JP3054521B2 - Induction motor control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an induction motor controller for performing vector control.

[0002]

2. Description of the Related Art FIG. 5 is a block diagram showing a configuration of a conventional induction motor control device. In FIG. 5, IM is an induction motor main body, SS is a power converter, CT is a current detector, SF
Is a slip frequency calculator, CAL is a voltage command value calculator, C
₁ and C ₂ are comparators, AD ₁ and AD ₂ are adder / subtracters, Gr
(S) is a speed control compensation circuit, Gq (s) is a torque current control compensation circuit, VEC-1 and VEC-2 are coordinate converters,
NT is an integration circuit, and ROM1 and ROM2 are memory tables.

The power converter SS supplies three-phase AC power with a variable voltage and a variable frequency to the motor IM.
There is a pulse width modulation control (PWM) inverter and the like.

The primary current i of the motor is determined by the current detector CT.
u, iv, and iw are detected and input to the coordinate converter VEC-1. The current detection values iu, iv, iw are converted from the stationary coordinates into the rotating coordinates by the coordinate converter VEC-1,
It becomes the torque current detection value iq and the excitation current detection value id.

The slip frequency calculator SF estimates the slip angular frequency ωs ^* of the induction machine IM by performing the following calculation using the torque current detection value iq and the excitation current detection value id. ωs ^* = (R ₂ iq) / (L ₂ id) where R ₂ and L ₂ are a secondary resistance value and a secondary inductance value of the induction machine IM.

[0006] Speed control circuit is constituted by a comparator C ₁ and the control compensating circuit Gr (s), and outputs the command value iq ^* of the torque current. That later detected rotational speed .omega.r illustrating the comparison by the comparator C ₁ the command value .omega.r ^*, amplifies the deviation εr = ωr ^* -ωr by the control compensation circuit Gr (s), the torque current command value iq ^* .

[0007] The torque current control circuit is composed of a comparator C ₂ and the control compensating circuit Gq (s), and outputs the primary angular frequency command value omega ₁ ^* of the induction machine IM. That is, the comparison by the comparator C ₂ a torque current command value iq ^* and the torque current detection value iq, the deviation εq = iq ^* -iq control compensation circuit Gq (s) is amplified primary angular frequency ω in Seeking ₁ ^* .

The difference between the primary angular frequency command value ω ₁ ^* and the slip angular frequency command value ωs ^* is calculated by an adder / subtracter AD ₁ to obtain a rotational angular frequency ωr ＾ = ω ₁ ^* −ωs ^* .
Feedback is provided to the speed control circuit.

The primary angular frequency ω ₁ ^* is integrated by an integrator INT, and a phase angle θ ^* = ∫ω ₁ ^* dt is obtained and input to the memory table ROM ₁ . The memory table ROM1 generates two-phase unit sine wave signals sin θ ^* and cos θ ^* for the input phase angle θ ^* , and supplies the signals to the coordinate converter VEC-1.

In the coordinate converter VEC-1, the three-phase current i
The detection values of u, iv, and iw are converted into currents id and iq of dq coordinates (rotation coordinates). That is, id = cos θ · iα−sin θ · iβ iq = sin θ · iα + cos θ · iβ where iα = k · (iu−iv / 2−iw / 2) iβ = k ′ · (iv−iw) k = 2/3, k '= 1/2. Generally, id is called a direct axis current (excitation current), and iq is called a horizontal axis current (torque current).

On the other hand, the excitation current command value id ^* , the torque current command value iq ^*, and the primary angular frequency ω ₁ ^* of the induction machine are input to the calculator CAL, and the following calculation is performed. The IM voltage command value v ₁ ^* and the phase angle δ ^* are obtained. ^{_{v 1 * = a {e +}} R 1 'iq * + ω 1 * σL 2 id *}
cos δ ^* -a ｛R ₁ 'id ^* -ω ₁ ^* σL ₂ iq ^* ｝ sin
δ ^* δ ^* = tan ^-1 ω (ω ₁ ^* σL ₂ iq ^* −R ₁ ′ i
^{d *) / (e + R} 1 'iq * + ω 1 * σL 2 id *)} _{However, σ = (L 1 L 2} / M-1) a = M / L 2 e = ω 1 * L 2 id * = ω ₁ ^* φ ₂ R ₁ ′ = R ₁ / a ² where R ₁ , L ₁ : primary resistance value of induction machine, primary inductance R ₂ , L ₂ : secondary resistance value of induction machine, secondary inductance M: Mutual inductance of the induction machine φ ₂ : Secondary magnetic flux of the induction machine.

[0012] Thus the phase angle determined by [delta] ^* is input to the adder-subtracter AD _2, is added to the output theta ^* of the integrator. The memory table ROM ₂ stores the sum of the phase angles θ ^*.
With the input of + δ ^* , a three-phase unit sine wave is created, and the signal is given to the second coordinate converter VEC-2.

The voltage value V ₁ ^* is converted from the rotating coordinates to the stationary coordinates by the second coordinate converter VEC-2, and three-phase AC voltage command values vu ^* , vv ^* , vw ^* are obtained. That is, vu ^* = V ₁ ^* sin (θ ^* + δ ^* ) vv ^* = V ₁ ^* sin (θ ^* + δ ^* −2π / 3) vw ^* = V ₁ ^* sin (θ ^* + δ ^* + 2π / 3) ).

The power converter SS generates a voltage proportional to the voltage command value, and supplies a voltage necessary for vector control to the induction motor IM.

FIG. 6 shows a voltage-current vector diagram of the induction motor IM at this time. Here, there is a relationship of v ₁ ′ = v ₁ / a and i ₁ ′ = a · i ₁ . The generated torque of the motor is proportional to the product of the exciting current id (that is, the secondary magnetic flux φ ₂ ) and the torque current iq, and the primary current vector i _{1 at} that time.
Is determined. Further, the internal induced voltage e of the motor is determined by the exciting current id and the primary angular frequency ω _{1 as} described above, and the voltage e is calculated by adding the primary resistance drop and the drop due to the primary and secondary leakage inductances. By this addition, the vector of the primary voltage v _{1 to} be applied to the electric motor is determined. The computing unit CAL performs the above-described computation so as to satisfy this relationship.

With such a conventional induction motor control device, vector control can be performed without a speed sensor.
Characteristics equivalent to those of a DC motor can be obtained.

