JPH02304380A - Measuring method for constant of induction motor - Google Patents

Abstract

PURPOSE:To surely and easily execute an automatic measurement of a stop state of a motor whose constant is unknown by identifying a constant of an induction motor by using an output voltage of an inverter which becomes a driving power source of the induction motor and a current of the induction motor. CONSTITUTION:An induction motor 1 which becomes an object for measuring a constant uses a PWM inverter 2 as a driving power source. Prior to a actual operation of this motor 1, the inverter 2 is allowed to function as a motor constant measuring instrument. Current detectors 3a - 3c output current feedback signals Iu - Iw responding to a primary current of each phase of the motor 1. A 3-phase/d<e>q<e> coordinate transformer 4 converts the signals iu - iw into an orthogonal axis component shifted by theta from a fixed orthogonal axis. A function generator 5 outputs a signal sintheta and costheta corresponding to a phase angle com mand theta. A d<e>q<e>/3-phase coordinate transformer 6 converts a d<e> axial component of a primary voltage command into 3-phase voltage commands Vu*, Vv* and Vw. A constant tuning device 7 identifies a primary and a secondary resistances, a primary and a secondary self-inductances and a mutual inductance from these voltage and current.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for measuring the constants of an induction motor, and more particularly to a method for automatically measuring the constants of an induction motor connected to an inverter.

[Conventional technology]

A vector control method has been developed as a control method for induction motors, and is known to achieve transient characteristics and stability equivalent to those of DC motors. An induction motor control device using this vector control method performs various controls based on constants of the induction motor to be controlled, such as primary resistance, secondary resistance, primary leakage inductance, secondary leakage inductance, mutual inductance, etc. A constant is set.

For this reason, conventionally, the design values of the motor or the "IEEJ, Electrical Standards Committee Standards J JEC" published by Denki Shoin were used.
-37-1979 (Storytelling Machine) As shown in Part 2,
Necessary motor constants are determined from the measured values of resistance measurement tests, no-load tests, and restraint tests, and control constants are set.

(Problems to be Solved by the Invention) In an induction motor control device using a conventional vector control method, calculations from design values and constant measurement tests are time-consuming to obtain the constants of the motor to be controlled. There were problems such as being complicated. In particular, in general-purpose variable speed drives, the constants of the motor to be controlled are unknown, so a constant measurement test is performed each time depending on the motor model, and control constants are set based on the obtained motor data. I needed to.

Furthermore, the data obtained from the design values may have a large error from the actual values, resulting in control calculation errors, requiring confirmation and adjustment or resetting of the control device.

The present invention has been made to solve the above-mentioned problems.In an induction motor using an inverter as a driving power source, the inverter functions as a motor constant measuring device before actual operation to generate torque in the motor. This invention shares a method for measuring the constants of an induction motor with high precision, without rotating or rotating the motor.

[Means to solve the problem]

This method for measuring constants of an induction motor according to the first invention is as follows:
In an induction motor connected to an inverter as a drive power source, an inverter output voltage is generated so as to achieve a single-phase power supply or an equivalent power supply state, and this voltage and the current flowing through the induction motor are used to generate a model-based adaptive system. The constants of the induction motor are identified, and the primary resistance Rs,
The secondary resistance R1, the primary self-inductance 1, the secondary self-inductance r, and the mutual inductance M are determined.

Further, the method for measuring the constants of an induction motor according to the second invention uses the inverter output voltage and frequency generated in the same manner as in the first invention, and the current flowing through the induction motor to convert the voltage to the current. Find the gain characteristics for the frequency of
The primary resistance R8, the secondary resistance Rs, the primary self-inductance L1, the secondary self-inductance Lr%, and the mutual inductance M are obtained by substituting Lr into the equation showing the transfer function of the induction motor.

[Effect]

The induction motor constant identification method based on the model norm adaptation system in this first invention employs a parallel type identifier that uses output error as the identification algorithm. When the Lr error becomes τ, the unknown parameter calculated matches the constant of the actual conductive motor. It works like this.

