JPH0413954B2 - - Google Patents

Description

[Detailed description of the invention] [Technical field to which the invention pertains] This invention relates to an induction motor (also referred to as an induction machine).
A so-called vector control method in which the spatial vector of the next current (also called winding current or stator current) is controlled by dividing it into a component in the same direction as the magnetic flux vector (magnetizing current) and a component orthogonal to this (torque current).
In particular, the present invention relates to a vector control method for an induction machine that can improve the calculation accuracy of the magnetic flux vector necessary for this purpose.

[Prior art and its problems]

In the following description of each figure, the same reference numerals indicate the same or corresponding parts. In addition, logic "High" and "Low" are simply written as "H" and "L".

In vector control that realizes high-performance variable speed operation of induction machines, it is important to accurately detect the magnetic flux vector that serves as the control reference. One of the practical methods for calculating the magnetic flux vector is the current model method described in Reference 2, which will be described later. This method calculates the magnetic flux from the motor winding current and its winding position. Specifically, the target value of the magnetizing current (referred to as the magnetizing current target value) i _M ^* , the target value of the torque current (torque current (referred to as target value) i _T
A configuration that calculates from ^* and motor constants is conceivable.
This method can calculate from the motor stop state to the highest frequency range, but the calculation formula includes the motor's secondary time constant, especially the secondary resistance R ₂ ' whose value changes significantly depending on the motor temperature, as a parameter. This has the disadvantage that unless the constant of the arithmetic unit is corrected accordingly, the magnetic flux calculation accuracy will be poor. This correction method uses the voltage model method (i.e., calculates the back electromotive force from the motor terminal voltage and integrates it to obtain the magnetic flux component), which is one of the other practical methods for calculating the magnetic flux vector. The magnitude of the magnetic flux vector to be found, the correction method using the steepness, the value representative of the magnetic flux, e.g. 2
Improving current model calculation accuracy in the low frequency region where motor voltage is small by using a correction method based on the magnitude or angle of voltage vectors such as secondary induced voltage, air gap induced voltage, and primary induced voltage, or by using the former correction method. A method to do this has been proposed. (References: (1) JP-A-5-79468 “Method and device for detecting rotor resistance of asynchronous machine”, (2) JP-A-56-115190
No. ``Magnetic flux phase control device'', (3) Japanese Patent Application Laid-open No. 58-139692
No. ``Magnetic flux vector calculator for induction motors'', (4) Japanese Patent Publication No. 2-26476 ``Vector control method for induction motors'').

In addition, in Japanese Patent Application Laid-open No. 61-39887 "Control Method for Induction Motor" (Reference (5)), the present applicant describes a regulator that zeros out the deviation in the magnitude of the magnetic flux calculated by each method of the current model and voltage model. , operate only in high speed and heavy load areas, and when entering low speed or light load areas, store the regulator output immediately before disabling this regulator, Alternatively, the accuracy of magnetic flux calculation in the light load region is improved.

Next, the outline and problems of this control method (reference document 5) will be explained using FIGS. 4 to 7. Figure 4 is a block diagram showing the circuit configuration of the reference document (5) method, and Figure 5 is the secondary time constant correction circuit 101 in Figure 4.
6 is a block diagram showing the detailed structure of the switching signal generator 90 in FIG. 5, and FIG. 7 is a vector diagram showing the magnetic flux and current vector of the induction motor. In FIG. 7, the primary current i ₁ of the induction motor can be understood as a space vector rotating around the rotor axis. Similarly, the rotor linkage flux φ ₂ produced by the primary current can also be regarded as a space vector rotating around the rotor axis. Of the primary current vector i ₁ →, the component in the same direction as the magnetic flux vector Φ ₂ → is the magnetizing current i _M as described above.
It corresponds to the field current of a DC machine. Furthermore, the component of the primary current vector in the direction perpendicular to the magnetic flux vector is called the torque current i _T as described above, and corresponds to the armature current of a DC machine. If these components of an induction motor can be controlled independently of each other, it is possible to make the induction motor, which is cheaper and more robust than a DC motor, exhibit variable speed control performance equivalent to that of a DC motor. . Controlling an induction motor based on this principle is called vector control of an induction motor. In addition,
Items with arrows (→) represent vector quantities; however, unless the vector quantities are particularly emphasized, the arrows will not be used to distinguish between them.

