JPH0884500A - Speed sensorless vector controller for induction motor - Google Patents

Abstract

PURPOSE: To control the speed and torque of an induction motor accurately even in a very low speed region. CONSTITUTION: A speed adaptive secondary flux observer comprising a flux observer 4 of same dimension and a speed adaptive mechanism 7 estimates the actual speed of an induction motor 1 and a current control section is controlled by means of a comparison error signal between an estimated motor speed ωr # and motor speed command ωr #. A PWM control converter 2 at the current control section compares primary voltage control command voltages vu , vv , vw with a carrier signal (triangular wave signal) and performs PWM control of an inverter depending on the comparison result. A PWM control inverter 2 sets the frequency of carrier signal at a conventional high level when an estimated motor speed ωr # is higher than a set value and decreases the frequency of carrier signal when the estimated motor speed ωr # drops below the set value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a speed sensorless vector control device for an induction motor, and is devised so that stable speed control can be performed even in a low speed range and accurate torque control can be performed.

[0002]

2. Description of the Related Art A slip frequency control type vector control method has become popular as a high performance speed control method for induction motors.
A speed sensorless vector control method for controlling this without a speed sensor is known.

FIG. 7 shows a control system of a conventional speed sensorless vector controller for an induction motor, which estimates the actual speed of the induction motor using a speed adaptive secondary magnetic flux observer.

First, the estimation of the motor speed (rotational angular frequency ω _r ) of the induction motor 1 will be described with reference to FIG.

The voltage equation of the induction motor 1 is given by the following equation (1) when it is represented by the ab axis for observing various quantities from the synchronous rotation coordinate system rotating at the power source angular frequency (ω ₀ ).

[0006]

[Equation 1]

However, v _1a , v _1b ... Primary excitation axis voltage, primary torque axis voltage (V) i _1a , i _{1b on} synchronous rotation coordinates (ab axis) i _1a , i _1b ... synchronous rotation coordinates (ab axis) ) Above primary excitation axis current, primary torque axis current (A) λ _2a , λ _2b ... Secondary excitation axis magnetic flux on secondary rotation coordinate (ab axis), secondary torque axis magnetic flux (Wb) ω ₀ ………… Power supply angular frequency (rad / sec) ω _r ………… Motor speed (rotational angular frequency, rad / sec) ω _s ………… Slip angular frequency (rad / sec) R ₁ , R ₂ … 1 Next, secondary resistance (Ω) L ₁ , L ₂ … Primary, secondary inductance (H) M ………… Mutual (excitation) inductance (H) Lσ ………… Equivalent leakage inductance (H) (Lσ = (L ₁ L ₂ −M ² ) / L ₂ ) s ………… Time fine molecule (d / dt) And power source angular frequency (ω ₀ ), motor speed (ω _r ),
Relationship slip angular frequency command value (omega _s ^*), and calculates the slip angular frequency command value (omega _s ^*) is represented by the following formula (2). ω ₀ = ω _r + ω _s ^* ω _s ^* = i _1b ^* / i _1a ^*・ τ ₂ …… (2) However, τ ₂ ………… Second-order time constant (τ ₂ = L ₂ / R ₂ ) Subscript (*): Indicates a command value or set value. Now, i _1a ^* = constant …………………………………………………………………………………………………… (3) When the primary voltage command values (v _1a ^* , v _1b ^* ) in the synchronous rotation coordinate system (ab-axis) are given by the following equation (4) which is a constitutive equation of the digital current controller 3, the following is obtained.

[0008]

[Equation 2]

When the control is performed so as to satisfy the above equation (4), the primary current detection value i ₁ on the synchronous rotation coordinate (ab axis) is determined.
Is the primary current command value i ₁ ^* (i _1a ^* , i _1b ^* ), and the secondary magnetic flux λ ₂ on the synchronous rotation coordinates (ab axis).
(Λ _2a , λ _2b ) is kept at λ _2a = Mi _1a (constant), λ _2b = 0 …………………… (5).

As a result, the torque (T) of the induction motor 1
Is T = M / L ₂ · (λ _2a · i _1b −λ _2b · i _1a ) = M ² / L ₂ · (i _1a · i _1b ) ………………………… (6) , The secondary magnetic flux λ on the synchronous rotation coordinates (ab axis)
₂ (λ _2a , λ _2b ) and the secondary current i ₂ (i _2a , i _2b ) are irrelevant to the non-interference control vector control.

By the way, as is clear from the above equation (2).
If the speed sensor is not used, the slip angular frequency command
Value ω_s ^*Even if is set, the motor speed ω_rIs unknown
Power source angular frequency ω₀Can't decide
Is the power source angular frequency ω₀Synchronous rotation coordinates (a-
secondary magnetic flux λ on the b-axis)₂(Λ_2a, Λ_2b) Is the above formula (5)
So that the power source angular frequency ω₀To control
Similarly, vector control of non-interference control is realized by
Can be That is, the same-dimensional magnetic flux observer 4
Speed adaptive secondary magnetic flux observer consisting of
Using synchronous rotation coordinates that satisfy the above equation (5)
Secondary magnetic flux λ on (ab axis)₂(Λ_2a, Λ _2b) Estimated
The secondary magnetic flux estimated value λ₂ ^#(Λ_2a ^#, Λ_2b ^#) Based on
Motor speed ω_rEstimate (ω_r ^#)
Equation (2) (ω₀= Ω_r ^#+ Ω_s ^*) From the power angular frequency
Number ω₀And the power source angular frequency ω₀Due to digital current
Vector of decoupling control by controlling the controller 3
Control can be realized.