[0017]

In the conventional induction motor control device described above, the primary voltage command values (v ₁ ^* and δ ^* ) required for vector control are generated by the calculator CAL.
In this calculation, motor constants (R ₁ , M, L ₂ , etc.) are used. If these constants are accurate values, the magnetic flux and the torque current of the motor always maintain a right angle position, and the DC motor It is possible to obtain characteristics equivalent to the above.

However, the constants of the induction machine differ depending on the type of motor and the output rating, and it is difficult to know an accurate value. Normally, calculations are performed using design values, but it remains questionable whether accurate vector control is realized. As a countermeasure, a veteran adjuster performs a test operation and adjusts the motor constant to a value considered to be optimal while measuring steady-state characteristics and the like. This adjustment requires a considerable amount of time and money, and general-purpose motors with a large number of motors can only make rough adjustments.
There have been problems such as a decrease in output and characteristics.

Since the primary resistance R ₁ of the induction machine changes due to a temperature rise during operation, and the mutual inductance M and the secondary inductance L ₂ change due to saturation of the iron core, it is assumed that an accurate constant is first inserted. Constant changes during operation,
The vector of the magnetic flux φ ₂ and the torque current iq is not maintained at a right angle, and problems such as a decrease in the motor generated torque occur.

On the other hand, the primary angular frequency ω ₁ ^* is
This is a correct value because it matches the output angular frequency of S. However, the slip angular frequency ωs ^* obtained by the arithmetic unit SF
Is calculated using the motor constants (R ₂ , L ₂ ), an error occurs if the values deviate. Therefore, the calculation speed ωr ＾ = ω ₁ fed back to the speed control circuit.
^* −ωs ^* also has an error, and accurate speed control cannot be performed. In particular, the secondary resistance R ₂ of the motor varies with the temperature of the rotor, there is a problem such that the motor speed during operation comes deviated from the set value.

The present invention has been made in view of the above circumstances, and aims at non-adjustment of motor constants used for vector control calculation, and even when the motor constants change during operation, the above-mentioned calculation is performed in accordance with the change. An object is to provide an induction motor control device capable of automatically adjusting a constant.

[0022]

Means for Solving the Problems As means for solving the above problems, a first invention is directed to an exciting current command value id ^* , a torque current command value iq ^* , and a primary angular frequency command value ω for an induction motor. obtains the voltage command value v ₁ ^* and a phase angle command value [delta] ^* with ₁ ^*, based on the voltage command value v ₁ ^* and a phase angle command value [delta] ^*, and supplies a variable voltage variable frequency AC power to the induction motor Output voltage command value v for power converter
In an induction motor control device that performs vector control on an induction motor by outputting u ^* , vv ^* , and vw ^* ,
A detection value of an output voltage and an output current from the power converter is input, and an excitation current operation value id ＾ and a torque current operation value iq ＾ are output based on an operation or estimation of a secondary magnetic flux of the induction motor. Magnetic flux calculating means, an exciting current calculated value idｉ from the magnetic flux calculating means and a torque current calculated value i
q ＾, the excitation current command value id ^* , the torque current command value iq ^* , and the primary angular frequency command value ω ₁ ^*, and while learning based on the back propagation algorithm, the two output signals A *, a neural network means for outputting a B *, the neural network hand
The voltage command value v based on the output signals A * and B * from the stage
₁ ^* and means for calculating the phase angle command value δ ^* .

According to a second aspect, in the configuration of the first aspect, a detection value of an output current from the power converter is input,
A coordinate converter that outputs an excitation current detection id and a torque current detection value iq based on the conversion of the detected value from the stationary coordinate system to the rotation coordinate system; and an excitation current detection value id and a torque current from the coordinate converter. And a slip frequency calculator for calculating the slip frequency command value ωs ^* based on the input of the detection value iq.

According to a third aspect, in the configuration of the first aspect, a detection value of an output current from the power converter is inputted.
Based on the conversion of the detected value from the stationary coordinate system to the rotating coordinate system,
A coordinate converter for outputting a torque current detection value iq;
Means for outputting a secondary magnetic flux command value Φ2 * of the induction motor;
A torque current detection value iq from the coordinate converter, and
The slip frequency command value ω based on the secondary magnetic flux command value φ ₂ ^*
and a slip frequency calculator for calculating s ^* .

According to a fourth aspect of the present invention, there is provided excitation of an induction motor.
Current command value id ^* , torque current command value iq ^* , primary angular circumference
Voltage command value v ₁ ^* and phase angle using wave number command value ω ₁ ^*
A command value δ ^* is obtained, and the voltage command value v ₁ ^* and the phase angle
Based on decree value [delta] ^*, exchange of the variable voltage variable frequency induction motor
Output voltage command value v for the power converter that supplies the
By outputting u ^* , vv ^* , vw ^* ,
In an induction motor control device that performs vector control,
The detected values of the output voltage and output current from the power converter
Input and calculate the secondary magnetic flux of the induction motor or
Based on the estimation, the excitation current calculation value id ＾ and the torque current
Magnetic flux calculating means for outputting a calculated value iq ＾, and the induction motor
Means for outputting a secondary magnetic flux command value Φ2 * of
Input the command value φ ₂ ^* and the secondary magnetic flux calculation value φ ₂ 、,
Learning based on back propagation algorithm
No while, front for outputting the excitation current command value id ^*
Neutral network means, and the magnetic flux calculating means
Excitation current calculation value id ＾ and torque current calculation value iq
＾, the exciting current command value id ^* , the torque current
Ordinance value iq ^*, 1 primary angular frequency command value ω ₁ ^* Enter the, back
Perform learning based on the propagation algorithm
While the latter stage that outputs two output signals A * and B *
Neural network means and the subsequent neural network
Based on the output signals A * and B * from the network means.
Means for calculating pressure command value v ₁ ^* and phase angle command value δ ^*
And the learning result of the neural network means at the preceding stage
And the torque current detection value from the coordinate converter
iq and the secondary magnetic flux command value Φ2 *,
A slip frequency calculator for calculating the wave number command value ωs ^* ,
It is provided with a configuration.

According to a fifth aspect of the present invention, in any one of the first to fourth aspects,
In the configuration of the invention described above, the neural network means performs learning based on the back propagation algorithm only when the rotation speed of the induction motor is equal to or higher than a predetermined level.

