qAC tan
L
L
i i   
28
operating point, the locus considerably close to the
circular locus is obtained around the quartic approximation.
As described above, the d-axis current pulsation generating
unit 6 generates the d-axis current pulsation command idAC
*
5 such that the locus of the vector of the voltage command is
maintained in the circumferential direction or the
tangential direction of the voltage limit circle 21 having
a specified radius, which in this case is set to the limit
value Vom.
10 [0056] Compared to expression (2), expression (6) has a
reduced number of motor constants used for calculation.
Expression (2) uses the dq-axis flux linkage Φa for
calculation, whereas expression (6) does not. Therefore,
expression (6) is robust against fluctuations in “Φa”.
15 Although expressions (2) and (6) both use the dq-axis
inductances Ld and Lq for calculation, the influence of the
inductance error is smaller in expression (6). Moreover,
in expression (2), both the low frequency component and the
high frequency component are affected by the inductance
20 error, but in expression (6), only the high frequency
component is affected by the inductance error. It can thus
be said that the flux weakening control of the present
embodiment is more robust against fluctuations in the motor
constant than the conventional method using expression (2).
25 [0057] FIG. 9 is a diagram illustrating an example of
operation waveforms as a comparative example when the
vibration suppression control according to the first
embodiment is not performed. FIG. 10 is a diagram
illustrating an example of operation waveforms when the
30 vibration suppression control according to the first
embodiment is performed. FIGS. 9 and 10 are drawn on the
same scale. Both waveforms are the waveforms when the
mechanical device 2 is a rotary compressor and is operated
29
at a high speed, and the only difference in conditions is
whether the vibration suppression control of the present
embodiment is applied or not. In the rotary compressor,
although depending on the number of compression chambers,
5 load torque pulsation of one cycle to M cycles occurs
during one rotation of the alternating-current motor 1.
Note that “M” is an integer of 2 or more. Here, the
feedback vibration suppression control is used to perform
control for matching fundamental wave components of the
10 load torque and the motor torque. In addition, the
integral flux weakening control is performed as the flux
weakening control of the low frequency component including
the direct current component.
[0058] First, when the torque waveforms are compared, it
15 can be seen that the fundamental waves of the load torque
and the motor torque roughly coincide with each other in
FIGS. 9 and 10, but the torque distortion is larger in FIG.
9. This torque distortion is not intended but causes noise,
vibration, or the like, and thus is not preferable.
20 [0059] Next, when the voltage amplitudes are compared,
it can be seen that the voltage fluctuates erratically in
FIG. 9 but is controlled to be roughly constant in FIG. 10.
In FIG. 9, the integral flux weakening control attempts to
control the voltage amplitude to be constant, but the
25 integral flux weakening control unit does not have
sufficient responsiveness to the high-frequency q-axis
current pulsation iqAC. Therefore, when the high-frequency
q-axis current pulsation iqAC is applied, the voltage
amplitude cannot be controlled to be constant. This
30 tendency is remarkable when the q-axis current pulsation
iqAC is large. On the other hand, in FIG. 10, the flux
weakening control for the q-axis current pulsation iqAC is
performed separately so that the voltage amplitude can be
30
made constant.
[0060] Finally, the d-axis current and the q-axis
current will be compared. As described above with
reference to FIGS. 3 to 5, a desired q-axis current cannot
5 be passed when the voltage is saturated. When the desired
q-axis current cannot be passed, the feedback vibration
suppression control unit attempts to increase the pulsation
of the q-axis current command. As a result, the q-axis
current command is very large in FIG. 9. However, a higher
10 voltage is required to pass a large q-axis current. Here,
in order to reduce the amplitude of the voltage command
vector |ν*|, the integral flux weakening control unit
increases the d-axis current id in the negative direction.
As a result, when absolute value averages of the d-axis
15 current in FIGS. 9 and 10 are compared, the absolute value
average is larger in FIG. 9. However, although the
absolute value average of the d-axis current is increased,
the q-axis current cannot follow the command value, and
distortion occurs at the peaks of the waveform. This
20 distortion of the q-axis current causes torque distortion.
In FIG. 9, the absolute value average of the d-axis current
increases in addition to the increase in the amplitude of
the q-axis current command due to the influence of voltage
saturation, so that a copper loss (not illustrated)
25 increases and efficient operation cannot be performed. On
the other hand, in FIG. 10, because the appropriate d-axis
current pulsation idAC is calculated, the amplitude of the
q-axis current command does not increase, and the absolute
value average of the d-axis current does not increase.
30 Therefore, the copper loss is reduced in FIG. 10 as
compared to FIG. 9, and efficient vibration suppression
control can be implemented.
[0061] For these reasons, when the vibration suppression
31
control is performed, it can be said that it is a
reasonable method to perform the flux weakening control by
separating the low frequency component and the high
frequency component.
5 [0062] An operation of the motor drive device 105 will
be described with reference to a flowchart. FIG. 11 is a
flowchart illustrating the operation of the motor drive
device 105 according to the first embodiment. The current
detection unit 4 detects a phase current of the
10 alternating-current motor 1 (step S1). The position/speed
specifying unit 5 specifies the magnetic pole position and
the rotational speed of the alternating-current motor 1 on
the basis of a voltage command output from the dq-axis
current control unit 7 and the phase current of the
15 alternating-current motor 1 detected by the current
detection unit 4 (step S2). The d-axis current pulsation
generating unit 6 generates the d-axis current pulsation
command idAC
* using the q-axis current pulsation iqAC (step
S3). The dq-axis current control unit 7 determines a dq20 axis current command using the magnetic pole position and
the rotational speed of the alternating-current motor 1
specified by the position/speed specifying unit 5, the
phase current of the alternating-current motor 1 detected
by the current detection unit 4, the q-axis current
25 pulsation iqAC, and the d-axis current pulsation idAC
generated by the motor drive device 105, and generates a
voltage command for controlling the current passing through
the alternating-current motor 1 on the dq rotating
coordinates that rotate in synchronization with the
30 magnetic pole position (step S4).