In the conventional speed sensorless vector control method for an induction motor in the vector control system shown in FIG. 7, in order to detect the actual speed of the induction motor 1 without using a speed sensor, the primary current of the induction motor 1 ( Phase current) i _u , _iv , i _w are detected and the 3-phase-2 phase phase number converter 1
The primary current detection value i ₁ (i _1d , i _1q ) on the stator coordinates (dq axes) converted in phase number in 4 is set. Then, the detected primary current i ₁ and the electric motor 1 on the stator coordinates (dq axes)
The secondary magnetic flux on the stator coordinates (d-q axes) is calculated by the same-dimensional magnetic flux observer 4 that receives the next voltage command value v ₁ ^* (v _1d ^* , v _1q ^* ) and the estimated speed value ω _r ^#. Estimated value λ ₂ ^#
(Λ _2d ^# , λ _2q ^# ) and the primary current estimated value i ₁ ^# (i _1d ^# ,
i _1q ^# ) and the speed adaptation mechanism 7 estimates the primary current i ₁ ^# (i _1d ^# , i _1q ^# ) and the detected primary current i.
₁ (i _1d , i _1q ) is compared with the estimation error signal (i ₁ −i
And a motor speed estimation value (omega _r ^#) calculation estimated by the induction motor 1 speed detected by the adaptive tuning strategy represented by the following formula (7) based on ₁ ^#).

Ω _r ^# = K _p (e _id λ _2q ^# −e _iq λ _2d ^# ) + K _i ∫ (e _id λ _2q ^# −e _iq λ _2d ^# ) dt (7) However, e _id = i _1d −i _1d ^# : estimation error e _iq = i _1q −i _1q ^# : estimation error K _p : velocity estimation unit proportional gain K _i : velocity estimation unit integration gain Note that the same-dimensional secondary magnetic flux observer 4 and velocity adaptation mechanism are used. 7
For the speed sensorless vector control method of the induction motor 1 that estimates the actual speed of the induction motor by the speed adaptive secondary magnetic flux observer consisting of and, see "Journal of the Institute of Electrical Engineers of Japan, Volume 11, No. 11, 1991" (Kubota, Ozaki, Matsuse, Nakano: "Velocity sensorless direct vector control of induction motor using adaptive secondary magnetic flux observer").

A conventional speed sensorless vector control system for estimating the actual speed of the induction motor using the speed adaptive secondary magnetic flux observer will be described below.

Explaining the operation in FIG. 7 (control system configuration), when the motor speed command value (ω _r ^* ) is given to the speed control unit (ASR), the speed command value (ω _r ^* ) and the motor speed are given. The estimated value (ω _r ^# ) is compared, and the comparison error signal is proportional-integral (PI) controlled by the speed controller 6 to obtain 1 on the synchronous rotation coordinate (ab axis).
It is converted to the primary torque axis current command value (i _1b ^* ) which is the torque axis component of the secondary current command value i ₁ ^* . Next, in the digital current controller 3 in the current control unit (ACR), the primary voltage command value v ₁ ^* (v
_1a ^* , _v1b ^* ) is the primary current command value i
₁ ^* (i _1a ^* , i _1b ^* ) and primary current detection value i ₁ (i _1a ,
i _1b ) is equal (i _1a ^* = i _1a , i _1b ^* = i _1b ), which is calculated by the above-mentioned expression (4) which is a conditional expression enabling decoupling control.

Primary voltage finger on the synchronous rotation coordinates (ab axis)
Command v₁ ^*(V_1a ^*, V_1b ^*) By the coordinate converter 9
Primary voltage command value v on the stator coordinates (dq axes)₁ ^*(V
_1d ^*, V_1q ^*2 phase to 3 phase phase number converter
The number of phases is converted by 15 and the PWM control inverter 2 has three phases.
Phase Primary voltage control command voltage v for each phase_u, V_v, V_wConversion to
The output voltage of each of the three phases of the PWM control inverter 2 is
Control. As a result, the induction motor 1 is driven at the desired motor speed.
Command value (ω_r ^*) According to the speed control. Also power
Angular frequency ω₀Synchronous rotation coordinates (a-b axis) that rotate at
Stator coordinates (dq) fixed to the stator of the induction motor 1
Used for coordinate converters 8 and 9 to convert between
Unit vector (sin θ₀, Cos θ₀) To produce
Basic phase angle of₀(Θ₀= Ω₀t) is calculated as follows.
Can be That is, the slip calculator 5
As shown in equation (2), on the synchronous rotation coordinates (ab axis)
Primary excitation axis current command value i_1a ^*Primary torque axis current command value
i_1b ^*, And the secondary time constant τ of the induction motor 1₂(= L₂/
R₂Slip angular frequency command value (ω_s
^*) And the speed adaptive secondary magnetic flux observer (4, 7)
Estimated motor speed estimate (ω_r ^#) And obtained from
Power angular frequency (ω₀) Is the integrator 11 for calculating the basic phase angle.
The fundamental phase angle θ₀Seeking
Can be.

The delay compensating element 12 is provided to moderate the rising and falling of the slip angular frequency command value (ω _s ^* ) in order to match the control delay in the digital current controller 3.

As described above, the conventional speed sensor
The torque control method is based on the motor speed (ω_r)
Estimated motor speed
(Ω_r ^#) And slip angular frequency command value (ω_s ^*) And addition
Power source angular frequency ω obtained by₀(Ω ₀= Ω_r ^#+ Ω_s
^*) Deviates from the true value, resulting in a digital current
Power source angular frequency ω in controller 3₀Is calculated based on
Primary voltage command value v for decoupling control₁ ^*of
For displacement and coordinate conversion in the coordinate converters 8 and 9
Basic phase angle used₀Misalignment, eventually coordinate conversion axis misalignment
The vector control of decoupling control is no longer valid.
There is a risk of injury.

This means that the induction motor 1
In the speed control of 1), a serious problem that the torque according to the torque command value cannot be obtained occurs.

Therefore, the applicant of the present application previously applied for a speed sensorless vector control method capable of performing complete vector control by preventing deviation of the vector control coordinate axis which is a "lag" in the process of estimating the motor speed. (Japanese Patent Application No. 5-265976).

The technique of Japanese Patent Application No. 5-265976 filed previously is developed based on the following knowledge.