[0027]

As an input signal of the neural network means, first, an excitation current command value id ^* and a torque current command value i
q ^* and the primary angular frequency command value ω ₁ ^* are used. The excitation current command value id ^* is usually given as a constant value, but when performing field weakening control or the like in accordance with the rotation speed of the electric motor, the command value is changed. The torque current command value iq ^* may be directly given, but generally the output from the speed control circuit is the torque current command value. The primary angular frequency command value ω ₁ ^* is given from the torque current control circuit.

On the other hand, a voltage / current supplied from the power converter to the induction motor is detected, and the detected value is inputted to a magnetic flux calculator which is a magnetic flux calculation means, and an excitation current id ＾ of the motor is obtained.
And the torque current iq # are calculated. Note that a magnetic flux observer may be used instead of the magnetic flux calculator as the magnetic flux calculator. The excitation current calculation value id # and the torque current calculation value id # are the other input signals of the neural network means.

The output signals of the neural network means are the primary voltage V ₁ ^* and the phase angle δ ^* of the motor, expressed in rotational coordinates. The primary voltage command value V ₁ ^{* is set} to the stationary coordinates (3
(Phase AC voltage command value), and a voltage proportional to the command value is generated from the power converter, thereby performing vector control of the induction motor.

That is, the neural network means includes an exciting current command value id ^* and a torque current command value iq.
^* A primary voltage command value of the motor v ₁ ^* in which the attaching Conclusion, serves conventional calculator CAL.

The coupling coefficient of the neural network means is learned by back propagation (back propagation). That is, first, an appropriate engagement coefficient is inserted, and the excitation current command value id ^* , the torque current command value iq ^*, and 1
A primary voltage command value v ₁ ^* and a phase angle command value δ ^* are derived from the secondary angular frequency command value ω ₁ ^* , and the motor is operated. An excitation current calculation value id # and a torque current calculation value iq # of the motor are calculated from the voltage and current at that time, and the values are held. Next, using the excitation current calculation value id ＾, the torque current calculation value iq ＾, and the primary angular frequency command value ω ₁ ^* , the voltage command value v ₁ ^* and the phase angle command value δ ^* via the neural network means ^. Ask for ＾. Then, the voltage command value v ₁ ^* and the phase angle command value δ ^* ＾ and the previously obtained v ₁
The coupling coefficient of the neural network means is changed according to the difference between ^* and δ ^*, and learning is repeated so that the difference finally becomes zero.

Finally, id ^* = id ＾, iq ^* = iq
＾, v ₁ ^* = v ₁ ^* ＾, δ ^* = δ ^* ＾, and settle down. At this time, assuming that the excitation current calculation value id ＾ and the torque current calculation value iq ＾ obtained by the magnetic flux calculator are sufficiently accurate, the voltage command value v ₁ ^* and the phase angle δ obtained by the neural network means are obtained. ^* Can be expected to be sufficiently accurate, and as a result, vector control with the same output characteristics as DC machines can be achieved.

In general, the accuracy of the excitation current calculation value id # and the torque current calculation value iq # by the magnetic flux calculator deteriorates when the rotation speed of the motor is low. Therefore, learning of the neural network means is performed when the rotation speed is high. It is preferable to temporarily stop the learning when the rotation speed becomes low.

Even when the temperature of the rotor rises during operation, the coupling coefficient of the neural network means is updated correspondingly, and the vector control of the induction motor can always be performed in an optimum state. . The same applies when the iron core of the motor is saturated.

As described above, the induction motor control device of the present invention does not require a seasoned adjuster, automatically selects the vector control constant to the optimum value, and is resistant to disturbances such as temperature rise. Vector control can be achieved.

Next, another neural network means is added to the neural network means, and as input signals, a secondary magnetic flux command value φ ₂ ^* and a torque current command value iq ^*.
And the case where the primary angular frequency ω ₁ ^* is used. The secondary magnetic flux command value φ ₂ ^* is normally given as a constant value, but when performing field weakening control or the like in accordance with the rotation speed of the electric motor, the command value is changed. The torque current command value iq ^* may be directly given, but generally the output from the speed control circuit is the torque current command value. The primary angular frequency command value ω ₁ ^* is given from the torque current control circuit.

On the other hand, the voltage / current supplied from the power converter to the induction motor is detected, and the detected value is input to the magnetic flux calculator, and the secondary magnetic flux calculation value φ ₂の of the motor, the excitation current calculation
The value id # and the torque current calculation value iq # are calculated. Note that a magnetic flux observer may be used instead of the magnetic flux calculator. The secondary magnetic flux calculation value φ ₂ ＾, the excitation current calculation value id ＾
And the torque current operation value iq ＾ are the other input signals of the neural network means and the other neural network means.

The other neural network means is connected to the input side of the neural network means, and alternately inputs a secondary magnetic flux command value φ ₂ ^* and a secondary magnetic flux calculation value φ ₂ 、 to generate an exciting current command value id ^*. Is output. That is, the other neural network means on the preceding stage connects the secondary magnetic flux command value φ ₂ ^* and the exciting current command value id ^* , and the coupling coefficient is learned by back propagation (back propagation).

The neural network means on the subsequent stage comprises:
The excitation current command value id ^* , the torque current command value iq ^*, and the primary angular frequency ω ₁ ^* which are the output signals of the other neural network means on the preceding stage, and the excitation current calculation value idｉ
And the torque current calculation value iq # are alternately input to output a primary voltage command value v ₁ ^* and a phase angle command value δ ^* of the electric motor represented by the rotation coordinates. That is, the latter-stage neural network means connects the excitation current command value id ^* and the torque current command value iq ^* with the primary voltage command value v ₁ ^* and the phase angle command value δ ^*, and the coupling coefficient is back. Learn by propagation.

The primary voltage command value v ₁ ^* is converted to the stationary coordinates (3
(Phase AC voltage command value), and a voltage proportional to the command value is generated from the power converter, thereby performing vector control of the induction motor.

The coupling coefficients of the neural network means and other neural network means are learned by back propagation.

First, an appropriate value is entered as the excitation current command value id ^* , and the coupling coefficient of the neural network means on the subsequent stage is learned. That is, an appropriate coupling coefficient is inserted in the neural network means on the subsequent stage, and the excitation current command value id ^* is used to convert the torque current command value iq ^* and the primary angular frequency command value ω ₁ ^* to the primary voltage command value v ₁ ^*. And the phase angle command value δ ^* are issued, and the motor is operated. From the voltage and current at that time, the excitation current calculation value idｉ of the motor and the torque current
An operation value iq ＾ is calculated and the value is held.