[0063] As described above, according to the present
embodiment, the d-axis current pulsation generating unit 6
in the motor drive device 105 generates the d-axis current
32
pulsation command idAC
* that is in synchronization with the
q-axis current pulsation iqAC and prevents or reduces the
increase or decrease in the amplitude of the voltage
command vector due to the q-axis current pulsation iqAC.
5 The dq-axis current control unit 7 generates the voltage
command for the power conversion unit 3 using the d-axis
current pulsation command idAC
*. As a result, the motor
drive device 105 can prevent a reduction in efficiency
while efficiently performing the vibration suppression
10 control with the small d-axis current id in the
overmodulation region. The motor drive device 105 can
perform the vibration suppression control in the high-speed
region by preventing or reducing the increase or decrease
in the amplitude of the voltage command.
15 [0064] Second Embodiment.
A second embodiment will describe a specific hardware
configuration of the motor drive device 105 described in
the first embodiment.
[0065] FIG. 12 is a diagram illustrating an example of
20 the hardware configuration of the motor drive device 105
according to the second embodiment. In FIG. 12, parts
identical or corresponding to those in the configuration
illustrated in FIG. 1 will be denoted by the same reference
numerals as those in FIG. 1.
25 [0066] In the motor drive device 105, the d-axis current
pulsation generating unit 6 and the dq-axis current control
unit 7 are implemented by control circuitry 101. The
control circuitry 101 includes a processor 102 and a memory
103 as hardware. Although not illustrated, the memory 103
30 includes a volatile storage device such as a random access
memory and a nonvolatile auxiliary storage device such as a
flash memory. Note that although not illustrated, the
memory 103 may include the volatile storage device such as
33
the random access memory and an auxiliary storage device
such as a hard disk instead of the nonvolatile auxiliary
storage device. The processor 102 executes a program input
from the memory 103. Because the memory 103 includes the
5 auxiliary storage device and the volatile storage device,
the program is input from the auxiliary storage device to
the processor 102 via the volatile storage device. The
processor 102 may also output data such as a calculation
result to the volatile storage device of the memory 103, or
10 may save the data in the auxiliary storage device via the
volatile storage device.
[0067] Various modes have been studied for the power
conversion unit 3 and the current detection unit 4, but
basically any mode may be used therefor. As for the
15 position/speed specifying unit 5, the position sensor 5a
may be included, but basically any type of sensor may be
used. Although not illustrated here, the motor drive
device 105 may further include a voltage detection unit
that detects an input/output voltage of the power
20 conversion unit 3, a direct current bus voltage, and the
like.
[0068] Basically any method may be used as a method of
transmitting and receiving data between the components.
The components may transmit data by a digital signal or an
25 analog signal. The digital signal may be communicated by
parallel communication or serial communication. The analog
signal and the digital signal may be converted as
appropriate by a converter (not illustrated). For example,
in a case where the phase current detected by the current
30 detection unit 4 is expressed by an analog signal, the
analog signal is converted into a digital signal by a
digital to analog (D/A) converter (not illustrated), and
data is transmitted to the processor 102. The D/A
34
converter (not illustrated) may be inside the control
circuitry 101 or inside the current detection unit 4.
Similarly, the signal of the voltage command transmitted
from the processor 102 to the power conversion unit 3 may
5 be an analog signal or a digital signal. The processor 102
may also include a modulation unit such as a carrier
comparison modulation unit or a spatial vector modulation
unit, and may transmit a pulse train after modulation as
the voltage command from the processor 102 to the power
10 conversion unit 3.
[0069] A similar method applies to the communication
with the position sensor 5a, the voltage detection unit,
and the control circuitry 101.
[0070] The processor 102 performs calculation of
15 expression (6) on the basis of the q-axis current pulsation
iqAC of the alternating-current motor 1, and calculates the
d-axis current pulsation command idAC
*. The processor 102
then performs current control on the basis of the dq-axis
current pulsation and determines the voltage command,
20 thereby implementing the vibration suppression control in
the high-speed region. Note that the q-axis current
pulsation iqAC may be a command value or a detection value.
The q-axis current pulsation iqAC may be given from another
computer (not illustrated) or may be calculated inside the
25 processor 102. In addition, the processor 102 may perform
other calculation processing if having the capacity to do
so. That is, the processor 102 may perform other control
calculation processing such as a speed control calculation,
a vibration suppression control calculation, or an integral
30 flux weakening control calculation not illustrated in FIGS.
1 and 12.
[0071] Third Embodiment.
The configuration of the motor drive device 105
35
illustrated in FIG. 1 described in the first embodiment is
the minimum configuration. A third embodiment will
specifically describe an example of the configuration of
the motor drive device more suitable for practical use.
5 Differences from the first embodiment will be described.
[0072] FIG. 13 is a block diagram illustrating an
example of the configuration of a motor drive device 105a
according to the third embodiment. The motor drive device
105a is obtained by adding a subtraction unit 8, a first
10 speed control unit 9, a second speed control unit 10, a
subtraction unit 11, and a flux weakening control unit 12
to the motor drive device 105 of the first embodiment
illustrated in FIG. 1. Note that the position/speed
specifying unit 5 may have any configuration but here
15 estimates the magnetic pole position and the rotational
speed of the alternating-current motor 1 from the phase
current of the alternating-current motor 1 detected by the
current detection unit 4 and the voltage command input to
the power conversion unit 3.