The "lag" in the estimation process of the estimated motor speed value (ω _r ^# ) of the induction motor 1 is the slip angular frequency command value (ω _s ^* calculated by the slip calculator 5 when viewed from the induction motor 1 side ^. ) Is offset.
(Equation (2) above, ω ₀ = ω _r + ω _s ^* , ω _s ^* = i _1b ^*
/ See i _1a ^* · τ ₂ )

Therefore, the electric power when the vector control is established is established.
Source angular frequency ω₀The requirement that determines
Estimated value of secondary magnetic flux λ on the mark (ab axis)₂ ^#Torque axis
Minute (λ_2b ^#) Is set to zero (see the above equations (5) and (6))
Therefore, the secondary torque axis magnetic flux estimated value λ_2b ^#Integral (ω_sc
= Kω_i∫λ_2b ^#・ Dt Kω_i: Integral gain)
Slip angular frequency correction value ω_scSlip angular frequency command value
ω_s ^*To (ω_s ^*+ Ω_sc) Allows the motor speed
Degree estimate ω_r ^#Power source angular frequency ω due to “delay”₀Nozomi
Re Δω₀Can be corrected and the coordinate axis is prevented from shifting.
And complete vector control is performed. (See the following formula) ω₀+ Δω₀= Ω_r ^#+ Ω_s ^*+ Ω_sc Δω₀= Ω_sc

FIG. 8 shows an embodiment of Japanese Patent Application No. 5-265976.

In the illustrated control system, the speed control section (ASR) is now instructed to drive the motor speed command value (ω _r ^* ).
Is given, the speed command value (ω _r ^* ) is compared with the motor speed estimated value (ω _r ^# ) which is a negative feedback signal, and the comparison error signal is proportional-integral (PI) controlled in the speed controller 6. , Primary current command value i on the synchronous rotation coordinates (ab axis)
Converted to ₁ ^* primary torque axis current command value (i _1b ^* ). Next, in the digital current controller 3 in the current controller (ACR), the primary torque axis current command value (i _1b ^* ) of the primary current command value i ₁ ^* and the primary excitation axis current command value (i _1a). ^* ) Is compared with the primary torque axis current detection value (i _1b ) of the primary current detection value i ₁ and the primary excitation axis current detection value (i _1a ), and i _1b ^* = i _1b , and i _1a ^* = I _{1a The} primary voltage command value v on the synchronous rotation coordinate axis (ab axis) that controls the PWM control inverter 2 so as to be controlled to
₁ ^* (v _1a ^* , v _1b ^* ) is calculated by the above equation (4).

The primary power output from the digital current controller 3
Pressure command value v₁ ^*(V_1a ^*, V_1b ^*) In the coordinate converter 9
Primary voltage command value v on the stator coordinates (d-q axes)₁ ^*
(V _1d ^*, V_1q ^*) Is converted to 2 phase-3 phase
The number of phases is converted by the converter 15, and the PWM control inverter
2 primary voltage control command voltage v for each of the three phases_u, V_v, V _w
To control the output voltage of the PWM control inverter 2.
As a result, the induction motor 1 has a desired motor speed command value.
(Ω_r ^*) According to the speed control.

The actual motor speed of the induction motor 1 (ω_r)
Is the estimated value of the motor speed estimated as follows.
ω_r ^#To use. That is, the stator coordinates (d-
Primary current detection value i on the q-axis)₁(I_1d, I_1q) Primary power
Pressure command value v₁ ^*(V_1d ^*, V _1q ^*) And motor speed estimation
Value (ω_r ^#) Input same-dimensional magnetic flux observer 4
And the primary estimated by the same-dimensional magnetic flux observer 4.
Current estimated value i₁ ^#(I_1d ^#, I_1q ^#) And the estimated secondary magnetic flux
λ₂ ^#(Λ_2d ^#, Λ_2q ^#) And the detected primary current i ₁(I
_1d, I_1qBased on the above equation (7)
A speed-adaptive secondary magnetic flux observer consisting of
The motor speed estimate ω_r ^#To estimate.

Then, the power source angular frequency ω ₀ (ω ₀ = ω _r ^# +) caused by the “lag” of the estimated motor speed (ω _r ^# ) in the process of estimating the actual motor speed (ω _r ).
In order to correct the deviation of ω _s ^* ), the secondary magnetic flux estimated value λ ₂ ^# (λ _2d ^# , λ _2q ^# ) on the stator coordinate (dq axis) estimated by the same-dimensional magnetic flux observer 4 is used as the coordinate. Converter 10
And the secondary magnetic flux λ ₂ ^# (λ
_2a ^# , λ _2b ^# ) and the secondary magnetic flux estimated value λ ₂ ^#
The secondary torque axis magnetic flux estimated value (λ _2b ^# ) is integrated by the slip angle frequency correction integrator 16 and the slip angular frequency correction value ω _sc
(Ω _sc = Kω _s ∫λ _2b ^# · dt) is obtained and added to the slip angular frequency command value ω _s ^* by the adder 17.

To add the slip angular frequency correction value ω _sc to the slip angular frequency command value ω _s ^* , the deviation of the motor speed estimated value (ω _r ^# ) is corrected by the slip angular frequency correction value ω _sc . Therefore, the basic phase angle (θ) due to the deviation of the power source angular frequency (ω ₀ ) determined by the above equation (2)
₀ ) is prevented, the coordinate conversion axis is prevented from being displaced, and accurate vector control of decoupling control is established.

[0030]

By the way, FIG. 7 and FIG.
In the previously proposed technique shown in (2), the following two problems remain.

<First Problem> First, the first problem will be described. Since it is difficult to detect the motor voltage of the induction motor 1 with the vector control device shown in FIGS. 7 and 8,
It is assumed that the primary voltage command value v ₁ ^* (v _1d ^* , v _1q ^* ) matches the actual motor voltage value. Based on this assumption, the same-dimensional observer 4 determines the primary voltage command value v ₁ ^*.
(V _1d ^* , v _1q ^* ), primary current detection value i ₁ (i _1d ,
i _1q ), motor speed estimated value ω _r ^# as input values, primary current estimated value i ₁ ^# (i _1d ^# , i _1q ^# ), secondary magnetic flux estimated value λ ₂ ^# (λ _2d ^# , λ _2q ^# ) Is estimated and calculated.

However, there is an error between the primary voltage command value v ₁ ^* and the actual motor voltage value (the cause of the error will be described below), and the error becomes larger especially in the low speed region where the output voltage is low. Due to this voltage error, the estimated motor speed ω _r
^An error occurs in ^# , and speed control becomes unstable.

Here, the two factors of the voltage error described above will be described.