In order to enable vector control of such an induction motor, a primary current vector is provided with command values in the form of two components orthogonal to each other with the axis of the rotating magnetic flux vector (magnetic flux axis; M axis) as a reference. It is understood that the position of the magnetic flux axis (M-axis) must be detected because it is necessary to convert it into a vector quantity based on a fixed axis (α-axis) that does not rotate. Note that this fixed axis (α-axis) is generally set at the winding axis of one phase of the stator. By doing so, the position of the magnetic flux axis (M axis) can be expressed by the rotation angle with respect to this fixed axis (α axis). The current or voltage model type magnetic flux vector calculator indirectly detects the position angle of Φ ₂ → of the magnetic flux vector or the magnitude of the magnetic flux vector by calculating Φ. In vector control, it plays an important role in determining control performance.

Next, in FIG. 4, 101 is a secondary time constant correction circuit, 206 is a multiplier, 12 is a speed regulator, 1
3,203 is a divider, 14,205 is a vector rotator, 15 is a 2-phase/3-phase converter, 16 is a current regulator, 17 is a power conversion device such as a cycloconverter, 18 is a current detector, and 19 is a Induction motor, 20 surrounded by a dotted line frame is a current model type magnetic flux calculator, 2
1 is a rotor position detector, 22 is a speed detector, 23
is a voltage detector, 24 is a voltage model type magnetic flux calculator,
201 is a differential circuit, 202 is a proportional element, 204 is a nonlinear function generator, and 30 is a rotor position calculator.

In other words, the setting value of the magnitude of magnetic flux (magnetic flux setting value)
It is given to the Φ ^* divider 13, a secondary time constant correction circuit 101, a differentiation circuit 201, and a proportional element 202, which will be described later, and the calculation output of the differentiation circuit 201 becomes the magnetizing current target value i _M ^* and is given to the vector rotator 14. It will be done.
On the other hand, the speed regulator 12 performs adjustment calculations to make the actual speed value n obtained through the speed detector 22 coincide with its target value n ^* , and the torque target value T ^* as the adjustment output is calculated using a divider. 13 to the vector rotator 14 as the torque current target value i _T ^* , and also to the correction circuit 101 as a quantity representing the magnitude of the load. The vector rotator 14 sets these stator current vector target values i _M ^* ,
Based on i _T ^* and the unit vector (cos, sin) of the magnetic flux vector Φ ₂ given via the calculator 20, the target value i ₁ ^* of the stator current vector I ₁ is calculated using the following equation (1). As ^shown _{in FIG} ^.

i ₁ 〓 ^* ＝i _M ^* cos−i _T ^* sin i ₁ 〓 ^* ＝i _M ^* sin＋i _T ^* cos……(1) The two-phase/three-phase converter 15 is constructed as shown in equation (1). The obtained two-phase target values i ₁ ^* , i ₁ ^* are converted into three-phase target values i _a ^* , i _b ^* , i _c ^* as shown in equation (2) below.

i _a ^* = i ₁ 〓 ^* i _b ^* = −1/2i ₁ 〓 ^* +√3/2i ₁ 〓 i _c ^* = −1/2i ₁ 〓 ^* −√3/2i ₁ 〓 ………(2) These target values i _a ^* , i _b ^* , i _c ^* are guided to the current regulators 16a, 16b, 16c, respectively, so that each regulator receives the electric power detected via the current detectors 18a, 18b, 18c. Each phase current of converter 17
Adjustment calculations are performed to make i _a , i _b , and _ic coincide with their target values i _a ^* , i _b ^* , and _ic ^* , respectively.

In this way, vector control of the induction motor 9 is performed, and the magnetizing current target value i _M ^* is required to obtain the magnitude of the magnetic flux vector ΦΦ ₂ → necessary for such control.
and the angular position cos from the fixed axis (α−β),
sin is determined by a current model type magnetic flux calculator indicated by a broken line frame 20.