20 [0073] The subtraction unit 8 calculates a difference
between a speed command indicating the rotational speed of
the alternating-current motor 1 and an estimated speed
being the rotational speed estimated by the position/speed
specifying unit 5, and outputs the difference as a speed
25 deviation eω.
[0074] The first speed control unit 9 controls the
average speed of the alternating-current motor 1 using the
speed deviation eω. The first speed control unit 9
generates a q-axis current command iqDC
* that has a lower
30 frequency than the q-axis current pulsation iqAC or the qaxis current pulsation command iqAC
* and controls the
average speed of the alternating-current motor 1. The
first speed control unit 9 typically performs feedback
36
control, but may perform feedforward control. The first
speed control unit 9 determines the q-axis current command
iqDC
* on the low frequency side including the direct current
component such that the speed deviation eω equals zero. It
5 is known that when a PI control unit is used as the first
speed control unit 9, a steady state speed deviation with
respect to a step response equals zero. Although the PI
control unit has a simple gain design, another control law
may be used. However, the first speed control unit 9 has a
10 restriction in response design, and it is considered that
the first speed control unit 9 alone cannot suppress highfrequency speed pulsation.
[0075] A current control response in general position
sensorless control is about 3000 to 4000 rad/s. Because
15 the first speed control unit 9 is configured as an outer
loop of current control, a speed control response of the
first speed control unit 9 can only be designed to be about
one-tenth of the current control response considering
control stability. Therefore, the speed control response
20 of the first speed control unit 9 is limited to 300 to 400
rad/s. In a case where the angular frequency of the load
torque pulsation of the mechanical device 2 is higher than
the speed control response, the first speed control unit 9
alone cannot suppress the speed pulsation so that the
25 second speed control unit 10 is required.
[0076] The second speed control unit 10 generates the qaxis current pulsation command iqAC
* for suppressing speed
pulsation caused by the load torque pulsation. The second
speed control unit 10 is referred to as a vibration
30 suppression control unit, a speed pulsation suppression
control unit, or the like, and various methods have already
been proposed therefor. Here, for convenience of
description, a feedback vibration suppression technique
37
described in the aforementioned Japanese Patent Application
Laid-open No. H01-308184 will be described as an example.
Note that characteristics of the load torque pulsation may
be evaluated in advance, and feedforward compensation may
5 be performed on the basis of a result of the evaluation
performed in advance. However, because the feedback method
requires less adjustment, the description will be made
assuming that the feedback method is employed.
[0077] In the method disclosed in the aforementioned
10 Japanese Patent Application Laid-open No. H01-308184, the
second speed control unit 10 performs Fourier series
expansion on the speed deviation eω at a frequency of a
periodic disturbance on the condition that the frequency of
the periodic disturbance is known, and extracts a cosine
15 component Eωcos and a sine component Eωsin. For example, when
the mechanical device 2 is a rotary compressor, the cosine
component Eωcos and the sine component Eωsin of the speed
deviation eω are calculated by expressions (7) and (8).
Note that although it is apparent to those skilled in the
20 art, the integration operation and the division processing
of expressions (7) and (8) may be substituted by
approximate integration using a low-pass filter.
[0078]
[Expression 7]
25 ... (7)
[0079]
[Expression 8]
... (8)
[0080] In expressions (7) and (8), “k” represents the
30 number of compression chambers. For example, a single
  m m
2
0
cos e cos k d
1
E  

 

 
  m m
2
0
sin e sin k d
1
E  

 

 
38
rotary compressor has k=1, and a twin rotary compressor has
k=2.
[0081] Next, the second speed control unit 10 performs
integral control on each of the cosine component Eωcos and
5 the sine component Eωsin. The control is simple because the
cosine component Eωcos and the sine component Eωsin are
direct current values. When “Iqcos
*” and “Iqsin
*” represent
results of the respective integral controls, the second
speed control unit 10 can determine the q-axis current
pulsation command iqAC
* 10 by restoring these to alternating
current values. Expression (9) is an example of the
arithmetic expression.
[0082]
[Expression 9]
15 ... (9)
[0083] The above is the method of vibration suppression
control disclosed in the aforementioned Japanese Patent
Application Laid-open No. H01-308184. This method is
simple and highly effective, and thus is widely used in
20 industry.
[0084] Note that although it is apparent to those
skilled in the art, the control block diagram can be
modified as appropriate. For example, because the q-axis
current and the motor torque are roughly proportional in
25 the permanent magnet synchronous motor, the outputs of the
first speed control unit 9 and the second speed control
unit 10 may be expressed in the dimension of the torque
instead of the dimension of the q-axis current.
[0085] In the low and middle speed region where the
30 voltage is not saturated by the alternating current voltage
output from the power conversion unit 3 to the alternatingcurrent motor 1, the two control units being the first
    iqAC  Iqcossin km  Iqsin cos km
  
39
speed control unit 9 and the second speed control unit 10
are sufficient. Meanwhile, in the high-speed region where
the voltage is saturated, a desired q-axis current cannot
be passed unless flux weakening control is performed. The
5 flux weakening control unit 12 is thus required. The flux
weakening control unit 12 generates a d-axis current
command idDC
* having a lower frequency than the d-axis
current pulsation command idAC
* for maintaining the
amplitude of the voltage command at a value less than or
10 equal to a specified value. Various proposals have been
made for the flux weakening control, and basically any
method may be used such as the integral flux weakening
control described in detail in the first embodiment. In
this method, first, the subtraction unit 11 calculates a
15 voltage deviation between a voltage limit value and the
amplitude of the voltage command. The flux weakening
control unit 12 performs integral control using the voltage
deviation calculated by the subtraction unit 11 as an input.