<First Voltage Error Factor Causing First Problem> The main switching element used in the inverter part of the PWM control inverter 2 has a switching delay,
In order to prevent a short circuit between the upper and lower arms due to this delay time, a short circuit prevention period (dead time) is provided. Due to this dead time, the actual motor voltage value and the primary voltage command value v ₁ ^* differ. So now,
In order to reduce this voltage error, a dead time compensating circuit for delaying the output voltage of the PWM control inverter 2 according to the dead time is provided. However, complete compensation cannot be performed, and the voltage error remains.

<Second Voltage Error Factor Causing First Problem> In the PWM modulation command section of the PWM control inverter 2, the primary voltage control command voltages v _u , v _v , v _w and the carrier signal (for example, (Triangular wave) and PW depending on the size
The M modulation command (base current) is generated. This PWM
The main switching element of the inverter section is turned on / off according to the high / low of the modulation command to produce an equivalent sine wave output voltage. When the PWM modulation command section is digitally configured, the command voltages v _u , v _v , and v _w are digitally compared with the carrier signal to generate the PWM modulation command.
Due to the carrier wave resolution and the voltage calculation value resolution, an error occurs between the actual motor voltage value and the primary voltage command value v ₁ ^* .

<Second Problem> Next, the second problem will be described. In the vector control device shown in FIG. 8, the slip angular frequency command value ω _s ^* is added to the estimated motor speed value ω _r ^#, and the slip angular frequency correction value ω _sc is further added to obtain the power supply angular frequency ω _0. There is. By adding the slip angular frequency correction value ω _sc in this way, the calculation delay of the speed estimation is corrected and the power source angular frequency ω ₀ becomes a true _value .

By the way, the motor speed estimated value ω _r ^# fed back to the same-dimensional magnetic flux observer 4 is not a value corrected for the delay due to the speed estimation calculation but a value deviated from the true value. Therefore, the estimated motor speed value ω _r ^# output from the speed adaptation mechanism 7 remains deviated from the true value, and as a result, the output torque according to the torque command value may not be obtained.

In view of the above-mentioned prior art, it is an object of the present invention to provide a speed sensorless vector control device for an induction motor capable of accurately performing speed control and torque control.

[0039]

[Means for Solving the Problems] First to solve the above problems
In the configuration of the present invention, the primary current command value (i₁ ^*) And the primary power
Flow detection value (i₁) And power angular frequency (ω₀) And invite
Conductor motive current decoupling control is performed and primary voltage command value
(V₁ ^*) Output current controller (3) and the current control
The primary voltage command value (v₁ ^*) Two
Primary voltage control command voltage (v_u,
v_v, V_w) And the carrier signal, and
Then, the PWM operation is performed, and the PWM-controlled equivalent sine wave power
Control inverter to supply the induction motor (1)
(2) and the primary current detection value (i₁) And the primary voltage command value
(V₁ ^*) And the estimated motor speed (ω _r ^#) Each
The primary current estimated value i₁ ^#And secondary magnetic flux estimated value λ₂ ^#To
The same-dimensional magnetic flux observer (4) to be estimated and the same order
Estimated primary current i, which is the output of the original magnetic flux observer (4)
₁ ^#And secondary magnetic flux estimated value λ₂ ^#And the primary current detection value
(I₁) Respectively, and estimate the motor speed
(Ω_r ^#) Estimating and calculating) and outputting it (7)
And the primary current command value (i₁ ^*) Excitation axis component and torque axis
Based on the component, the slip angular frequency command value (ω_s ^*)
And a slip calculator (5) for calculating and outputting
Slip angular frequency command value that is the output of the output device (5)
(Ω_s ^*) To the estimated motor speed (ω_r ^#) Before adding
The power supply angular frequency (ω which is the control input of the current controller (3)
₀), An adder (17) for outputting
In the vector control device without speed sensor of the machine, the P
The WM control inverter (2) outputs the estimated motor speed value ω_r ^#
Is larger than the set value, the carrier signal with high frequency is
Primary voltage control command voltage (v_u, V_v, V_w)
And estimate the motor speed ω_r ^#Is less than the set value
When it becomes low, use the carrier signal with a low frequency to
Pressure control command voltage (v_u, V_v, V _w) And PW
It is characterized by performing M operation.

In the configuration of the second invention, in the PWM control inverter (2), the first set value used when the estimated motor speed value ω _r ^# decreases and the estimated motor speed value ω _r ^#. Is set, and a second set value used when the value rises is set.

In the third configuration of the present invention, the current controller (3) outputs the primary voltage command value (v ₁ ^* ) immediately before switching at the control timing immediately after switching the carrier signal, and thereafter. 1 was calculated by control timing
It is characterized in that the next voltage command value (v ₁ ^* ) is output.

The configuration of the fourth invention is that the primary current command value
(I₁ ^*) And the primary current detection value (i₁) And power angular frequency
(Ω₀) To control the current decoupling of the induction motor.
No primary voltage command value (v₁ ^*) Output current controller
(3) and the primary voltage which is the output of the current controller (3)
Command value (v₁ ^*) Based on the electric power to control the speed of the induction motor
The converter (2) and the primary current detection value (i₁) And primary voltage finger
Command value (v₁ ^*) And the estimated motor speed (ω _r ^#)
Input the estimated primary current i₁ ^#And secondary magnetic flux estimated value λ₂
^#Same-dimensional magnetic flux observer (4) for estimating
Estimation of the primary current output from the one-dimensional magnetic flux observer (4)
Value i₁ ^#And secondary magnetic flux estimated value λ₂ ^#And the primary current detection value
(I₁) Respectively, and estimate the motor speed
(Ω_r ^#) Estimating and calculating) and outputting it (7)
And the output of the same-dimensional magnetic flux observer (4)
Secondary torque axis magnetic flux estimated value λ_2d ^#Coordinate conversion
Estimated value of secondary torque axis magnetic flux λ on synchronized rotation coordinates_2b ^#To
Integral calculation and slip angular frequency correction value ω_scAn integral that outputs
Device (16) and the primary current command value (i₁ ^*) Excitation axis component
And the slip angle frequency command value of the motor based on the torque axis component
(Ω_s ^*) And a slip calculator (5)
Slip angular frequency command output from slip calculator (5)
Value (ω_s ^*) To the estimated motor speed (ω_r ^#)
Power source angular frequency which is the control input of the current controller (3)
(Ω₀And an adder (17) for outputting
In the speed sensorless vector controller of the electric motor,
Motor speed estimated value ω_r ^#Slip angular frequency correction value ω_sc
Corrected motor speed estimate ω_r ^#’
The one-dimensional magnetic flux observer (4) includes the motor speed estimation
Value ω_r ^#Instead of the calculated modified motor speed estimate ω_r
^#′ Inputting an adder (12)
It

[0043]

The first cause of the voltage error, which causes the first problem, is the compensation error of the delay time of the main switching element. Therefore, in the present invention, by lowering the frequency of the carrier wave signal and increasing the pulse width of the voltage sent from the PWM control inverter 2 to the induction motor 1, it is possible to apparently reduce the error component contained in this voltage. The cause of the voltage error of 1 can be solved.