Next, using the excitation current calculation value id ＾, the torque current calculation value iq ＾ and the primary angular frequency ω ₁ ^* , the voltage command value v ₁ ^* ＾ and the phase Obtain the angle command value δ ^* ＾. And
The voltage command value v ₁ ^* and the phase angle command value δ ^* ＾ and the engagement coefficient of the neural network means on the subsequent stage are changed according to the respective differences between v ₁ ^* and δ ^* obtained earlier,
Finally, learning is repeated so that the difference becomes zero.

Finally, id ^* = id ＾, iq ^* = i
q ＾, v ₁ ^* = v ₁ ^* ＾, δ ^* = δ ^* ＾, and settle down.

Next, the coupling coefficients of the other neural network means on the preceding stage are learned. That is, an appropriate coupling coefficient is inserted into the other neural network means on the preceding stage, an exciting current command value id ^* is output from the secondary magnetic flux command value φ ₂ ^*, and the primary voltage is output via the neural network means on the subsequent stage. The command value v ₁ ^* and the phase angle command value δ ^* are issued, and the motor is operated. The secondary magnetic flux φ ₂の of the motor is calculated from the voltage and current at that time, and the value is held. Next, the excitation current command value id ^* is calculated using the secondary magnetic flux calculation value φ ₂ ＾ ^.
Ask for ＾. Then, according to a difference between the excitation current command value id ^* ^ and previously determined excitation current command value id ^* by changing the coupling coefficient of the other neural network means of the front side, eventually the difference is zero Repeat learning to become. Finally, φ ₂ ^* = φ ₂ ＾, id ^* = id ＾
Calm down.

In this manner, all the coupling coefficients of the other neural network means on the preceding stage and the neural network means on the subsequent stage are learned, and the secondary magnetic flux calculation value φ ₂求 め obtained by the magnetic flux calculator and the excitation current calculation value id ＾
If the calculated torque current value iq ＾ is a sufficiently accurate value, the voltage command value v ₁ ^* and the phase angle command value δ ^* obtained by the neural network means can be expected to be sufficiently accurate values. Vector control with the same output characteristics as the machine can be achieved.

Even if the temperature of the rotor rises during operation, the coupling coefficient of the other neural network means is updated correspondingly, and the vector control of the induction motor can always be performed in an optimum state. It becomes possible. The same applies when the iron core of the motor is saturated.

The engagement coefficient learned by the preceding neural network means is related to the secondary resistance of the electric motor and can be reflected as an operation constant of the slip frequency operation. Usually, the slip frequency calculation using the secondary resistance value R _2, etc., when the temperature of the rotor rises, R ₂ is changed, not be determined the correct slip frequency. Therefore,
Accurate speed control cannot be expected with speed sensorless vector control that estimates the rotation speed of the motor using the slip frequency. Therefore, the coupling coefficient learned by the other neural network means on the preceding stage is substituted into the slip frequency calculator to perform the calculation. As a result, the motor constant is learned according to the temperature rise during operation, and accurate speed control can be performed.

As described above, according to the induction motor control apparatus of the present invention, the vector control constant is automatically selected to the optimum value without the need for a veteran adjuster, and the disturbance such as temperature rise is prevented. Strong vector control can be performed.

[0050]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS. FIG. 1 is a block diagram showing the configuration of the first embodiment.

In FIG. 1, IM is an induction motor main body, S
S is a power converter, CT is a current detector, PT is a voltage detector, FCAL is a magnetic flux calculator as magnetic flux calculator,
C ₁ and C ₂ are comparators, Gr (s) is a speed control compensation circuit,
Gq (s) is a torque current control compensation circuit, VEC-1, V
EC-2 is a coordinate conversion circuit, SF is a slip frequency calculator,
AD ₁ and AD ₂ are adders, INT is an integration circuit, ROM
1, ROM2 is a memory table, NNW is a neural network circuit, BP is a back propagation (back propagation) circuit, and NCAL is an external arithmetic unit.

Then, the neural network means includes:
It comprises a neural network circuit NNW, a back propagation circuit BP, and an external arithmetic unit NCAL.

The power converter SS supplies three-phase AC power having a variable voltage and a variable frequency to the motor IM.
There are a pulse width modulation control (PWM) inverter, a cycloconverter, and the like.

As input signals to the neural network NNW, an excitation current command value id ^* and a torque current command value iq
^{* And the} primary angular frequency command value ω ₁ ^* are used.

The excitation current command value id ^* is usually given as a constant value, but when performing field weakening control or the like in accordance with the rotation speed of the motor, the command value is changed.

The torque current command value iq ^* may be given directly, but generally the output from the speed control circuit is the torque current command value.

The speed control circuit comprises a comparator C ₁ and a control compensation circuit Gr (S), and outputs a torque current command value iq ^* . That is, compared by the comparator C ₁ speed detection value .omega.r determined by calculating ^ and the command value .omega.r ^*,
The deviation εr = ωr ^* −ωr is calculated by the control compensation circuit Gr (S).
To give the torque current command value iq ^* to the neural network circuit NNW.

The torque current control circuit includes a comparator C ₂ and a control compensation circuit Gq (S), and includes a primary angular frequency command value ω ₁
Output ^* . The ω ₁ ^* is supplied to the neural network NNW and is also input to the integrator INT to generate a phase angle signal θ ^* = ∫ω ₁ ^* dt for coordinate transformation. Memory table ROM1 will make the phase angle signal theta ^* from 2 phase unit sine wave sin θ ^* · cos θ ^*, it gives the coordinate converter VEC-1.

On the other hand, from the power converter SS to the induction motor IM
Are supplied to the coordinate converter VEC-1 by detecting the currents iu, iv, iw supplied to the coordinate converter VEC-1.

In the coordinate unit VEC-1, the three-phase currents iu, i
The detected values of v and iw are converted to the current i of the dq coordinates (rotational coordinates).
Convert to d, iq. That is, id = cos θ ^* · iα−sin θ ^* · iβ iq = sin θ ^* · iα−cos θ ^* · iβ where iα = k · (iu−iv / 2−iw / 2) iβ = k ′ · (Iv-iw) k = ２, k ′ = １ ／ Generally, id is called a direct axis current (excitation current), and iq is called a horizontal axis current (torque current).

[0061] torque current detection value iq are fed back to the comparator C ₂ of the torque current control circuit.