Note that a block for calculating the amplitude of the
20 voltage command from the vector of the voltage command is
not illustrated. The flux weakening control unit 12
determines the d-axis current command idDC
* on the low
frequency side including the direct current component. As
with the first speed control unit 9, the flux weakening
25 control unit 12 has a restriction in the control response,
and thus the flux weakening control response is limited to
300 to 400 rad/s. In a case where the angular frequency of
the load torque pulsation of the mechanical device 2 is
higher than the flux weakening control response, the flux
30 weakening control unit 12 alone cannot follow a voltage
change due to the q-axis current pulsation command iqAC
*
generated by the second speed control unit 10, so that
appropriate vibration suppression control cannot be
40
performed.
[0086] Therefore, the motor drive device 105a calculates
the d-axis current pulsation command idAC
* from the q-axis
current pulsation command iqAC
* using the d-axis current
5 pulsation generating unit 6. On the basis of the q-axis
current pulsation command iqAC
*, the d-axis current
pulsation generating unit 6 generates the d-axis current
pulsation command idAC
* that is in synchronization with the
q-axis current pulsation command iqAC
* and prevents or
10 reduces the increase or decrease in the amplitude of the
voltage command. The motor drive device 105a increases or
decreases the d-axis current in accordance with the
pulsation of the q-axis current to maintain the locus of
the tip of the voltage command vector in the tangential
15 direction or the circumferential direction of the voltage
limit circle 21. The motor drive device 105a can thus
perform the vibration suppression control with high
efficiency and high performance.
[0087] The dq-axis current control unit 7 determines the
20 dq-axis current command from the magnetic pole position,
the rotational speed, the phase current, the d-axis current
command idDC
*, the q-axis current pulsation command iqAC
*,
the q-axis current command iqDC
*, and the q-axis current
pulsation command iqAC
*. The dq-axis current control unit 7
25 then performs current control on the dq rotating
coordinates and determines a dq-axis voltage command. The
dq-axis current control unit 7 generates a voltage command
by performing coordinate transformation on the dq-axis
voltage command to obtain a command value on three-phase
30 coordinates, and outputs the voltage command to the power
conversion unit 3 to drive the alternating-current motor 1.
[0088] An operation of the motor drive device 105a will
be described with reference to a flowchart. FIG. 14 is the
41
flowchart illustrating the operation of the motor drive
device 105a according to the third embodiment. The current
detection unit 4 detects the phase current of the
alternating-current motor 1 (step S11). The position/speed
5 specifying unit 5 specifies the magnetic pole position and
the rotational speed of the alternating-current motor 1 on
the basis of the voltage command output from the dq-axis
current control unit 7 and the phase current of the
alternating-current motor 1 detected by the current
10 detection unit 4 (step S12). The first speed control unit
9 generates the q-axis current command iqDC
* using the speed
deviation eω (step S13). The second speed control unit 10
generates the q-axis current pulsation command iqAC
* for
suppressing the speed pulsation caused by the load torque
15 pulsation (step S14). The d-axis current pulsation
generating unit 6 generates the d-axis current pulsation
command idAC
* using the q-axis current pulsation command
iqAC
* (step S15). The flux weakening control unit 12
generates the d-axis current command idDC
* for maintaining
20 the amplitude of the voltage command at a value less than
or equal to a specified value (step S16). The dq-axis
current control unit 7 determines the dq-axis current
command using the magnetic pole position and the rotational
speed of the alternating-current motor 1 specified by the
25 position/speed specifying unit 5, the phase current of the
alternating-current motor 1 detected by the current
detection unit 4, the q-axis current command iqDC
*, the qaxis current pulsation command iqAC
*, the d-axis current
pulsation idAC, and the d-axis current command idDC
*, and
30 generates the voltage command for controlling the current
passing through the alternating-current motor 1 on the dq
rotating coordinates that rotate in synchronization with
the magnetic pole position (step S17).
42
[0089] Note that as for a hardware configuration of the
motor drive device 105a, the subtraction unit 8, the first
speed control unit 9, the second speed control unit 10, the
subtraction unit 11, and the flux weakening control unit 12
5 are implemented by the control circuitry 101 as with the daxis current pulsation generating unit 6 and the dq-axis
current control unit 7 of the motor drive device 105 of the
first embodiment.
[0090] As described above, according to the present
10 embodiment, in the motor drive device 105a, the first speed
control unit 9 generates the q-axis current command iqDC
*,
the second speed control unit 10 generates the q-axis
current pulsation command iqAC
*, and the flux weakening
control unit 12 generates the d-axis current command idDC
*.
15 In this case as well, effects similar to those of the first
embodiment can be obtained.
[0091] Fourth Embodiment.
In the first to third embodiments, the d-axis current
pulsation generating unit 6 determines the d-axis current
pulsation command idAC
* 20 such that the locus of the tip of
the voltage command vector coincides with the voltage limit
circle 21 as much as possible. However, in a state where
the voltage is not saturated, the flux weakening control is
not required in the first place so that the copper loss of
25 the alternating-current motor 1 can be reduced by setting
the d-axis current pulsation command idAC
* to zero. A
fourth embodiment will describe a case where the copper
loss of the alternating-current motor 1 is reduced in the
state where the voltage is not saturated.
30 [0092] FIG. 15 is a block diagram illustrating an
example of a configuration of a motor drive device 105b
according to the fourth embodiment. FIG. 16 is a diagram
illustrating an example of a voltage vector locus when the
43
voltage is not saturated in the motor drive device 105b
according to the fourth embodiment. Note that the d axis
and the locus of the tip of the voltage command vector are
illustrated to be parallel in FIG. 16, but are seldom
5 completely parallel in practice. In actual operation, the
locus has a certain degree of slope. The locus of the tip
of the voltage command vector moves roughly parallel to the
d-axis.