In the current model type magnetic flux calculator 20, the magnetic flux set value Φ ^* is inputted to the differentiator circuit 201 as described above, and the differential calculation of the following equation (3) is performed to output the magnetizing current target value i _M ^* . , to the vector rotator 14.

i _M ^* = (1/lm') (1 + ST ₂ ) Φ ^* ......(3) where lm' is the mutual inductance between the stator circuit and rotor circuit of the induction machine, and T ₂ is the electrical time of the rotor circuit. It is a constant (also called a secondary time constant).

Further, as described above, the magnetic flux target value Φ ^* is also input to the proportional element 202, and the value (Φ ^* /lm') is calculated and outputted and given to the divider 203 as a divisor. On the other hand, the torque current target value i _T is input to the divider 203 as the target to be removed.
Since ^* is input, the value (lm'/Φ ^* )i _T ^* is output from the divider 203 and given to the multiplier 206.

On the other hand, the multiplier 206 inputs the corrected second-order time constant (the reciprocal of the reciprocal, hereinafter the term "reciprocal" will be omitted) 1/T ₂ from the second-order time constant correction circuit 101 as described later. From 206, an output corresponding to the slip frequency ωsl (=d/dt−dθ/dt), which is expressed by the calculation (multiplication) obtained by the following equation (4), is obtained.

ωsl=(lm'/Φ ^* )(1/ _T2 ) _iT ^* ......(4) The quantity corresponding to this slip frequency _ωsl is further led to the nonlinear function generator 204, and from its output, The quantities cosλ and sinλ are obtained. At this time, the following relationship holds between the slip frequency ω _sl and λ: λ=∫ω _sl dt =∫(d/dt−dθ/dt) dt=−θ (5). The outputs cosλ and sinλ of the nonlinear function generator 204 are transmitted to the vector rotator 205 along with unit vectors cosθ and sinθ representing the rotor position.
guided by. Note that the unit vectors cosθ and sinθ are
It is determined by a rotor position detector 21 and a rotor position calculator 30 connected to the rotor shaft of the induction motor 19. The vector rotator 205 outputs cos and sin, which are the α and β axis components of the unit magnetic flux vector, through the following calculation: cosλcosθ−sinλsinθ=cos(λ+θ) =cos sinλcosθ+cosλsinθ=sin(λ+θ) =sin (6) . Note that the output of the vector rotator 205 is given to the vector rotator 14 as described above.

On the other hand, the voltage model type magnetic flux calculator 24 calculates instantaneous line voltage values v _ab and v _bc of the motor terminal voltages obtained via the voltage detectors 23a and 23b, and the current detector 1
By performing a predetermined calculation from the motor phase currents i _a , i _b , and i _c obtained through 8a, 18b, and 18c,
The actual magnetic flux value Φ as the magnitude of the magnetic flux vector with the α-β axis as a reference is outputted and given to the secondary time constant correction circuit 101 .

The secondary time constant correction circuit 101 inputs the actual magnetic flux position Φ, magnetic flux set value Φ ^* , torque current target value i _T ^* , and actual speed value n obtained from the voltage model type magnetic flux calculator 24, and calculates the The quadratic time constant 1/ with different corrections depending on the region and other regions.
This circuit outputs T ₂ and its configuration is shown in FIGS. 5 and 6.

That is, in FIG. 5, 90 is a switching signal generator, 52 is a regulator, 54 and 55 are changeover switches,
57 is a secondary time constant setter, 58 is an operational amplifier, 5
3, 56 is an addition point, 59 is a capacitor,
The changeover switch 55, capacitor 59, and operational amplifier 58 constitute a so-called sample-and-hold circuit SH. Here, the initial value (value at startup) of the voltage of the capacitor 59 is 0.