The torque current detection value iq and the excitation current detection value id are input to the slip frequency calculator SF, and the following calculation is performed to determine the slip angular frequency ω _s ^* . ω _s ^* = (R ₂ · iq) / (L ₂ · id) Further, the terminal voltage Vu, Vv, Vw of the motor is detected by the voltage detector PT, and the magnetic flux is detected together with the three-phase current detection values iu, iv, iw. Input to the arithmetic unit FCAL. Magnetic flux calculator FCA
L is the secondary magnetic flux calculation value φ of the motor using the voltage and current.
₂ ＾, excitation current calculation value id ＾ and torque current calculation value i
Calculate q ＾.

First, the detected three-phase (U, V, W) voltage / current is coordinate-converted into two-phase (α, β-phase) voltage / current. That is, iα = k · (iu-iv / 2−iw / 2) iβ = k ′ · (iv−iw) and eα = k · (Vu−Vv / 2−Vw / 2) eβ = k ′ · (Vv −Vw). Here, k = ２ and k ′ = １ ／.

The secondary magnetic fluxes φ ₂ α and φ ₂ β in the β coordinate system can be obtained by performing the following calculations. _{φ 2 α = (L 2 /} M) (eα-R 1 · iα) dt-σ ·
L ₁ · (L ₂ / M) · iα φ ₂ β = (L ₂ / M) (eβ-R ₁ · iβ) dt-σ ·
L ₁ · (L ₂ / M) · iβ where R ₁ and R ₂ are primary and secondary resistances, and L ₁ and L ₂ are 1
Next, secondary inductance, M is mutual inductance, σ
= 1−M ² / (L ₁ · L ₂ ) is a leakage coefficient.

The secondary magnetic flux φ ₂ α, φ ₂ β in the αβ coordinate system
Is converted to the dq coordinate system, the following calculation is performed. φ _2d = cos ψ · φ ₂ α + sin ψ · φ ₂ β φ _2q = −sin ψ · φ ₂ α + cos ψ · φ ₂ β where cos ψ = φ ₂ α / (φ ₂ α ² + φ ₂ β ² ) sin _{ψ = φ 2 β / (φ} 2 α 2 + φ 2 β 2) holds becomes the ^{_{φ 2d = (φ 2 α 2}} + φ 2 β 2) φ 2q = 0. φ _2d is the calculated value of the secondary magnetic flux φ ₂ ＾.

Further, the current operation values id ＾, i in the dq coordinate system
q ＾ is obtained as follows using the above cos ψ and sin ψ. id ＾ = cosψ · iα + sinψ · iβ iq ＾ = − sinψ · iα + cosψ · iβ That is, id ＾ is an excitation current calculation value, and iq ＾ is a torque current calculation value. The excitation current calculation value id # and the torque current calculation value iq # are the other input signals of the neural network NNW.

Signals A ^* and B ^* for obtaining the output voltage command value V ₁ ^* and the phase angle command value δ ^* of the power converter SS are selected as output signals of the neural network NNW. The external arithmetic unit NCAL selects the voltage command value V ₁ ^* from the signals A ^* and B ^* by performing the following operation. The external arithmetic unit NCAL performs the following operation to obtain the signal A
^* , B ^* , voltage command value V ₁ ^* and phase angle command value δ ^*
Ask for. ^{δ * = tan -1 (B *} / A *) V 1 * = A * · cos δ * -B * · sin δ * the phase angle [delta] ^* is inputted to the adder AD _2, the output of the integrator INT Added to the signal θ ^* . The phase angle signal (θ ^* + δ ^* ) is input to the memory table ROM2,
Create a three-phase unit sine wave sin (θ ^* + δ ^* ) sin (θ ^* + δ ^* -2π / 3) sin (θ ^* + δ ^* + 2π / 3).

The coordinate converter VEC-2 converts the voltage command value V1 ^* into a three-phase AC voltage command value Vu ^* , Vv ^* ,
Vw ^* , which is given to the power converter SS. Vu ^* = V ₁ ^* · sin (θ + δ ^* ) Vv ^* = V ₁ ^* · sin (θ + δ ^* −2π / 3) Vw ^* = V ₁ ^* · sin (θ + δ ^* + 2π / 3) Command values Vu ^* , Vv ^* , Vw
A voltage proportional to ^* is generated to drive and control the induction motor IM.

FIG. 2 shows a neural network circuit NNW.
FIG. 3 is a block diagram showing the configuration of FIG. In FIG. 2, SW ₁
To SW ₄ is switched circuits, ML1, ML2 multiplier, X ₁ ~
X ₄ input elements, Y ₁ and Y ₂ are output elements (addition elements), W
₁₁ to _W-42 is the coupling coefficient.

The inputs X _{1 to} X ₄ are connected to switch circuits SW ₁ ,
When SW ₂ is connected to the a side, X ₁ = iq ^* , X ₂ =
id ^* , X ₃ = ω ₁ ^* · id ^* , X ₄ = ω ₁ ^* · iq ^*
And when connected to the b side, X ₁ = iq ＾, X ₂ =
id ＾, X ₃ = ω ₁ ^* · id ＾, X ₄ = ω ₁ ^* · iq ＾
Becomes

The coupling coefficients W _{11 to} W ₄₂ are (4 × 2) =
Eight are conceivable, but the minimum required four are described here.

Next, the neural network circuit N
The operation of the learning function of the NW will be described. First, connect the switch circuits SW ₁ to SW ₄ to a side, put an appropriate coupling coefficient W ₁₁ to _W-42 to the neural network circuit NNW, excitation current command value id ^*, the torque current command value iq ^* and the primary angular frequency command value omega ₁ ^* from the output signal A ^*, B ^*
The out voltage command value V ₁ ^* and position via the external operation unit NCAL
The phase angle δ ^* is obtained, and the motor IM is operated.

[0073] That is, the output element Y ₁ signal A ^* and signal output element Y ₂ B ^* is as follows. _{^{A * = iq * · W 11}} + ω 1 * · id * · W 31 B * = id * · W 22 + ω 1 * · iq * · W 42 U 1 This value back Pro holidays circuit BP =
A ^* , U ₂ = B ^* . Further, the voltage command value V ₁ ^* and the phase angle command value δ ^* are sent to the external arithmetic unit NCAL.
It is calculated more as:. δ ^* = tan ⁻¹ (B ^* / A ^* ) V ₁ ^* = A ^* · cos δ ^* −B ^* · sin δ ^* The motor IM is performed using the voltage command value V ₁ ^* and the phase angle command value δ ^*. Are input to the excitation calculator FCAL by inputting the voltages Va, Vb, Vc and the currents ia, ib, ic when
The excitation current calculation value id # and the torque current calculation value iq # of the motor are calculated by the above-described method, and the calculated values are stored as inputs to the neural network NMW.