The operation of the present invention for eliminating the second cause of the voltage error that causes the first problem will be described. When the PWM calculation unit is composed of digital circuits, the PWM after PWM modulation
The resolution of the modulation command is determined by the command voltages v _u , v _v , v _w and the resolution of the carrier signal. However, generally, the carrier signal has a lower resolution than the command voltage. The reason for this is that the carrier wave is often configured in hardware, and its resolution greatly depends on the clock frequency of the digital circuit that generates the carrier wave. On the other hand, the control clock frequency has an upper limit in terms of circuit operation, and infinite frequency increase is not allowed. Also, I do not want to change the hardware due to changes in the clock frequency.

In order to solve this, the present invention proposes a method of reducing the carrier frequency to equivalently increase the resolution of the carrier signal. This makes it possible to equivalently improve the resolution of the carrier signal and improve the command voltage resolution after PWM modulation. This allows the actual output voltage and the command voltage to match as much as possible.

After all, the first and second causes of the voltage error, which cause the first problem, can be eliminated by lowering the frequency of the carrier wave signal of the PWM calculation section. However, when the motor is operated at high speed, the carrier frequency is preferably as high as possible. Therefore, in the present invention, the control is performed at a low carrier frequency only in the low speed region where the voltage error is a problem, and is controlled at the high carrier frequency in the high speed operation range.

In the actual control, the magnetic flux observer calculation, the speed estimation calculation and the current control calculation are performed in synchronization with the half cycle of the carrier wave, and these must be controlled according to the control cycle. Therefore, simply changing the carrier frequency causes a control problem. So in the example,
Even if the frequency of the carrier wave signal changes, these arithmetic units can be controlled stably.

In order to solve the second problem, in the present invention, the estimated value of the motor speed ω _r ^# is added to the slip angular frequency correction value ω.
By inputting the corrected motor speed estimated value ω _r ^{# '} to which _sc is input to the same-dimensional magnetic flux observer, accurate magnetic flux observer calculation can be performed and stable torque control can be performed.

[0049]

Embodiments of the present invention will be described below in detail with reference to the drawings. The same parts as those of the conventional technique are designated by the same reference numerals, and duplicate description will be omitted.

<First Embodiment> FIG. 1 shows a first embodiment of the present invention. The first embodiment is different from the conventional technique shown in FIG.
PW the estimated motor speed value ω _r ^# obtained by the speed adaptation mechanism 7.
Change to send to M control inverter 2 and PWM
The functions of the control inverter 2, the digital current controller 3, the same-dimensional magnetic flux observer 4, and the basic phase angle calculating integrator 11 are changed. These changes will be sequentially described.

First, referring to FIG. 2, the PWM in the first embodiment
The control inverter 2 will be described. The carrier wave signal generation unit 2a can generate two kinds of carrier wave signals (triangular wave signals) H _H and H _L , and one of the carrier wave signals H _H and H _L is received according to a command from the carrier wave signal change command unit 2b. PWM one
It is sent to the modulation command unit 2c. The carrier signal H _H has a high frequency as in the conventional case, and the carrier signal H _L has a low frequency. The amplitude of H _L is H _H
It is designed to be larger than the amplitude of.

The carrier signal change command section 2b has a first base.
Quasi value V_ref1And this first reference value V _ref1Greater than value
Second reference value V_ref2Is remembered. Carrier signal change
The update command unit 2b uses the estimated motor speed value ω_r ^#Is descending
When estimated, ω_r ^#And the first reference value V_ref1Compare with
ω_r ^#<V_ref1From the carrier signal generator 2a
The carrier wave signal to output is H_HTo H_LSwitch to. this
The switching point is a carrier wave (triangular wave) signal as shown in FIG.
At the top of. On the other hand, the estimated motor speed ω_r ^#Rises
Estimated value ω_r ^#And the second reference value V_ref2And ratio
Compare ω_r ^#> V _ref2Becomes, the carrier wave signal generation unit 2a
The carrier wave signal output from_LTo H_HSwitch to.
This switching point is the carrier wave (triangular wave) as shown in FIG.
At the top of the signal. Carrier wave signal H_H, H_LChange at the top of
The reason why the operation cycle timing is_H,
H_LThis is because it is the top of
The signals can be changed relatively easily by and. Two more
The standard value of is used to prevent chattering.
You can

The PWM modulation command section 2c compares the primary voltage control command voltages v _u , v _v , v _w with one of the carrier signals H _H , H _L , and outputs the PWM modulation command I to the inverter 2
send to d. At this time, the operation clock of the PWM modulation command section 2c is the same as the conventional one.

The inverter 2d operates according to the PWM modulation command I and sends an equivalent three-phase alternating current to the induction motor 1.

As described above, the PWM in the first embodiment
In the control inverter 2, the estimated motor speed value ω_r ^#Is the first
Reference value V_ref1When it gets smaller and becomes extremely slow
Is a low frequency carrier signal H_LUses PWM modulation
The operation clock cycle in the command unit 2c is the same as the conventional one.
However, the apparent PWM calculation resolution is improved. Therefore
Estimating the actual voltage value output from the inverter 2d
Primary voltage command value v₁ ^*The voltage error with
Calculation in the same dimension magnetic flux observer 4 is more accurate even in the range
This enables the more accurate speed estimation calculation.