In addition, the method for measuring the constants of an induction motor in this second invention is to obtain the gain characteristic of the induction motor from voltage to current with respect to frequency in a single-phase power supply state, and to calculate the gain at a plurality of frequencies from among the gain characteristics of the induction motor in a single-phase power supply state. By substituting Lr into the equation showing the transfer function from voltage to current in the single-phase power supply state and calculating them as a simultaneous equation, the primary resistance R
1. Secondary resistance Rs, primary self-inductance Lr, secondary self-inductance L, and mutual inductance M are determined.

〔Example〕

Embodiments of the first invention will now be described in detail with reference to the drawings. FIG. 1 is a block diagram showing a constant measuring device for an induction motor according to this embodiment. In the figure, (1) is the induction motor whose constants are to be measured, (2) is the PWM inverter, and (3a), (3b), and (3c) are the U phase, ■ phase, and W phase of the induction motor (1), respectively. (4) is the current feedback signal i, which outputs current period signals 11x, lv, and 1w in response to the primary current. ,L,iv
The orthogonal axis component (d) is shifted by θ from the orthogonal axis with fixed.

Axis primary current component id@, and qa-axis primary current component i
3-phase/d"q' coordinate converter to convert to q a-, (5
) is a function generator that outputs sinusoidal signals sin θ and cos θ corresponding to the phase angle command θ, (6) is the d@axis component Vd' of the primary voltage command, and "is the 3-phase voltage command vu", vv
d@q″/three-phase coordinate converter that converts to “, v,” (7
) is a constant tuning device that identifies the conductive vJm constant from these voltages and currents.

Here, in order to make the induction motor (1) equivalent to a single-phase power supply state, the phase angle command θ is set to 0, and the orthogonal axes are shifted by θ from the fixed orthogonal axes (d-axis, q-axis). d@
A predetermined value is applied to the voltage Vd'' on the axis and the q@ axis (coinciding with the d axis and the q axis here), and LrVQ''g is set to 0. At this time, voltages Vu and Vv shown by the following equations are applied to the U phase, ■ phase, and W phase of the induction motor (1).

V, is applied.

V, = ffVd”, = V.

vV=v,=-JXva8.
(+) Also, the currents 1u, lv, and 1w flowing in the U phase, ■phase, and W phase of the induction motor (1) are expressed by the following equations.

iu=Fableid@,=i.

lv= iw= --rrXxd@*
(2) In this way, a single-phase power supply state can be configured.

Next, a method for identifying constants using the constant tuning device (7) will be explained in detail.

In the present invention, as shown in FIG. 2, motor constants are identified using a parallel type identifier that uses output errors in the identification algorithm. In Figure 2, the mathematical model has exactly the same shape as the induction motor (1) in a single-phase power supply state, and the output (current here) when the same input (voltage here) is applied to both.
The identifier calculates the unknown parameters such that the error of is; and the calculated unknown parameters are fed back to the mathematical model.

Once the mathematical model is determined, a constant identification method can be derived using the following steps.

(1') Determine an equivalent nonlinear feedback system.

Its configuration is shown in FIG.

(2) In the equivalent nonlinear feedback system, the linear stationary block is determined so that its transfer function is strongly real.

(3) Among the equivalent nonlinear feedback, the nonlinear block is determined so as to satisfy Boboff's integral inequality expressed by the following equation.

tV"Wdt≧-ro'""
(3) Here, r 02 is a finite positive constant.

The single-phase equivalent circuit of the induction motor (1) is shown in FIG. 4, and its state equation is expressed by the following equation.

Here, Rs, R, are the primary and secondary resistances, L, , L are the primary and secondary self-inductances, M is the mutual inductance, a = 1-M2/L, L is the leakage coefficient, 1.
．． 11 is a primary and secondary current, v3 is a primary voltage, and P=- is a differential operator.

In t (4), the detectable variables are v8 and 11,
l, is undetectable. Therefore, equation (4) is transformed as shown in the following equation.

Here, λ, = L, i, + Mi, is the primary flux linkage.