Next, the switch circuits SW _{1 to} SW ₄ are connected to the b side, and the neural network circuit NM is used by using the excitation current calculation value id ＾ and the torque current calculation value iq ＾.
Output signals A ^{* *} and B ^* ＾ are obtained via W. A ^* ＾ = iq ＾ · W ₁₁ + ω ₁ ^* · id ＾ · W ₃₁ B ^* ＾ = id ＾ · W ₂₂ + ω ₁ ^* · iq ＾ · W ₄₂ This value is applied to the back-propagation circuit BP and V ₁ = A
^* ＾ and V ₂ = B ^* ＾. Then, the coupling coefficients W _{11 to} W ₄₂ of the neural network are changed in accordance with the respective differences between the A ^* ＾ and B ^* ＾ and the previously obtained A ^* and B ^*, and finally, the differences become zero. Repeat learning to become. That is, the difference δ ₁ ,
If δ ₂ is δ ₁ = U ₁ −V ₁ = A ^* −A ^* ＾ δ ₂ = U ₂ −V ₂ = B ^* −B ^*補正, the correction amount △ W _{11 of the} coupling coefficients W _{11 to} W ₄₂ ~ W ₄₂
The correction amount ΔW ₄₂ of ΔW ₁₁ = K ₁ · δ ₁ · X ₁ ΔW ₂₂ = K ₂ · δ ₂ · X ₂ ΔW ₃₁ = K ₁ · δ ₁ · X ₃ ΔW ₄₂ = K ₂・ It will be δ ₂ · X ₄ . Here, K ₁ and K ₂ are learning gains, and X ₁ = iqｎ (n): iq ＾ at sample time X ₂ = id ＾ (n): id at sample time ＾ X ₃ = ω ₁ ^*・ id ＾ (n ): Ω ₁ ^* · i at the time of sample
d ＾ X ₄ = ω ₁ ^* · iq ＾ (n): ω ₁ ^* · i at the time of sampling
q ＾. Therefore, the coupling coefficient W ₁₁ to _{W-42 is, W 11 (k + 1)} = W 11 (k) + △ W 11 W 22 (k + 1) = W 22 (k) + △ W 22 W 31 (k + 1) = W 31 ( k) + ΔW ₃₁ W ₄₂ (k + 1) = W ₄₂ (k) + ΔW ₄₂

For example, when A ^* > A ^*} and δ ₁ becomes positive, the coupling coefficients W ₁₁ and W ₃₁ decrease, and the signal A ^* used in actual operation decreases. As a result, learning is performed so that A ^* = A ^* ＾.

When B ^* > B ^*} and δ ₂ becomes positive, the coupling coefficients W ₂₂ and W ₄₂ decrease, and the signal B ^* used in actual operation decreases. As a result, learning is performed so that B ^* = B ^* ＾.

Finally, id ^* = id ＾, iq ^* = i
q ＾, A ^* = A ^* ＾, B ^* = B ^* ＾ and calm down.
At this time, if the excitation current calculation value id 演算 and the torque current calculation value iq ＾ obtained by the magnetic flux calculator FCAL are sufficiently accurate values, the neural network circuit NN
The signals A ^* and B ^* obtained by W, that is, the voltage command value V ₁ ^* and the phase angle command value δ ^* can be expected to be sufficiently accurate. As a result, vector control having output characteristics equivalent to those of a DC machine can be achieved. Can be achieved.

However, the accuracy of the excitation current calculation value id ＾ and the torque current calculation value iq ＾ by the magnetic flux calculator FCAL generally deteriorates when the rotation speed of the motor is low. It is better to do it occasionally, and to suspend the learning when the rotation speed becomes low.

Even if the temperature of the rotor rises during operation and the primary resistance value R ₁ changes, the coupling coefficient of the neural network is updated correspondingly. Similarly, even when the core of the motor is saturated and the mutual inductance M is changed, the coupling coefficient of the neural network is automatically updated, and the vector control of the induction motor can always be performed in an optimal state.

Next, a second embodiment of the present invention will be described with reference to FIGS.
It will be described based on FIG. In FIG. 3, IM is an induction motor main body, SS is a power converter, CT is a current detector, PT is a voltage detector, FCAL is a magnetic flux calculator, C ₁ and C ₂ are comparators, and Gr.
(S) is a speed control compensation circuit, Gq (S) is a torque current control guarantee circuit, VEC-1 and VEC-2 are coordinate conversion circuits,
SFN is a slip frequency calculator, AD ₁ and AD ₂ are adders, INT is an integration circuit, ROM 1 and ROM 2 are memory tables, NN 1 and NN 2 are neural network circuits, BP 1 and BP 2 are back propagation (back propagation) circuits, NCAL Is an external arithmetic circuit.

Then, the neural network circuit NN
2, a back propagation circuit BP2 and an external operation circuit NCAL constitute a neural network means on the subsequent stage, and a neural network circuit N connected to the input side of the neural network means, that is, the preceding stage.
N1 and the back propagation circuit BP1 constitute "another neural network means".

In the second embodiment, only the parts different from the first embodiment will be described. The magnetic flux calculator FCAL outputs a secondary magnetic flux calculation value φ ₂ ＾ in addition to the excitation current calculation value id ＾ and the torque current calculation value iq ＾. This secondary magnetic flux performance
Calculated values phi ₂ ^ the exciting current calculated value id ^ and torque current Starring
What is obtained when calculating the arithmetic value iq ＾ is described above.

The neural network circuit NN2
Is the same as the neural network NNW in FIG. The back propagation circuit BP2 is the same as the back propagation circuit BP in FIG.

FIG. 4 shows other network means on the preceding stage,
FIG. 3 is a block diagram showing a network unit on the subsequent stage. 4, the neural network circuit of the first-stage NN1, front side of the back-propagation circuit BP1, the subsequent stage of the backpropagation circuit BP2, SW ₁ to SW ₆ is a switch circuit, x _1, x ₂ and X ₁ to X ₄ are the input element, y ₁ and Y _1, Y ₂ is output element (adding element), w _11, w ₂₁ and W ₁₁ to _W-42 is the coupling coefficient, S is a differential element, ML1, ML2 a multiplier is there.