When the estimated motor speed value ω _r ^# becomes larger than the second reference value V _ref2 , the carrier signal H _H having the same high frequency as in the conventional case is used, so that the accurate control at the time of high speed operation is ensured. be able to.

Next, the digital current controller 3 of the first embodiment will be described.
Details will be described with reference to FIG. This digital current control
The device 3 has a primary current command value i₁ ^*(I_1a ^*, I_1b ^*) And 1
Next current detection value i₁(I_1a, I_1b) Is equal to
Primary voltage command value v₁ ^*(V_1a ^*, V_1b ^*) Is calculated and output
Force This calculation cycle is the carrier signal H (H_H, H_LNou
The signal is selected every half cycle. Immediately
In FIG. 4, the multiplication units 3-11, 3-12, 3-13,
3-14 is an excitation axis voltage integral gain G₁₁, Excitation axis voltage ratio
Example gain G₁₂, Multiplication value ω₀L₁, Primary resistance R₁Multiplication
The subtraction units 3-15 and 3-16 perform subtraction calculation and add.
The section 3-17 performs addition calculation, and the integration section 3-18 performs integration calculation.
Then, the addition / subtraction unit 3-19 performs addition / subtraction calculation. And add
The part 3-17 and the integrating part 3-18 work together to perform an equivalent integration operation.
And the integral term YO is
Add the integral term PRYO (this is the previous value of the integral term YO)
Calculated and calculated. Similarly, the multiplication units 3-21, 3-22,
3-23 and 3-24 are torque axis voltage integral gain G_{twenty one},
Torque shaft voltage proportional gain G_{twenty two}, Multiplication value ω₀Lσ, primary resistance
Anti-R₁And the subtraction unit 3-25 performs subtraction calculation,
The adders 3-26 and 3-27 perform addition operations and perform integration 3-2.
8 performs integral calculation, and the adder / subtractor 3-29 performs adder / subtractor operation.
U Then, the adding unit 3-27 and the integrating unit 3-28 work together.
Equivalent integral calculation is performed, and the integral term YT is calculated by the multiplication unit 3-
21 output and the previous integral term PRYT (this is the integral term YT
It is calculated by adding (previous value). R₁, Ω
₀L₁, Ω₀The value obtained by multiplying Lσ becomes the interference term.

In the digital current controller 3 shown in FIG.
The secondary excitation axis voltage command value V _1a ^* and the primary torque axis voltage command value V _1b ^* are obtained by the following equations (8) and (9).

V _1a ^* = YO + i _1a ^* · R ₁ −i _1a · G ₁₂ −i _1b ^* · ω ₀ Lσ (8) V _1b ^* = YT + i _1b ^* · R ₁ −i _1b · G ₂₂ + i _1a ^* · omega is ₀ L ₁ ... (9) where ^{_{YO = (i 1a * -i 1a}} ) · G 11 + PRYO ... (10) YT = (i 1b * -i 1b) · G 21 + PRYT ... (11). Thus ^{_{YO = V 1a * -i 1a *}} · R 1 + i 1a · G 12 + i 1b * · ω 0 Lσ ... (12) YT = V 1b * -i 1b * · R 1 + i 1b · G 22 -i 1a *・ Ω ₀ L ₁ (13)

The following can be understood by comparing the above equations (4) and (8) and (9). That is, the primary current command value i ₁ ^*
(I _1a ^* , i _1b ^* ) and the primary current detector i ₁ (i _1a ,
i _1b ), the influence of the feedback term disappears, and if the differential term (SLσ) in equation (4) is ignored, equations (4) and (8) and (9) are in agreement. I understand that

In the digital current controller 3 shown in FIG.
Period is synchronized with half cycle of carrier signal H
When the frequency of the carrier signal H is switched to
Gain G according to the cycle (calculation cycle)₁₁, G_12,G_{twenty one,}G
_{twenty two}Need to be converted. Moreover, the gain G₁₁, G_12,G
_{twenty one,}G_{twenty two}Output voltage (voltage command value V _1a
^*, V_1b ^*) Is devised to prevent sudden changes (see below).

In the current controller 3 of this embodiment, since the current control is performed digitally, the gain is also calculated in a discrete system. Therefore, the gains G ₁₁ , G _21, change depending on the sine pulling time (calculation cycle for calculation). Also,
Since the proportional gain G _12, G ₂₂ does not change theoretically, actually uses a value such that the control is follow-up of stably performed and the current control better maximize the experiments, varied by the carrier frequency ing.

In the digital current controller 3, the carrier signal is
Gain G when switching₁₁, G _12,G_{twenty one,}G_{twenty two}The same
Sometimes switch. When I wanted to take any measures at this time
The operating state will be described. Explaining the excitation axis, the gain G₁₂
And primary excitation axis current detection value i_1aThe proportional term i multiplied by_1a・ G₁₂
Is the current detection value i immediately before and after switching the carrier signal.
_1aIs almost unchanged (actual sampling
Since the period is short, there is little change in one sample time), gain
G₁₂The change is large due to the change. For example, gay
G₁₂Is doubled, the proportional term i_1a・ G₁₂Doubles.
However, the integral term YO is the detected value i_1aAnd command value i_1a ^*Difference
(Actually, the current is controlled, so the value is small)
G₁₁Is the sum of the product of
, There is little change immediately before and after switching the carrier signal.
Yes. As a result, the integral term YO and the proportional term i_1a・ G₁₂And the interference term
Excitation axis voltage command value V obtained from_1a ^*Is the proportional term i_1a・ G
₁₂Abrupt changes in. Torque axis voltage command value
V_1b ^*Also suddenly changes, and as a result, when switching carrier signals
The control becomes unstable and operation is impossible.