The state equation of the mathematical model corresponding to equation (5) is expressed by the following equation.

Here, the subscript represents constants and variables within the mathematical model.

However, V, which is the Lr input variable, is the same for both the actual machine and the mathematical model.

The state equation of the linear stationary block is the state equation of the actual machine (5
) is expressed by the following equation by subtracting the state equation (6) of the mathematical model.

...(7) The second and third terms on the right side of equation (7) indicate inputs that cause errors between the state variables of the actual machine and the mathematical model, and the errors between these human inputs and the state variables that can be detected. , that is −, −
1, there is the following relationship.

As mentioned earlier, the transfer function of a linear stationary block must be strongly real. In this case, its transfer function G (s
) is a part of equation (8) and the transfer function C(s) of the linear compensator
It is expressed by the following formula.

...(9) G <s) is strongly real when C(s) satisfies the following equation.

(:(s) −C+ ” Co/S
・” (1
0) However, LrGo > O, C+/Go > 1 / (R
−/σI−m”Rr/OLr).

v-C(s) (i, -1s)
= (11) Next, the nonlinear time-varying block is determined so as to satisfy Boboff's integral inequality expressed by equation (3).

V in equation (3) is expressed by equation (11), and W is expressed by the following equation from equation (8).

■, When W is expressed by equations (11) and (12), (
3) In order to satisfy equation (14), the manipulable quantities in equation (14), i.e., l/σL, , 1/Rs, R#/σLg+
Rr/σL,.

This is possible by manipulating R, Rr/σL, and L as shown in equations (13) to (16).

...(13) fr, R, R, P ...(14) ...(15) ...(16) Here, kPl ~ kP4≧0, k11 ~ is 14〉o. The reason why the latter does not include 0 is to set the steady-state deviation of constant identification to 0. Further, the subscript 0 of each constant indicates the initial value of the integrator. They may be arbitrary constants. moreover,(
14) Regarding kP2° and 12° in the formula, the product of these and σL, Lr/R, R, may be treated as kP2+k12 as in formulas (17) and (18).

kp2= (σL-L-/R-Rr) kp2'
...(17) k+z = (σL-Lr/R
mRr) k+2' (18) Summarizing the above, the identification algorithm can be determined as shown in equations (19) to (22).

...(21) ...(22) An example of a block diagram of the numerical model obtained from equation (6) is shown in FIG. i-1λ, , /R required for constant identification
, are obtained, and constants to be identified are arranged. FIG. 6 is an example of a block diagram of a constant identifier obtained from equations (19) to (22). FIG. 7 is an example of an integrator of V required for identification. Here, a first-order lag with a time constant T is used to avoid the influence of offset that occurs when an integrator, which happens to be a car, is used. Figure 1 (
The constant identification device 8) is constructed by combining these FIGS. 5, 6, and 7. The constant directly identified from equations (19) to (22) is the sum of the products of the modal constants, that is, 1
1 R, R, R, R.

There are four types: σL,, Rs, σLR2σLr, σL, and Lr, but these are commonly practiced, ~L,
By approximation, R3゜Rs, L, J
=rL,,M can be decomposed into.

Furthermore, in the case of identification based on a model norm adaptive system, the convergence of identification is compensated when one person uses one output and the human effort contains more than half of the different frequency components of those to be identified.

Therefore, ■ is 2f! It is sufficient that the waveform contains frequency components of the same type or higher, and various shapes such as a superimposed waveform of different sine waves, a square wave, a triangular wave, or other composite waveforms are possible. Alternatively, direct current components may be superimposed. This V may be used by measuring the actual inverter output voltage, or may be used as a command to the inverter.

Furthermore, instead of identifying all four at the same time, it is also possible to identify only one or only 2.3 in any combination, and not to identify the others.

If 00 in equation (lO) is O, the linear stationary block is real, but G (s) in equation (9) is ω≠0 and Ro(G(
jw) ) > O, and if it is an AC input, it can be identified, so C0=0 may be sufficient.

In addition, in the above embodiment, d is required to configure the single-phase power supply state.
I tried to match the axis, q axis, d@ axis, and q@ axis, but
A constant phase angle may be given between these two.