In this example, the regulator 52 works so that the actual magnetic flux value Φ from the voltage model type magnetic flux calculator 24 matches the magnetic flux set value Φ ^* , and at high speed and high load (also called high frequency time, in this case, the switching switch 54,
55 switches to the dotted line position), the controller 52 changes at the addition point 56 to the secondary time constant (reciprocal) set value 1/T ₂ ^* set by the secondary time constant setter 57.
Correction is performed by adding the adjustment output of , and the corrected secondary time constant 1/T ₂ is output.

On the other hand, in the case of low speed or light load, in the former case, to prevent the influence of deterioration of the calculation accuracy of the voltage model type magnetic flux calculator 24, and in the latter case, to prevent the deviation of the operating point of the regulator 52 or In order to prevent the effect of loss of function due to saturation, the switching signal generator 90 inputs the actual speed value n and the torque current target value i _T ^* , and outputs the switching signal CH under the condition that the actual speed value n and the torque current target value i T * fall below a predetermined speed level or load level. The changeover switches 54 and 55 are set to the solid line positions shown in the figure to disconnect the regulator 52 from its regulation system. This allows the sample and hold circuit SH
The correction amount as the adjustment amount that was output from the regulator 52 immediately before the switching of the changeover switch 55 is stored in the capacitor 59, and is sent to the addition point 5 via the operational amplifier 58.
6, and the second-order time constant 1/T ₂ thus corrected in a new form is output from this correction circuit 101.

The details of the switching signal generator 90 are shown in FIG. In the figure, reference numerals 91 and 92 are absolute value circuits, into which torque current target value i _T ^* speed actual value n is input, respectively. Note that the target value i _T ^*
represents the magnitude of the load, and instead of this value i _T ^* , an actual torque current value i _T , a torque target value, or an actual torque value may be used.
Further, 93 and 95 are comparators, 94 is a load level setter, 96 is a speed level setter, and 97 is an AND circuit.

Absolute values of the torque current target value i _T ^* and the actual speed value n are taken by absolute value circuits 91 and 92, and the respective level setters 9 are set in comparators 93 and 95, respectively.
The comparator 93,
The outputs of 95 are ANDed in an AND circuit 97, and the switching signal CH as the output signal is used as an operation or inoperation signal for the regulator 52 in FIG. 5 and a signal for operating the changeover switches 54 and 55. used. In addition, switch 5 is here.
When 4 and 55 are at the solid line position, that is, at low speed (at low circumferential speed) or light load as described later, the output of the regulator 52 is maintained at zero (i.e., inactive state) by the switching signal CH. It is assumed that The setting levels for the load and speed in the level setters 94 and 96 are normally selected to be a low load and low speed level of about 10% of the rated value.

Next, the principle of magnetic flux correction operation of the control circuit shown in FIG. 4 will be described. In the current model type magnetic flux calculator 20 of this example, the slip frequency ω _sl is corrected by the corrected secondary time constant 1/T ₂ through the multiplier 206 . At this time, in the case of a heavy load, since the torque current i _T ^* is large, the range of the slip frequency ω _sl that changes due to the corrected and outputted secondary time constant 1/T ₂ is large.

For example, when these circuits are implemented using analog electronic circuits, the maximum value is usually 10V, so when the load is heavy, the input signal level as the multiplicand of the multiplier 206 ((lm'/Φ ^* in equation (4) above) When i _T ^* ) is 10V, the quadratic time constant 1/
When T ₂ changes as 0 ^v ≦(1/T ₂ ) ≦10 ^v , the level of the output slip frequency ω _sl changes as 0 ^v ≦ω _sl ≦10 ^v .

On the other hand, in the case of a light load, when the input signal level corresponding to (lm'/Φ ^* )i _T ^* is 2 ^V , and the quadratic time constant changes as described above, the slip frequency ω _sl
The level of V only changes from 0 to 2 ^V.