As described above, current control is performed by IP control.
If the gain is switched when the carrier signal is switched,
The proportional term changes greatly, while the integral terms YO and YT change.
Voltage command value V_1a ^*, V_1b ^*Sudden change
I will end up. Immediately before and after switching the carrier signal, output to the motor.
The voltage that must be applied is almost unchanged (the
The pulling time is extremely short). Therefore, this embodiment
In the current controller 3 of, only once after switching the carrier signal
Is the same as the voltage immediately before switching (primary excitation voltage command V_1a
^*And the primary torque voltage command V_1b ^*Is assumed to remain unchanged) and
Assuming this time voltage command value V_1a ^*, V_1b ^*And its assumption
Voltage command value and current command value i_1a ^*, I _1b ^*And current detection
Value i_1a, I_1bExcitation and torque integral terms (Y in the figure)
O and YT) is calculated and the calculation is performed from the second time.
The current is controlled by IP control using the integrated term
U With this method, a large change in the proportional term is reflected in the integral term.
Because it can be projected, the primary voltage command value V_1a ^*, V_1b ^*Is
Does not change suddenly. The formula for calculating the integral term is the current control
The system is obtained from the equations (12) and (13) described by equations.

Since this is done, when the current control cycle changes
Stable voltage command value v_1a ^*, V _1b ^*To calculate
Therefore, the output current when changing the carrier is
Stable control is possible without delaying even one sample
Current change does not occur.

Next, the same-dimensional magnetic flux observer of the first embodiment
The operation of No. 4 will be described. This same-dimensional magnetic flux observer 4
Performs discrete-time operation, and also performs half-cycle of carrier signal frequency
Since the sampling period is the period of time, the carrier wave
Carrier frequency (= observer calculation frequency when changing signals)
It is necessary to perform the calculation method according to the period. So same next
When the carrier wave signal H is changed, the original magnetic flux observer 4
Previous primary voltage command value v _1d ^*, V_1q ^*And this primary current
Detected value i_1d, I_1qIs used to estimate the primary current i₁ ^#And 2
Estimated secondary magnetic flux λ₂ ^#Estimate calculation. In this way
Estimated value i calculated by constant calculation₁ ^#, Λ₂ ^#Enters speed adaptation mechanism 7
Forced motor speed estimate ω_r ^#Is estimated and calculated.

That is, as shown in FIG. 5, the operation constants used in the same-dimensional magnetic flux observer 4 are changed and made effective two operation cycles after the carrier wave signal is changed. In this way, the observer calculation is performed up to the calculation cycle after two samples in consideration of the calculation cycle before changing the carrier wave, and therefore the estimation of the primary current estimation value i ₁ ^# and the secondary magnetic flux estimation value λ ₂ ^# by the discrete time system The calculation can be executed accurately. Therefore, the estimated value i ₁ ^# ,
The estimated motor speed value ω _r ^{# obtained} by the speed adaptation mechanism 7 using λ ₂ ^# becomes a continuous value, and the carrier signal can be changed smoothly.

On the other hand, the phase calculation of the output frequency performed by the basic phase angle detecting integrator 11 needs to be calculated according to the carrier frequency. When the carrier signal is changed, the integrator 11 that performs the phase calculation of the output frequency is
Up to the calculation cycle after sampling, the phase calculation is performed in consideration of the cycle before changing the carrier, so that the output voltage phase when changing the carrier can be calculated accurately. By thus continuously changing the output voltage phase when changing the carrier wave, a transient current does not flow in the induction motor 1. Therefore, the carrier wave can be changed without a transient change in the motor torque.

<Second Embodiment> Next, referring to FIG. 6, the present invention will be described.
A second embodiment will be described. In this second embodiment, as shown in FIG.
The following points differ from the related art shown. That is, slip
Slip angular frequency correction obtained by the integrator 16 for angular frequency correction
Value ω_scAnd the estimated motor speed ω obtained by the speed adaptation mechanism 7
_r ^#And adder 21 to add corrected motor speed estimated value
ω_r ^#′ Is obtained, and the corrected motor speed estimated value ω_r ^#'same
It is input to the one-dimensional magnetic flux observer 4. Still other parts
Is the same as the prior art shown in FIG.₀
= Ω_r ^#+ Ω_sc+ Ω_s ^*Has become.

As described in the prior art of FIG. 8, the slip angular frequency correction value ω _sc is calculated so that the secondary torque axis magnetic flux estimated value λ _2b ^# becomes zero, which is the same as the theoretical value shown in the equation (5). It is a thing. Therefore, the corrected motor speed estimated value ω _r ^# 'input to the same-dimensional magnetic flux observer 4 becomes a true value in which the delay due to the speed estimation calculation is corrected. Therefore, the speed estimation calculation can be accurately performed, and the torque according to the torque command value can be output from the induction motor 1.

[0071]

According to the first aspect of the invention, when the induction motor is rotated in the low speed range, the frequency of the carrier signal used in the PWM control inverter is lowered, so that the apparent PW is obtained.
The M calculation resolution is increased, and accurate speed control can be performed.

According to the second aspect of the invention, since hysteresis is provided in the set value, chattering does not occur when the carrier signal is switched, and stable operation can be performed.

According to the third aspect of the invention, the change of the primary voltage command value is suppressed even when the carrier signal is switched, so that the carrier signal can be stably switched.

According to the fourth invention, the same-dimensional magnetic flux observer is provided with the corrected motor speed estimated value ω with the calculation delay corrected.
_{Since r} ^# '(= ω _r ^# + ω _sc ) is input, there is no error in the speed estimation calculation, and accurate torque control can be performed.

[Brief description of drawings]

FIG. 1 is a block configuration diagram showing a first embodiment of the present invention.

FIG. 2 is a block diagram showing a PWM control inverter used in the first embodiment.

FIG. 3 is a waveform diagram showing a carrier signal used in the first embodiment.

FIG. 4 is a block diagram showing a digital current controller used in the first embodiment.

FIG. 5 is a time chart showing a calculation schedule in the first embodiment.

FIG. 6 is a block diagram showing a second embodiment of the present invention.

FIG. 7 is a block diagram showing a conventional technique.