[0093] The motor drive device 105b is obtained by adding
10 a d-axis current pulsation command selecting unit 13 to the
motor drive device 105a of the third embodiment illustrated
in FIG. 13. In the fourth embodiment, the dq-axis current
control unit 7 includes therein a component (not
illustrated) for determining a voltage saturated state and
15 a voltage non-saturated state, and outputs a result of
determination as a voltage non-saturated flag to the d-axis
current pulsation command selecting unit 13. In the
voltage non-saturated state, the d-axis current pulsation
command selecting unit 13 forcibly changes the d-axis
current pulsation command idAC
* 20 to zero. In FIG. 15, one
input of the d-axis current pulsation command selecting
unit 13 is set to zero, but when the d-axis current
pulsation command idAC
* is a minute amount, the purpose of
not generating extra d-axis current pulsation idAC can be
25 achieved. Accordingly, a specified minute amount may be
input instead of zero. The minute amount may be referred
to as a first value. That is, the d-axis current pulsation
command selecting unit 13 selects the d-axis current
pulsation command idAC
* or the first value capable of
30 reducing the loss in the alternating-current motor 1 on the
basis of the control of the dq-axis current control unit 7,
and outputs the selected d-axis current pulsation command
idAC
* or first value to the dq-axis current control unit 7.
44
When the output voltage of the power conversion unit 3 is
under the maximum value, the dq-axis current control unit 7
causes the d-axis current pulsation command selecting unit
13 to output the first value.
5 [0094] As described above, the motor drive device 105b
switches the d-axis current pulsation command idAC
* in the
voltage non-saturated state and the voltage saturated state,
thereby being able to achieve highly efficient vibration
suppression control in either state.
10 [0095] An operation of the motor drive device 105b will
be described with reference to a flowchart. FIG. 17 is the
flowchart illustrating the operation of the motor drive
device 105b according to the fourth embodiment. In the
flowchart of FIG. 17, operations other than step S21 are
15 similar to those in the flowchart of the third embodiment
illustrated in FIG. 14. In the motor drive device 105b,
the d-axis current pulsation command selecting unit 13
selects the d-axis current pulsation command idAC
* or the
first value as the output value to the dq-axis current
20 control unit 7 on the basis of the voltage non-saturated
flag from the dq-axis current control unit 7 (step S21).
[0096] Note that as for a hardware configuration of the
motor drive device 105b, the d-axis current pulsation
command selecting unit 13 is implemented by the control
25 circuitry 101 as with the d-axis current pulsation
generating unit 6 and the dq-axis current control unit 7 of
the motor drive device 105 of the first embodiment.
[0097] As described above, according to the present
embodiment, in the motor drive device 105b, the d-axis
30 current pulsation command selecting unit 13 selects and
outputs the d-axis current pulsation command idAC
* or the
first value on the basis of the voltage non-saturated flag
from the dq-axis current control unit 7. The dq-axis
45
current control unit 7 generates the voltage command using
the d-axis current pulsation command idAC
* or the first
value. As a result, the motor drive device 105b can
improve the motor drive efficiency when the voltage is not
5 saturated as compared to the cases of the first to third
embodiments.
[0098] Fifth Embodiment.
In the fourth embodiment, the d-axis current pulsation
command selecting unit 13 switches the d-axis current
pulsation command idAC
* 10 in the voltage non-saturated state
and the voltage saturated state. Here, the switching
processing is desirably performed seamlessly. A fifth
embodiment will describe a case where the d-axis current
pulsation command idAC
* is switched seamlessly.
15 [0099] FIG. 18 is a block diagram illustrating an
example of a configuration of a motor drive device 105c
according to the fifth embodiment. The motor drive device
105c is obtained by adding an offset correction unit 14 and
an addition unit 15 to the motor drive device 105a of the
20 third embodiment illustrated in FIG. 13. The offset
correction unit 14 receives the voltage command and the
voltage limit value as inputs, and outputs an offset
correction amount. That is, the offset correction unit 14
calculates the offset correction amount with respect to the
d-axis current pulsation command idAC
* 25 on the basis of the
voltage command and the voltage limit value indicating the
maximum voltage that can be output as the voltage command.
The addition unit 15 adds the offset correction amount to
the d-axis current pulsation command idAC
* output from the
30 d-axis current pulsation generating unit 6. The addition
unit 15 outputs a result of the addition to the dq-axis
current control unit 7 as a new d-axis current pulsation
command idAC
*. The dq-axis current control unit 7 includes
46
a current limiter (not illustrated), and limits the d-axis
current command to an appropriate value when the new d-axis
current pulsation command idAC
* takes a positive value or
exceeds a command value calculated by a maximum
5 torque/current control unit (not illustrated).
[0100] Here, a reason for performing offset correction
in the motor drive device 105c will be described. FIG. 19
is a first diagram for explaining the necessity of offset
correction performed by the motor drive device 105c
10 according to the fifth embodiment. FIG. 20 is a second
diagram for explaining the necessity of offset correction
performed by the motor drive device 105c according to the
fifth embodiment. The first to fourth embodiments have
described the locus of the tip of the voltage command
15 vector suitable for the vibration suppression control. As
described above, the circumferential locus or the
tangential locus is suitable for the vibration suppression
control when the voltage is saturated, and the locus
parallel to the d-axis is suitable when the voltage is not
20 saturated. From these two matters, in a case where the
voltage is slightly saturated, it is conceived that a
combination of the circumferential locus and the locus
parallel to the d-axis is suitable for the vibration
suppression control. FIG. 19 specifically illustrates such
a locus. The d-axis current pulsation command idAC
* 25 at this
time has a non-sinusoidal waveform by the above-described
current limiting processing. As a method of determining
the d-axis current command in the low and middle speed
region, id=0 control, maximum torque/current control, and
30 the like are widely used. It is apparent that the copper
loss increases when idDC
*+idAC
* is larger in the positive
direction than the command value determined by these
control units. Therefore, the d-axis current pulsation
47
command idAC
* is limited. As a result, the d-axis current
pulsation command idAC
* becomes a non-sinusoidal wave. At
first glance, it seems that it is sufficient to simply
apply the current limiting processing, but in such a case,
5 the locus illustrated in FIG. 20 is obtained instead of the
locus as in FIG. 19. This is because the amplitude of the
voltage command vector is not considered in expression (6).