Furthermore, although block diagrams of the various parts constituting the constant identification device are shown in FIGS. 5, 6, and 7, it is of course possible to configure them arbitrarily without departing from the identification algorithm described in the present invention. be.

Next, an embodiment of the second invention will be described with reference to the drawings. FIG. 8 is a configuration diagram showing a constant measuring device for a conductive motor according to this embodiment. In the figure, the same reference numerals as in FIG. 1 indicate the same or corresponding parts. In the figure, (9) is voltage and current,
and a constant calculation device that calculates induction motor constants from frequency.

Next, the operation will be explained based on the above configuration.

Here, in order to make the induction motor (1) equivalent to a single-phase power supply state as in the first invention, the phase angle command θ is set to 0. Then, a predetermined value is applied to the voltage Vd'' on the d'' axis and the qa axis (here, the d axis, which coincides with the q@), which are orthogonal axes shifted by θ from the fixed orthogonal axes (d axis, q glaze). The applied LrVq'''8 is set to 0. At this time, the U phase, ■ phase, W phase of the induction motor (1)
The voltage on the phase is shown by the following formula! , , Vv, V are applied.

v,=　FHva'',=　v.

Vv=Vw=-FKVd"g
”・(la) Also, the U phase, ■ phase of the induction motor (1),
The current iLl+lv+1w flowing in the W phase is expressed by the following equation.

I u = lfd" * = 11 iv = i, = --/"7id@m
= (2a) In this way, a single-phase power supply state can be configured.

Next, the method of calculating constants by the constant calculating device (8) under this single-phase power supply state will be explained in detail.

The single-phase equivalent circuit of the induction motor (1) is shown in FIG. 4, and its state equation is expressed by the following equation.

Here, Rs, R, are the primary and secondary resistances, L,, Lr are the primary and secondary self-inductances, M is the mutual inductance, σ = 1-M''/L, and L is the leakage coefficient, i.

11 is the primary and secondary current, ■ is the primary voltage, and P=- is the differential operator.

t In equation (3a), if 18 is constant (DC flow rate), Pi, = Pi, = O, and the following equation holds true.

)1,1. = v, ”
(4a) Therefore, the primary resistance R8 can be found from equation (4a) by giving V as a constant (DC voltage) so that i is constant (DC amount). Therefore, from now on, assuming that R1 has been measured, the other motor constants, that is, the secondary resistance Rr, the primary self-sinductance, and 3,
A method for measuring the secondary self-inductance Lr% mutual inductance M will be explained.

From equation (3a), the transfer function G (s) with input V and output i becomes the following equation.

1tS◆- Here, in order to make it easier to see the future development of the equation, we will use the sum of the products of the induction motor constants, that is, R, R1-= -c, -(6a)
σL, σL.

”, -C,・” (7a) σL,L.

1・C3...(8a) σL3 At this time, equation (5a) can be expressed as the following equation.

52-(:IS+R5C2 Also, once C1, C2, and C3 are determined, R3 can be calculated as L, , R, / by using the value measured by equation (4a).
L, ,σ can be calculated.

Furthermore, by using the commonly used approximation =L, Rs
,L, =L, ,M is found.

I Rr= --Rs...(10
a) If Vt+1g is constant (DC flow rate), only R and shi will affect the relationship between vl and is as shown in equation (4a). Therefore, to obtain CI, (:2.C3, vl is given as an AC sine wave voltage. At this time, the angular frequency of V is set to ω- and Lr. is the formula (13a).

...(13a) Also, the gain is expressed by the following formula.

Now, we can use the already measured value for R8, and there are three unknowns: CI, C2, and C3, so (14a)
They can be obtained by making the equations into three-dimensional simultaneous equations. In other words, V is set to three different angular frequencies ω6-1. ω
Literary 2. ω is given as an AC voltage of 3, and the gains G, , G2 . By measuring G, the following three-dimensional simultaneous equations are established for equation (14a).