Neural network circuit NN on the subsequent stage
2 is the same as the neural network circuit NNW in FIG.

Next, the learning operation of the neural network circuit NN1 at the preceding stage will be described. However, for convenience, exit the learning of the neural network circuit NN2 second-stage, the coupling coefficient W ₁₁ to _W-42 is described as being converged into the optimum value.

The neural network circuit NN on the preceding stage
1 inputs the secondary magnetic flux command value φ ₂ ^* and the secondary magnetic flux calculation value φ ₂ ＾, and outputs the exciting current command value id ^* . The exciting current command value id ^* becomes one of the input signals of the neural network circuit NN2 on the subsequent stage.

Input elements x ₁ and x ₂ are connected to switch circuit S
When W ₅ is connected to a ^{_{side, x 1 = φ 2 *,}} x
₂ = △ φ ₂ ^* / △ t, and when connected to the b side, X ₁ = φ ₂ ＾ and X ₂ = △ φ ₂ ＾ / △ t.

First, appropriate values are inserted into the coupling coefficients w ₁₁ and w ₂₁ of the NN 1 to switch the switch circuits SW ₅ and SW ₆ to a
To the side. At this time, the output y ₁ is given by the following equation. y ₁ = x ₁ · w ₁₁ + x ₂ · w ₂₁ = φ ₂ ^* · w ₁₁ + (△ φ ₂ ^* / △ t) · w ₂₁ The output y ₁ is held as u ₁ = y _{1 by the} back propagation circuit BP1, and becomes an excitation current command value id ^* which is one of the input signals of the neural network circuit NN2 on the subsequent stage. Here, assuming that the learning of the neural network circuit NN2 on the subsequent stage has already been completed, iq ^* = iq ＾, id ^* = id
Ｕ, U ₁ = V ₁ and U ₂ = V ₂ . In this state, the motor IM is driven, and the secondary magnetic flux calculation value φ ₂ ＾ is obtained from the magnetic flux calculator FCAL. This value φ ₂ } is input to the neural network circuit NN1 on the preceding stage.

Next, when the switch circuits SW ₅ and SW ₆ are connected to the b side, the output y ₁ is given by the following equation. That is, y ₁ ＾ = x ₁ · w ₁₁ + x ₂ · w ₂₁ = φ ₂ ＾ · w ₁₁ + ({φ ₂ ＾ / △ t) · w ₂₁ . The output v ₁ = y ₁ ^ is input to the back-propagation circuit BP1, wherein u ₁ difference between [delta] ₁₁ =
u ₁ −v _{1 is} determined. At this time, the coupling coefficients w ₁₁ and w
₂₁ of the correction amount △ w _11, △ a w _21, and _{△ w 11 = k 1 · δ} 11 · x 1 △ w 21 = k 1 · δ 11 · x 2. Here, k ₁ is a learning gain, and x ₁ = φ ₂ ＾ (n): φ ₂ ｘx ₂ = {φ ₂ ＾ (n) / △ t: sampled value of φ ₂微分 at the time of sample X 3 = ω ₁ ^* · id ＾ (n): ω ₁ ^* · i at the time of sampling
d ＾. Therefore, the coupling coefficients w ₁₁ and w ₂₁ are corrected as follows: w ₁₁ (k + 1) = w ₁₁ (k) + △ w ₁₁ w ₂₁ (k + 1) = w ₂₁ (k) + △ w ₂₁

For example, if u ₁ > v ₁ and δ ₁₁ becomes positive, the coupling coefficients w ₁₁ and w ₂₁ decrease, and the signal u ₁ used in actual operation decreases. As a result, learning is performed so that u ₁ = v ₁ .

Finally, φ ₂ ^* = φ ₂ ＾, u ₁ = v ₁
Calm down. At this time, if the secondary magnetic flux calculation value φ ₂求 め obtained by the magnetic flux calculator FCAL is a sufficiently accurate value, the signal y = id ^* obtained by the neural network circuit NN1 is also expected to be a sufficiently accurate value. it can. As a result, if vector control is performed on the induction motor IM via the neural network circuit NN2 on the subsequent stage using the excitation current command value id ^* , an output characteristic equivalent to that of a DC machine can be obtained.

In the above description, the learning of the neural network circuit NN1 of the preceding stage has been described on the assumption that the learning of the neural network circuit NN2 of the succeeding stage has been completed first. However, the neural network circuits NN1 of the preceding stage and the succeeding stage have been described. , NN2 can simultaneously learn.

When the learning is completed and the vector control is operating as theoretically, the relationship between the secondary magnetic flux command value φ ₂ ^* and the exciting current command value id ^* is expressed by the following equation. id ^* = φ ₂ ^* · (1 / M) + (dφ ₂ ^* / dt) ·
(L ₂ / (R ₂ · M)) That is, the coupling coefficients w ₁₁ and w ₂₁ of the NN1 converge to the following values. w ₁₁ = (1 / M) w ₂₁ = (L ₂ / (R ₂ · M)) where M is the excitation inductance, and L ₂ and R ₂ are the secondary inductance and the secondary resistance.

On the other hand, the slip frequency calculator SF of FIG. 3 inputs the torque current detection value iq and the secondary magnetic flux command value φ ₂ ^* ,
The slip angular frequency ω _s ^* is calculated using the following equation.

Ω _s ^* = (M · R ₂ / L ₂ ) · (iq / φ ₂ ^* ) The motor constant (M · R ₂ / L ₂ ) used here was obtained by learning of the neural network circuit NN 1. which is the inverse of the coupling coefficient w _21. Therefore, the learning result of the neural network circuit NN1 is reflected on the calculation constant of the slip frequency calculator SFN, and the slip angle frequency command value ω _s ^* is calculated as follows. ω _s ^* = (1 / w ₂₁ ) · (iq / φ ₂ ^* ) As a result, the secondary resistance value R ₂
Even if it changes, learning is done correspondingly,
The slip frequency can always be calculated with an accurate value. In other words, the estimated speed ωrr = ω of the motor IM
_An accurate value is always obtained for ₁ ^* −ω _s ^* , and accurate speed control can be performed without a normal mechanical speed detector.

In the above, the magnetic flux calculator FCAL was used to obtain the secondary magnetic flux calculation value φ ₂ ＾ or the excitation current calculation value id ＾ and the torque current calculation value iq ＾. Instead, a magnetic flux observer was used. It goes without saying that the same can be achieved even if used. Further, the above-described neural network circuit has been described by taking a two-layer linear model as an example, but a model having three or more layers may be used.
L may be incorporated in the network to directly output the voltage command value v ₁ ^* and the phase angle command value δ ^* .