In this way, the range of change in the slip frequency ω _sl (that is, the signal effective for correcting the position angle of the calculated magnetic flux) calculated by the current model type magnetic flux calculator 20 depends on the magnitude of the torque current target value i _T ^* . Therefore, if the regulator 52 continues to operate regardless of the load size, especially when there is no load, even at high speed, i _T ^* ≒ 0, therefore ω _sl ≒ 0, which is effective for correction. Although the signal level is extremely small, if there is a detection error in the voltage model type magnetic flux calculator 24 or if there is an offset voltage in the various calculation circuits, the values of these error components will be In other words, the magnetic flux setting value Φ ^* and the actual value Φ
This error component accounts for the deviation from the current value (in other words, an incorrect deviation is applied to the regulator 52).
For this reason, for example, this incorrect deviation may continue for a long time and be integrated by a capacitor (not shown) in the regulator 52, causing the operating point of the regulator 52 to deviate from the correct position or become saturated and stop functioning as a regulator. may be lost. At this time, if the induction machine is operated at steady state with no load, deviation or saturation of the operating point of the regulator will not affect the characteristics, but if the speed or load suddenly changes. In this case, the torque current target value i _T ^* changes greatly, so
The time it takes for the regulator 52 to come out of the deviation or saturation state and operate at the normal operating point is determined by the secondary time constant 1/2, which is corrected and output via the regulator 52.
T ₂ becomes significantly different from the true value, and the slip frequency calculation includes an error. in this case,
Since a slip frequency larger than the true value is given,
The actual magnetic flux value Φ decreases, causing deterioration of transient characteristics such as a large amount of torque current flowing to generate the necessary torque. Therefore, in the control circuit shown in FIG. 4, the accuracy of correction for the current model calculation magnetic flux is reduced. Under low speed or light load conditions, a method is used in which the regulator 52 is rendered inoperative and disconnected from the regulation system, and the regulation output immediately before the regulator is rendered inoperable is stored and used for magnetic flux correction. ing.

Next, to explain this correction operation based on FIGS. 4 to 6, in FIG. 6, when the output of the AND circuit 97, that is, the switching signal CH is "H", the high load, high speed side ,5
5 is the dotted line position, and the regulator 52 is in operation.

Next, when the speed n is smaller than the value set by the speed level setter 96, the output signal of the comparator 95 becomes "L", and therefore the output of the AND circuit 97, that is, the switching signal CH is also "L", and the changeover switch is 54,
55 is the position indicated by the solid line, and the input capacitor 59 of the operational amplifier 58 stores the output of the regulator 52 immediately before the switch 55 is turned off. At this time, the comparator 93 is logically ANDed with "H" or "L" and the load condition (which varies depending on the magnitude of the torque current i _T ^* , but the output "L" of the comparator 95 is In addition, when the speed n increases, the output of the comparator 95 becomes "H". In this region, the output signal of the comparator 93 turns the changeover switches 54 and 55 on and off. For example, when the torque current target When the value i _T ^* is smaller than the level set by the load level setter 94, the output of the comparator 93 becomes "L", so the switching signal CH becomes "L", so the regulator 52 becomes inoperable and at the same time A capacitor 59 stores the adjustment output immediately before the regulator 52 becomes inoperable after being disconnected from the adjustment system in the same way as in the low-speed region, and this stored adjustment output is applied to the multiplier 206 via the summing point 56. , the current model calculation magnetic flux is corrected.

Furthermore, when the load increases above the set level, the output signal of the comparator 93 becomes "H", and therefore the switching signal CH
becomes "H", the regulator 52 enters the operating state, and is incorporated into its regulation system, and the regulation output of the regulator 52 is given to the multiplier 206 via the addition point 56, and the current model calculation magnetic flux is corrected. will be exposed.

By the way, in the above control method, when the load is light, the operation of the regulator 52 is stopped, the adjustment output immediately before the stop is stored, and the 2
The current vector calculation magnetic flux is corrected by obtaining the following time constant 1/ _T2 . However, if there is a setting error in the mutual inductance lm' in the differentiating circuit 201 or the like, a large deviation may occur between the magnetic flux setting value Φ ^* and the actual value Φ due to the adjustment function of the regulator 52 being stopped during this time. be. In this case, there is usually no problem if steady operation continues, but if the actual magnetic flux value Φ is the set value Φ ^*
If the load is suddenly applied, an excessive rush current (torque current) will flow to the induction machine, or the transient response time to settle down to a steady state will be abnormally long. Further, if the actual magnetic flux value Φ exceeds the set value Φ ^* , problems such as an abnormal increase in iron loss will occur.