When “νave
*” represents an average amplitude value of the
voltage command vector, the locus in FIG. 20 overlaps a
10 tangent of a circle 22 with the radius corresponding to the
average amplitude value νave
*. Compared to the locus of FIG.
19, the locus of FIG. 20 has a margin in voltage, and thus
it can be said that an extra d-axis current is passed. In
other words, the locus of FIG. 19 cannot be achieved unless
15 the d-axis current is reduced by the amount corresponding
to the voltage margin. In the present embodiment, the
processing of reducing the d-axis current is referred to as
the offset correction.
[0101] Although various means of the offset correction
20 are conceivable, here, a method focusing on simplicity of
calculation rather than geometric strictness will be
described. FIG. 21 is a diagram illustrating an example of
a method of calculating the offset correction amount in the
motor drive device 105c according to the fifth embodiment.
25 An offset correction amount idoff when the average amplitude
value νave
* of the voltage command vector has a margin with
respect to the limit value Vom, which is the radius of the
voltage limit circle 21, will be considered. When “ωeLdidoff”
represents a voltage margin amount in the q-axis direction,
30 a relationship of expression (10) can be derived from FIG.
21.
[0102]
[Expression 10]
48
... (10)
[0103] Depending on the way of thinking, the voltage
margin amount is estimated to be small on purpose in
expression (10). This is the procedure for canceling an
5 approximation error at the tip in the tangential locus. If
it is desired to perform the calculation more strictly, the
calculation may be performed by considering a portion
written as “unconsidered part” on which “locus deviation
due to unconsidered part” is based in FIG. 21.
10 [0104] Expression (11) is obtained by solving expression
(10) for the offset correction amount idoff.
[0105]
[Expression 11]
... (11)
15 [0106] When the current limiting processing is performed
after translating the d-axis current pulsation idAC by the
offset correction amount idoff, a voltage change
ωeLd(iqAC+idoff) in the q-axis direction with respect to
“ωeLqiqAC” has a constant limit value except for the
20 vicinity of the peak of the sinusoidal signal in a downward
direction. This processing can obtain the locus close to
that in FIG. 19. As a result, the vibration suppression
control having high efficiency and high performance can be
achieved in both the low and middle speed region and the
25 high-speed region.
[0107] An operation of the motor drive device 105c will
be described with reference to a flowchart. FIG. 22 is the
flowchart illustrating the operation of the motor drive
device 105c according to the fifth embodiment. In the
30 flowchart of FIG. 22, operations other than steps S31 and
S32 are similar to those in the flowchart of the third
    e d doff om ave ave L i V cos 
      
   
e d
om ave ave
doff L
V cos i

   



49
embodiment illustrated in FIG. 14. In the motor drive
device 105c, the offset correction unit 14 calculates the
offset correction amount with respect to the d-axis current
pulsation command idAC
* on the basis of the voltage command
5 and the voltage limit value (step S31). The addition unit
15 adds the offset correction amount to the d-axis current
pulsation command idAC
* output from the d-axis current
pulsation generating unit 6, and outputs a result of the
addition to the dq-axis current control unit 7 (step S32).
10 [0108] Note that as for a hardware configuration of the
motor drive device 105c, the offset correction unit 14 and
the addition unit 15 are implemented by the control
circuitry 101 as with the d-axis current pulsation
generating unit 6 and the dq-axis current control unit 7 of
15 the motor drive device 105 of the first embodiment.
[0109] As described above, according to the present
embodiment, the offset correction unit 14 of the motor
drive device 105c calculates the offset correction amount
with respect to the d-axis current pulsation command idAC
*.
20 The addition unit 15 adds the offset correction amount to
the d-axis current pulsation command idAC
* and outputs the
result of the addition to the dq-axis current control unit
7. As a result, the motor drive device 105c can obtain an
effect similar to that of the fourth embodiment, and can
also switch the d-axis current pulsation command idAC
* 25
seamlessly.
[0110] Sixth Embodiment.
A sixth embodiment will describe a compressor drive
system including the motor drive device 105 described in
30 the first embodiment. Note that although the description
will be made using the motor drive device 105, the motor
drive devices 105a, 105b, and 105c described in the third
to fifth embodiments can also be applied.
50
[0111] FIG. 23 is a diagram illustrating an example of a
configuration of a compressor drive system 200 according to
the sixth embodiment. The compressor drive system 200
includes the motor drive device 105 and a refrigerant
5 compressor 2a. The refrigerant compressor 2a is, for
example, a rotary compressor of a type called rolling
piston, but may be another type of compressor. The
refrigerant compressor 2a includes the alternating-current
motor 1, a shaft 201, a compression chamber 202, an intake
10 pipe 203, and a discharge pipe 204. The refrigerant taken
in from the intake pipe 203 is compressed by the
alternating-current motor 1 in the compression chamber 202
and discharged from the discharge pipe 204. In the
refrigerant compressor 2a, the alternating-current motor 1
15 is structured to be immersed in the refrigerant in many
cases, where it is difficult to attach the position sensor
5a to the alternating-current motor 1 because the
temperature changes greatly. Therefore, in the application
of the refrigerant compressor 2a, the alternating-current
20 motor 1 needs to be driven without a position sensor. Note
that the alternating-current motor 1 is driven by the motor
drive device 105.