[0098]

As described above, according to the present invention, a vector control constant is automatically selected to an optimum value without requiring a seasoned adjuster, and a vector which is resistant to disturbances such as temperature rise. Control can be performed. Further, accurate speed control can be performed without using a speed detector.

[Brief description of the drawings]

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

FIG. 2 is a block diagram showing details of a neural network circuit in FIG. 1;

FIG. 3 is a block diagram showing a configuration of a second embodiment of the present invention.

FIG. 4 is a block diagram showing details of a neural network circuit in FIG. 3;

FIG. 5 is a block diagram showing a configuration of a conventional device.

FIG. 6 is a voltage-current vector diagram for explaining the operation of FIG. 5;

[Explanation of symbols]

IM induction motor SS power converter CT current detector PT voltage detector FCAL flux calculator C _1, C ₂ comparator Gr (S) speed control compensation circuit Gq (S) torque current control compensation circuit VEC-1, VEC-2 Coordinate conversion circuit AD1, AD2 Adder INT Integration circuit ROM1, ROM2 Memory table SF, SFN Slip frequency calculator NNW, NN1, NN2 Neural network circuit SW _{1 to} SW ₆ Switch circuit BP, BP1, BP2 Back propagation (back propagation) Circuit NCAL External computing unit

Continuation of the front page (56) References JP-A-5-207776 (JP, A) Kazuo Shimane et al., "Vector Control of Induction Machines Using Neural Networks", Proceedings of the Institute of Electrical Engineers of Japan, Proceedings of the Institute of Electrical Engineers of Japan Association, 1992, No. 6, p. 6-13 to 6-14 Jong-Whan et al., "AN ADAPTIVE LEARNING CURRENT CONTROLLER FOR FILE-ORIENTED ED COUTROLED INDUCTION MOTION BY NEO RION NETWORK, 1991. , p. 469-474 (58) Fields investigated (Int. Cl. ⁷ , DB name) H02P 21/00 G05B 13/02 G06F 15/18 H02P 5/408-5/412 H02P 7/628-7/632 JICST file ( JOIS)

Claims (5)

(57) [Claims] 1. An exciting current command value id for an induction motor.
^* , Torque current command value iq ^* , primary angular frequency command value ω ₁
^* Calculated voltage command value v ₁ ^* and a phase angle command value [delta] ^* with, based on the voltage command value v ₁ ^* and a phase angle command value [delta] ^*, power supplies variable voltage variable frequency AC power to the induction motor Output voltage command values vu ^* , vv ^* ,
In an induction motor control device that performs vector control on an induction motor by outputting vw ^* , a detection value of an output voltage and an output current from the power converter is input, and calculation or estimation of a secondary magnetic flux of the induction motor is performed. Magnetic flux calculating means for outputting an exciting current calculation value id ＾ and a torque current calculation value iq ＾ based on the magnetic field, an excitation current calculation value id ＾ and a torque current calculation value iq ＾ from the magnetic flux calculation means, and the excitation current Command value id
^* , Torque current command value iq ^* , primary angular frequency command value ω ₁
^* , A neural network means for outputting two output signals A *, B * while performing learning based on the back propagation algorithm, and an output signal A *,
The voltage command value v ₁ ^* and the phase angle command value based on B *
and a means for calculating δ ^* . 2. A detection value of an output current from the power converter is input, and based on a conversion of the detection value from a stationary coordinate system to a rotating coordinate system, an excitation current detection id and a torque current detection value i are determined.
and the slip frequency command value ωs based on the input of the excitation current detection value id and the torque current detection value iq from the coordinate converter that outputs q.
^2. The induction motor control device according to claim 1 , further comprising: a slip frequency calculator for calculating ^* . 3. The detected value of the output current from the power converter is
Input and change this detection value from the stationary coordinate system to the rotating coordinate system.
Coordinate conversion that outputs a torque current detection value iq based on the conversion
And a means for outputting a secondary magnetic flux command value Φ2 * of the induction motor
And a torque current detection value iq from the coordinate converter, and
The slip frequency command value ω based on the secondary magnetic flux command value φ ₂ ^*
2. The induction motor control device according to claim 1 , further comprising: a slip frequency calculator for calculating s ^* . 4. An exciting current command value id for an induction motor.
^* , Torque current command value iq ^* , primary angular frequency command value ω ₁
^* To determine the voltage command value v ₁ ^* and the phase angle command value δ ^* .
Based on the voltage command value v ₁ ^* and the phase angle command value δ ^* .
Supply AC power with variable voltage and variable frequency to the induction motor
Output voltage command values vu ^* , v
v ^*, vector for the induction motor by outputting a vw ^*
In an induction motor control device that performs torque control, the detected values of the output voltage and output current from the power converter are
Input and calculate the secondary magnetic flux of the induction motor or
Based on the estimation, the excitation current calculation value id ＾ and the torque current
Magnetic flux calculating means for outputting a calculated value iq ＾, and means for outputting a secondary magnetic flux command value Φ2 * of the induction motor
And the secondary magnetic flux command value φ ₂ ^* and the secondary magnetic flux calculation value φ ₂
Enter に for backpropagation algorithm
While performing based learning, the excitation current command value id ^*
A neutral network means at the preceding stage for outputting, an exciting current operation value id ＾ from the magnetic flux operation means and a torque
Current calculation value iq ＾ and the excitation current command value id
^* , Torque current command value iq ^* , primary angular frequency command value ω ₁
Enter ^* to set backpropagation algorithm
Outputs two output signals A * and B * while performing base learning.
A post-stage neural network means, and an output signal from the post-stage neural network means
The voltage command value v ₁ ^* and the phase angle finger based on A * and B *
Means for calculating the prescriptive value δ ^* and the learning result of the neural network means at the preceding stage.
While detecting the torque current value i from the coordinate converter.
q, and the slip frequency based on the secondary magnetic flux command value Φ2 *.
An induction motor control device comprising: a slip frequency calculator for calculating a number command value ωs ^* . 5. The neural network means performs learning based on the back propagation algorithm only when the rotation speed of the induction motor is equal to or higher than a predetermined level. 3. The method according to claim 1, wherein
5. The induction motor control device according to any one of 4 .