[Purpose of the invention]

The purpose of the present invention is to correct the secondary time constant through a regulator that adjusts so that the deviation between the magnetic flux target value and the magnitude of the voltage model calculation magnetic flux becomes 0, thereby correcting the current model calculation magnetic flux. It is an object of the present invention to provide a control method for a vector control circuit for an induction machine, which can maintain the magnetic flux value at a predetermined value even under light load, while eliminating the above-mentioned drawbacks.

[Key points of the invention]

The main points of the present invention are the torque target value of the induction motor,
A current model type magnetic flux calculation means that calculates the motor magnetic flux vector from the magnetic flux target value and rotor position information, and a back electromotive force that is calculated from the motor terminal voltage, stator current, and motor constant, and is integrated to calculate the motor magnetic flux vector. a voltage model type magnetic flux calculation means for calculating the voltage model type magnetic flux calculation means; and an adjustment means for adjusting the difference between the magnitude of the magnetic flux vector calculated by the voltage model type magnetic flux calculation means and the magnetic flux target value to zero, In the vector control method for an induction motor in which control is performed by correcting the magnetic flux vector calculated by the current model type magnetic flux calculation means (via a correction amount of a secondary time constant, etc.), the motor rotation speed is separately detected. mode (first mode and ), the adjustment output of the adjustment means is adjusted to a magnetization current correction means (such as a magnetization current target value addition point) that corrects the magnetization current.
When in a mode other than the first mode (referred to as the second mode), the initial value is memorized in advance, and after passing through the first mode, the value immediately before transitioning from the first mode to the second mode is The present invention is characterized in that a stored value of a storage means (sample and hold circuit) for storing the adjustment output of the adjustment means is provided to the magnetizing current correction means.

[Embodiments of the invention]

Embodiments of the present invention will be described below based on FIGS. 1 to 3. FIG. 1 is a block diagram showing an example of the configuration of a control circuit according to the present invention, and corresponds to FIG.
The diagram shows the magnetic flux/secondary time constant correction circuit 3 in Figure 1.
5 is a block diagram showing an example of the detailed configuration of 01, in which the functional unit that outputs the corrected secondary time constant 1/T ₂ corresponds to that in FIG. 5.

FIG. 3 is a block diagram showing an example of the detailed configuration of the switching signal generator 50 in FIG. 2, in which the functional section that outputs the switching signal CH corresponds to that in FIG.

The difference between FIG. 1 and FIG. 4 is that a new magnetic flux/secondary time constant correction circuit 301 replaces the conventional second-order time constant correction circuit, and an addition point 302 is added. The magnetizing current correction amount △i _M output from the correction circuit 301 is added to the conventional magnetizing current target value (output of the differentiating circuit 201) to obtain a new magnetizing current target value i _M ^* , which is given to the vector rotator 14. This is what we did.

In FIG. 2, 50 is a switching signal generator, 64,
65 is a changeover switch, 68 is an operational amplifier, 69 is a capacitor, and the switch 65, capacitor 69, and operational amplifier 68 are a new sample and hold circuit SH.
1. Here, the initial value (starting value) of the voltage of the capacitor 69 is assumed to be zero. The new switching signal generator 50 generates conventional and unusual switching signals.
CH is applied to the changeover switches 54, 55, and a new changeover signal CH1 is applied to the changeover switches 64, 55.
Similarly, the switching signal CH2 is applied to the regulator 52 at 65.

Further, in FIG. 3, 83 is a comparator, 84 is a new load level setter, 86 is a NOT circuit, and 87
is a new AND circuit, and 82 is an OR circuit. Also
AND circuits 97, 87 and OR circuit 82 output the switching signals CH, CH1 and CH2, respectively.