[0112] The structure and load torque of the refrigerant
compressor 2a will be described. FIG. 24 is a schematic
25 diagram illustrating an example of a detailed configuration
of the compression chamber 202 included in the compressor
drive system 200 according to the sixth embodiment. The
compression chamber 202 includes the shaft 201, a piston
205, a discharge valve 208, a vane 210, a spring 209, a
30 discharge port 207, and an intake port 206. The shaft 201
connects the alternating-current motor 1 and the piston 205.
The compression chamber 202 is partitioned into two by the
piston 205, the vane 210, and the spring 209. The piston
51
205 is eccentric, and the volumes on a discharge side and
an intake side change depending on the rotational angle.
The refrigerant taken in from the intake side is compressed
by the piston 205. When the pressure in the compression
5 chamber 202 increases, the discharge valve 208 is opened so
that the refrigerant is discharged from the discharge port
207. At the same time as the refrigerant is discharged,
the refrigerant flows to the intake side. When the
alternating-current motor 1 continues to be rotated, the
10 refrigerant is discharged once per rotation, in the
mechanical angle, of the refrigerant compressor 2a.
[0113] FIG. 25 is a graph illustrating an example of a
waveform of the load torque of the refrigerant compressor
2a according to the sixth embodiment. The refrigerant
15 compressor 2a illustrated in FIG. 23 includes only one
compression chamber 202, but may include a plurality of
compression chambers. Here, the number of compression
chambers is represented by “k”. When there is only one
compression chamber (k=1), the load torque fluctuates
20 greatly in the cycle of mechanical angle. Although second
and third harmonics are included in the load torque
waveform, the fluctuation due to the first is the largest.
When a plurality of compression chambers is provided in the
refrigerant compressor 2a, the load torque pulsation can be
25 reduced by shifting the angle of the piston 205. As the
number of compression chambers is increased, a waveform
with smaller pulsation can be obtained, which however
involves a complicated structure and an increase in cost.
The cycle of the load torque pulsation is reduced in
30 inverse proportion to the number of compression chambers.
When there are two compression chambers (k=2), the second
harmonic component increases. When there are three
compression chambers (k=3), the third harmonic component
52
increases.
[0114] The load torque pulsation of the refrigerant
compressor 2a becomes a periodic disturbance to the
alternating-current motor 1, and thus is a factor of the
5 speed pulsation. It is generally known that when the speed
pulsation is large, noise, vibration, and the like increase.
[0115] However, the frequencies of the load torque
pulsation, the speed pulsation, and the like are known
because the frequencies are determined by the structure of
10 the refrigerant compressor 2a. There is a known case that
uses this fact to establish the feedback vibration
suppression control. Moreover, in the present embodiment,
the motor drive device 105 causes the d-axis current
pulsation generating unit 6 to generate the d-axis current
pulsation command idAC
* 15 in synchronization with the q-axis
current pulsation iqAC. As a result, the motor drive device
105 can achieve the vibration suppression control with high
efficiency and high performance in the high-speed region
and the overmodulation region. In addition, when the motor
20 drive device 105c is applied, the offset correction unit 14
manages the voltage margin so that the vibration
suppression control can be achieved with high efficiency in
a wide range from the low-speed region to the high-speed
region. Application of the motor drive devices 105, 105a,
25 105b, and 105c to the compressor drive system 200 can
reduce vibration, noise, and the like of the refrigerant
compressor 2a under various conditions.
[0116] Seventh Embodiment.
A seventh embodiment will describe a refrigeration
30 cycle system including the compressor drive system 200
described in the sixth embodiment.
[0117] FIG. 26 is a diagram illustrating an example of a
configuration of a refrigeration cycle system 300 according
53
to the seventh embodiment. The refrigeration cycle system
300 includes the compressor drive system 200, a condenser
301, a liquid receiver 302, an expansion valve 303, and an
evaporator 304. The refrigerant compressor 2a of the
5 compressor drive system 200, the condenser 301, the liquid
receiver 302, the expansion valve 303, and the evaporator
304 are connected by piping so that the refrigerant
circulates. The refrigeration cycle system 300 transfers
heat while allowing the refrigerant to repeat phase changes
10 from liquid to gas and from gas to liquid through
circulation in the processes of evaporation, compression,
condensation, and expansion.
[0118] The role of each device will be described. The
evaporator 304 evaporates the refrigerant liquid at a low
15 pressure, takes heat from the surroundings, and performs a
cooling action. The refrigerant compressor 2a compresses
the refrigerant gas into high-pressure gas in order to
condense the refrigerant. The refrigerant compressor 2a is
driven by the motor drive device 105. The condenser 301
20 releases heat and condenses the high-pressure refrigerant
gas into the refrigerant liquid. The expansion valve 303
performs throttle expansion on the refrigerant liquid and
turns it into low-pressure liquid in order to evaporate the
refrigerant. The liquid receiver 302 is provided for
25 adjusting the amount of the refrigerant circulating, and
may be omitted in a small system.
[0119] Because the vibration generated by the
refrigerant compressor 2a causes breakage of the
refrigerant piping, noise, and the like, the alternating30 current motor 1 needs to be controlled so as to minimize
the vibration. As the feedback vibration suppression
control technique in the low and middle speed region has
been widely studied to be gradually established and spread,
54
the vibration suppression control in the high-speed region
has become required. In the high-speed region, the output
voltage of the power conversion unit 3 is saturated so that
proper torque control cannot be performed. In order to
5 perform the vibration suppression control in the high-speed
region, the flux weakening control needs to be used
together. However, as described above, the existing flux
weakening control method has a problem with robustness and
low control responsiveness. It has thus been considered
10 difficult to perform the vibration suppression control in
the high-speed region. A method is conceivable in which
“extra d-axis current is always passed to secure a voltage
margin for the vibration suppression control”, which
however is unrealistic because energy saving performance is
15 also considered to be important in the refrigeration cycle
system 300.