FIG. 8 is a block diagram showing a conventional technique.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 induction motor 2 PWM control inverter 2a carrier wave signal generation part 2b carrier wave signal change command part 2c PWM modulation command part 3 digital current controller 4 same dimension magnetic flux observer 5 slip calculator 6 speed controller 7 speed adaptation mechanism 8 coordinate converter 9 Coordinate converter 10 Coordinate converter 11 Integrator for basic phase angle calculation 16 Integrator for slip angle frequency correction 17 Adder 21 Adder ω _s Slip angular frequency ω _s ^* Slip angular frequency command value ω _sc Slip angular frequency correction value ω ₀ power supply angular frequency omega _r motor speed omega _r ^* motor speed command value omega _r ^# motor speed estimated value omega _r ^# 'modified motor speed command value omega _{r max} motor speed maximum value v 1 ₁ primary voltage v _1a 1 primary excitation axis voltage v _1b 1 primary torque-axis voltage v ₁ ^* 1 primary voltage command value v _1a ^*, v _1d ^* 1 primary excitation axis voltage command value v _1b ^*, v _1q ^* 1 primary torque-axis voltage command value i _u, i _v, i _w 1 Current i ₁ 1 primary current detected values i _1a, i _1d 1 primary excitation axis current detection value i _1b, i _1q 1 primary torque axis current detection value i ₁ ^* 1 primary current command value i _1a ^* 1 primary excitation axis current command value i _1b ^* 1 primary torque axis current command value i ₁ ^# 1 primary current estimated value i _1d ^# 1 primary excitation axis current estimated value i _1q ^# 1 primary torque axis current estimated value lambda ₂ 2 rotor flux λ _2a, λ _2d 2-order Excitation axis flux λ _2b , λ _2q Secondary torque axis flux λ ₂ ^# Estimated value of secondary flux λ _2a ^# , λ _2d ^# Estimated value of secondary excitation axis flux λ _2b ^# , λ _2q ^# Estimated value of secondary torque axis flux v _u , v _v , v _w Primary voltage control command voltage θ ₀ Basic phase angle H _H , H _L Carrier wave signal I PWM modulation command

Claims (4)

[Claims] 1. A primary current command value (i₁ ^*) And primary current detection
Outgoing price (i₁) And power angular frequency (ω₀) And enter induction
The primary voltage command value (v₁
^*), And a primary voltage command value (v) that is the output of the current controller (3).
₁ ^*) Is a two-phase three-phase conversion primary voltage control command voltage
(V_u, V_v, V_w) With the carrier signal and compare
The PWM operation is performed according to the result, and the PWM controlled equivalent positive
PWM control in that supplies chord wave power to the induction motor (1)
The barter (2) and the primary current detection value (i₁) And the primary voltage command value (v₁ ^*)When
Estimated motor speed (ω _r ^#) For each
Flow estimate i₁ ^#And secondary magnetic flux estimated value λ₂ ^#Estimate the same
Dimensional magnetic flux observer (4) and the primary power output from the same dimensional magnetic flux observer (4)
Flow estimate i₁ ^#And secondary magnetic flux estimated value λ₂ ^#And primary current detection
Outgoing price (i₁), And estimate the motor speed (ω
_r ^#) And a speed adaptation mechanism (7) that estimates and outputs the primary current command value (i₁ ^*) Excitation axis component and torque axis component
Based on the slip angular frequency command value (ω_s ^*Play)
A slip calculator (5) that calculates and outputs, and a slip angular frequency finger that is an output of the slip calculator (5).
Command value (ω_s ^*) To the estimated motor speed (ω_r ^#) Is added
Power source angular frequency which is the control input of the current controller (3)
(Ω₀) Output adder (17), and an induction motor speed sensorless vector control device comprising:
In the above arrangement, the PWM control inverter (2) is provided with an estimated motor speed value.
ω_r ^#Is larger than the set value, the
The primary voltage control command voltage (v_u, V_v,
v_w) And estimate the motor speed ω_r ^#Is the set value
If it becomes smaller, use a carrier signal with a low frequency
Primary voltage control command voltage (v_u, V_v, V _w)
Speed of induction motor characterized by PWM operation
Sensorless vector controller. 2. In the PWM control inverter (2),
Estimated motor speed ω _r ^#To be used when
1 set value and motor speed estimated value ω_r ^#Is rising
It is special that the second setting value used at the time is set.
The speed sensor-less vector of the induction motor according to claim 1.
Control device. 3. The current controller (3) outputs the primary voltage command value (v ₁ ^* ) immediately before switching at the control timing immediately after switching the carrier signal, and calculates it at the subsequent control timing. Primary voltage command value (v ₁ ^* )
The speed sensorless vector control device for an induction motor according to claim 1, wherein 4. A primary current command value (i₁ ^*) And primary current detection
Outgoing price (i₁) And power angular frequency (ω₀) And enter induction
The primary voltage command value (v₁
^*), And a primary voltage command value (v) that is the output of the current controller (3).
₁ ^*), A power converter that controls the speed of the induction motor
(2) and the primary current detection value (i₁) And the primary voltage command value (v₁ ^*)When
Estimated motor speed (ω _r ^#) For each
Flow estimate i₁ ^#And secondary magnetic flux estimated value λ₂ ^#Estimate the same
Dimensional magnetic flux observer (4) and the primary power output from the same dimensional magnetic flux observer (4)
Flow estimate i₁ ^#And secondary magnetic flux estimated value λ₂ ^#And primary current detection
Outgoing price (i₁), And estimate the motor speed (ω
_r ^#) And a velocity adaptation mechanism (7) for estimating and outputting
Secondary torque axis magnetic flux estimated value λ_2d ^#Coordinate conversion
Secondary torque axis magnetic flux estimated value λ on synchronous rotation coordinates_2b ^#Integrate
Calculated slip angular frequency correction value ω_scIntegrator that outputs
(16) and the primary current command value (i₁ ^*) Excitation axis component and torque axis component
Based on the slip angular frequency command value (ω_s ^*Play)
A slip calculator (5) that calculates and outputs, and a slip angular frequency finger that is an output of the slip calculator (5).
Command value (ω_s ^*) To the estimated motor speed (ω_r ^#) Is added
Power source angular frequency which is the control input of the current controller (3)
(Ω₀) Output adder (17), and an induction motor speed sensorless vector control device comprising:
, The estimated motor speed value ω_r ^#Slip angular frequency correction value ω
_scCorrected motor speed estimate ω_r ^#’
The same-dimensional magnetic flux observer (4) includes the motor speed
Constant value ω_r ^#Instead of the calculated modified motor speed estimate ω
_r ^#'Is provided with an adder (12) for inputting
Speed sensorless vector control device for induction motor.