Next, the operations shown in FIGS. 1 to 3 will be explained. As shown in FIG. 3, absolute values of the torque current target value i _T ^* and actual speed value n as input signals are taken by absolute value circuits 91 and 92, respectively, and comparators 93, 95, 83
are compared with the setting levels of setting devices 94, 96, and 84, respectively. It is assumed that the setting level of 94 is higher than the setting level of 84. When the speed is lower than the level set by the setting device 96 (referred to as the low speed region), the output of the comparator 95 is “L”.
Switching signal as output of AND circuits 97, 87
CH, CH1 are also "L", and accordingly, the switching signal CH2 as the output of the OR circuit 82 is also "L", and as a result, the regulator 52 in FIG. 65
is at the position indicated by the solid line, and the adjustment output of the regulator 52 immediately before this switching is sampled and held in the sample hold circuit.
The input capacitors 59 and 69 of SH and SH1 memorize the 2
The following time constant 1/T ₂ and magnetizing current correction amount Δi _M are output.

Next, when the speed is above the setting level of the setting device 96 (referred to as high-speed region) and the magnitude of the torque current target value i _T ^* is below the setting level of the setting device 84 (referred to as light load), the comparison 93, 95, 83, NOT
The outputs of the circuit 86 are "L", "H", "H",
It becomes “H” and therefore AND circuit 97, 87, OR
Output of circuit 82 (switching signal CH, CH1, CH2)
are "L", "H", and "H", respectively. As a result, the regulator 52 is brought into operation and incorporated into its regulation system, and the changeover switches 54 and 55 remain in the solid line position, while the changeover switches 64 and 65 are switched to the dotted line position. Therefore, the regulator 52 newly applies the magnetizing current correction amount Δi _M so that the magnetic flux set value Φ ^* and the actual magnetic flux value (detected value) Φ match.

Next, from this state, when the magnitude of the torque current target value i _T ^* becomes a value larger than the setting level of the setting device 84 and smaller than the setting level of the setting device 94 (referred to as medium load), , comparator 93,9
The outputs of 5 and 83 are "L", "H", "L", respectively.
As a result, the outputs of AND circuits 97, 87 and OR circuit 82 (switching signals CH, CH1, CH2) are all “L”.
becomes. As a result, the regulator 52 becomes inoperative again, the changeover switches 54 and 55 remain as they are (in the solid line position), and the changeover switches 64 and 65 return to the solid line position. This allows the sample and hold circuit to
SH1 stores the adjustment output before switching and supplies this output as the magnetizing current correction value Δi _M.

Next, from this state, the torque current target value i _T ^*
increases and its magnitude becomes a value larger than the set level of the setting device 94 (referred to as heavy load), the outputs of the comparators 93, 95, and 83 become "H", "H", and "H", respectively. By “L”, AND circuit 97,8
7, Output of OR circuit 82 (switching signal CH, CH1,
CH2) becomes “H”, “L”, and “H”. As a result, the regulator 52 becomes operational again and is incorporated into its regulation system, the changeover switches 64 and 65 remain as they are (solid line positions), the changeover switches 54 and 55 switch to the dotted line positions, and the adjuster 52 changes to the magnetic flux setting value. Φ ^* is the quadratic time constant 1/ which is corrected so that the actual value Φ matches.
Give T ₂ .

In this way, the regulator 52 is incorporated into the regulation system during light loads and heavy loads in the high-speed region, and uses its regulation output to set the magnetizing current target value i _M ^* and the secondary time constant setting value 1/T ₂ ^* , respectively. Under conditions other than the above, the adjustment output is stored in the memory circuit (sample and hold circuit SH1, SH) corresponding to each correction and used for the correction operation, and is used to adjust the speed and load. Correct magnetic flux can be calculated regardless of size. Note that normally, the setting level of the setting device 96 is selected to be approximately 10% of the maximum speed, the setting level 94 is approximately 10% of the maximum load, and the setting level 84 is approximately 5% of the maximum load. In the above example, the torque current target value i _T ^* was used to detect the load level, but instead of this, the torque current actual value i _T , the rectified value of the converter output current, the speed regulator output signal (torque target value), etc. It is also possible to use a signal that varies monotonically depending on the magnitude of the load.