[0120] For these reasons, in order to achieve the
vibration suppression control in the high-speed region, it
is necessary to review not only the technique of the
20 vibration suppression control but also the method regarding
the flux weakening control. The motor drive devices 105 to
105c solve the above problem. The motor drive devices 105
to 105c have the mechanism for causing the d-axis current
pulsation in synchronization with the q-axis current
25 pulsation generated by the vibration suppression control
unit. The motor drive devices 105 to 105c separate the qaxis current into the “low frequency component including
the direct current component” and the “high frequency
component”, and perform the flux weakening control
30 separately. The motor drive devices 105 to 105c can thus
achieve the vibration suppression control with high
efficiency and high performance in the high-speed region
that has been difficult. This as a result can reduce the
55
cost of measures against breakage of the refrigerant piping,
noise, and the like and reduce the cost of the
refrigeration cycle system 300.
[0121] The configuration illustrated in the above
5 embodiment merely illustrates an example of the content of
the present invention, and can thus be combined with
another known technique or partially omitted and/or
modified without departing from the scope of the present
invention.
10
Reference Signs List
[0122] 1 alternating-current motor; 2 mechanical
device; 2a refrigerant compressor; 3 power conversion
unit; 4 current detection unit; 5 position/speed
15 specifying unit; 5a position sensor; 6 d-axis current
pulsation generating unit; 7 dq-axis current control unit;
8, 11 subtraction unit; 9 first speed control unit; 10
second speed control unit; 12 flux weakening control unit;
13 d-axis current pulsation command selecting unit; 14
20 offset correction unit; 15 addition unit; 101 control
circuitry; 105, 105a, 105b, 105c motor drive device; 200
compressor drive system; 300 refrigeration cycle system.

56
We Claim :
1. A motor drive device that controls driving of an
alternating-current motor connected to a mechanical device
with periodic load torque pulsation by using dq rotating
5 coordinates, the motor drive device comprising:
a power conversion unit to convert a direct current
voltage into an alternating current voltage on the basis of
a voltage command, and output the alternating current
voltage to the alternating-current motor;
10 a current detection unit to detect a phase current
flowing to the alternating-current motor;
a position/speed specifying unit to specify a magnetic
pole position and a rotational speed of the alternatingcurrent motor;
15 a d-axis current pulsation generating unit to generate
a d-axis current pulsation command on the basis of periodic
q-axis current pulsation or a periodic q-axis current
pulsation command, the d-axis current pulsation command
being in synchronization with the q-axis current pulsation
20 or the q-axis current pulsation command and preventing or
reducing an increase or decrease in amplitude of the
voltage command; and
a dq-axis current control unit to generate the voltage
command that controls the phase current on the dq rotating
25 coordinates, which rotate in synchronization with the
magnetic pole position, by using the magnetic pole position,
the rotational speed, the phase current, the q-axis current
pulsation or the q-axis current pulsation command, and the
d-axis current pulsation command.
30
2. The motor drive device according to claim 1, wherein
the d-axis current pulsation generating unit generates
the d-axis current pulsation command on the basis of a
57
result of multiplication of a tangent of an average value
of a voltage advance angle and the q-axis current pulsation
or the q-axis current pulsation command.
5 3. The motor drive device according to claim 1 or 2,
further comprising:
a first speed control unit to generate a q-axis
current command that has a lower frequency than the q-axis
current pulsation or the q-axis current pulsation command
10 and controls an average speed of the alternating-current
motor;
a second speed control unit to generate the q-axis
current pulsation command that suppresses speed pulsation
caused by the load torque pulsation; and
15 a flux weakening control unit to generate a d-axis
current command that has a lower frequency than the d-axis
current pulsation command and maintains the amplitude of
the voltage command to be less than or equal to a specified
value, wherein
20 the d-axis current pulsation generating unit generates
the d-axis current pulsation command on the basis of the qaxis current pulsation command, the d-axis current
pulsation command being in synchronization with the q-axis
current pulsation command and preventing or reducing an
25 increase or decrease in the amplitude of the voltage
command, and
the dq-axis current control unit generates the voltage
command using the magnetic pole position, the rotational
speed, the phase current, the q-axis current command, the
30 q-axis current pulsation command, the d-axis current
pulsation command, and the d-axis current command.
4. The motor drive device according to any one of claims
58
1 to 3, wherein
the d-axis current pulsation generating unit generates
the d-axis current pulsation command such that a locus of a
vector of the voltage command is maintained in a
5 circumferential direction or a tangential direction of a
voltage limit circle having a specified radius.
5. The motor drive device according to any one of claims
1 to 4, further comprising
10 a d-axis current pulsation command selecting unit to
select and output, to the dq-axis current control unit, the
d-axis current pulsation command or a first value that can
reduce a loss in the alternating-current motor on the basis
of control of the dq-axis current control unit, wherein
15 the dq-axis current control unit causes the d-axis
current pulsation command selecting unit to output the
first value when an output voltage of the power conversion
unit is under a maximum value.
20 6. The motor drive device according to any one of claims
1 to 4, further comprising:
an offset correction unit to calculate an offset
correction amount with respect to the d-axis current
pulsation command on the basis of the voltage command and a
25 voltage limit value indicating a maximum voltage that can
be output as the voltage command; and
an addition unit to add the offset correction amount
to the d-axis current pulsation command output from the daxis current pulsation generating unit, and output a result
30 of the addition to the dq-axis current control unit.
7. A compressor drive system comprising the motor drive
device according to any one of claims 1 to 6.

8. A refrigeration cycle system comprising the compressor
drive system according to claim 7.