〔Effect of the invention〕

As is clear from the above description, in the present invention, adjustment is made by correcting the secondary time constant and magnetizing current target value so that the difference between the magnetic flux target value and the magnitude of the calculated magnetic flux based on the voltage model method becomes 0. A regulator is provided to perform the operation, and the calculated magnetic flux vector based on the current model method is corrected, and the following correction method is used depending on the speed range and load conditions. That is, in a low speed region or a light load region where the magnetic flux correction operation via the correction of the secondary time constant by the regulator is impossible, the correction signal (adjustment output) immediately before the correction operation is stopped is stored, and the correction signal (adjustment output) is stored. The signal corrects the secondary time constant and continues to correct the magnetic flux calculated by the current model, and when the load is light in the high speed region, the (magnitude) of the magnetic flux is ) correction operation is performed,
If the load increases further in the high-speed region, the adjustment output is diverted to correction of the secondary time constant, and at this time, the adjustment output immediately before the adjustment system of the regulator 52 is switched is memorized, and the adjustment output is continued using the stored signal. By continuously correcting the magnetizing current target value, it is possible to obtain the same high magnetic flux calculation accuracy in the low speed region and light load region as in the high speed region and heavy load region.
It is also possible to shorten the response time to load fluctuations and the like.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a control circuit as an embodiment of the present invention, FIG. 2 is a block diagram showing a detailed configuration example of the main part of FIG. 1, and FIG. 3 is a block diagram of the main part of FIG. 2. FIG. 4 is a block diagram showing an example of a detailed configuration of a conventional control circuit. FIG. 5 is a block diagram showing an example of a detailed configuration of the main part of FIG. 4. FIG. FIG. 7 is a block diagram showing a detailed configuration example of the main part of the induction machine, and FIG. 7 is a vector diagram showing the magnetic flux and current vector of the induction machine. 12... Speed regulator, 13,203... Divider, 14,205... Vector rotator, 15...
2-phase/3-phase converter, 16...Current regulator, 17...
...Power converter, 18...Current detector, 19...
Induction motor (induction machine), 20... Current model magnetic flux calculator, 21... Rotor position detector, 22... Speed detector, 23... Voltage detector, 24... Voltage model magnetic flux calculator, 201... Differential circuit, 202...
Proportional element, 204... Function generator, 206... Multiplier, 301... Magnetic flux/secondary time constant correction circuit, 3
02...Summing point, 50...Switching signal generator, 52
...Adjuster, 53,56...Additional point, 54,5
5, 64, 65...Selector switch, 57...Secondary time constant setter, 58,68...Operation amplifier, 5
9, 69... Capacitor, SH, SH1... Sample hold circuit, 82... OR circuit, 86...
NOT circuit, 91, 92...absolute value circuit, 84,
94...Load level setter, 96...Speed level setter, 87,97...AND circuit, 83,93,
95...Comparator, CH, CH1, CH2...Switching signal.

Claims (1)

[Scope of Claims] 1. A current model type magnetic flux calculation means for calculating a motor magnetic flux vector from the torque target value, magnetic flux target value, and rotor position information of the induction motor; and calculating a back electromotive force from the motor terminal voltage and stator current. a voltage model type magnetic flux calculation means that calculates a motor magnetic flux vector by calculating and integrating the magnetic flux vector; In the vector control method for an induction motor, the vector control method for an induction motor performs control by correcting a magnetic flux vector calculated by the current model type magnetic flux calculation means, wherein the separately detected motor rotation speed is set. When in a mode (referred to as the first mode) in which the value of the quantity representative of the motor current or load size, which is detected separately, is higher than the set value and smaller than the set value, the adjustment output of the adjustment means is changed to the magnetizing current. When in a mode other than the first mode (referred to as the second mode), the initial value is stored in advance, and after passing through the first mode, the 2. A method for controlling an induction motor, characterized in that a value stored in a storage means for storing the adjustment output of the adjustment means immediately before shifting to the second mode is given to the magnetizing current correction means.