Here, when formulas (15a) to (17a) are expanded and rearranged by Lr, formulas (18a) to (20a) are obtained.

= Cl-2R-h”・(18a)-CI −
2R-C2"'(19a)-CI-2R-C2・
・' (20a) L, L.

Therefore, the conditions C1>0, C2>0, and C5>0 hold true. Under these conditions, solving equations (18a) to (20a) yields C, C2, and C3 as follows.

However, Lr...(24a) Yl--...(26a) G, G2 Y2 praise--(27a) GI G3 z, =-0 grip 〒 +ω~2...(
28a) Z2 = -0ml +ωa4 3
- (29a) (21a) - Using CI, C2, C3 found from formula (29a)), (10a) - (1
rtr, L, = Lr, M can be calculated using equation 2a).

Here, a large calculation error may occur due to the selection of the three angular frequencies of v8. Therefore, how to select the three optimal angular frequencies will be explained.

When formula (5a) is expanded into the product of l,=L, -L, Lr (proportional+integral) element, and first-order lag element, formula (30a) is obtained.

2σ L 2σ L...
・(30a) However, Lr C= , +R, -4σR,R.

G1(s
), G2 (s), Gs (s), then
The soresolenocainto time constant is as follows.

G (s) = G, (s)・G2 (s)・G3
(S) 1”T2S 1+T3S However, Lr C12Rr When these gain curves are drawn, it becomes as shown in Fig. 9.G
(s) are ωr =1/T+, ω2 =1/T2, respectively. ω3=1/T3
A gain curve 201oglG+ with a corner frequency of
l.

This results in the synthesis of 201oglG21.201oglGsl.

Now, divide the gain curve of G (s) into regions sandwiched by each corner angular frequency, and divide each region by ω C3 (・1/T3) ... I
Area ω3 (-1/T3) <ω<C1 (・1/TI)
... Bow area ωI (-1/TI) <ω〈C2 (-1
/T2) -...nrW1ffcωkuC2(・1/T
2) ...■ Area.

At this time, the gain of G (s) in each region is as follows.

■Area 201ogl Gx l~201og r” 201o
g to 2◆201ogK== 201og(to1・K2
・K3) R3 If area 01ogl Ggl 201og K+4201
ogK2+ (201ogK.

−201ogT3−201ogω = 201og (2 to 3 to K+”) −201ogTs 201ogω ... (33a) I11 area 201ogl Gg l Phantom (to 201og, +2Ql
ogT, + 201og ω) + 201og z” (20
1og to 3-201ogT33-201oω) = 201og(toI'l'[2'to 3)+2010g
(TI/T3)...(34a) ■Area 201ogIGt l'= (r+ to 201og
201ogT, +201ogω)+ (201ogK2
-201ogTz-201ogω)+ (201og, -201ogT, -201ogω) = 201og
(K+ ”KfK,)+2010g(T+/T3)-2
010gω -(35a) Therefore, in the 5 (2) region, the gain is a function only of R8. Now, R9 is (4a
) can be obtained using the formula (21a) to (29a
) are assumed to be known at the stage of calculating the unknowns C, ,C2,C. (2) The gain in the region does not include information on these unknowns, and the gain in this region cannot be used for calculations.

When two angular frequencies of v8 are selected in the regions (1) and (2), the difference between them becomes a function of ω, and the above unknown quantity does not appear there. Furthermore, when two angular frequencies are selected in the region (■), the difference in gain becomes zero and is not related to the unknown quantity.

Therefore, V used when setting up three-dimensional simultaneous equations.

It is necessary to select one each of the three angular frequencies from the regions ■, ■, and ■. In the present invention, in order to judge these inclinations, the gain from 3 to i at several angular frequencies is measured. , find l/T3. As shown in Figure 10, the angular frequencies of multiple points are selected at approximately equal intervals on the logarithmic scale, and the points where the amount of change in gain from v8 to i between each measurement point increases. I can find it.

Note that in the above embodiment, the d-axis and q-axis are made to coincide with the d@-axis and q'' axis to configure a single-phase power supply state, but a certain phase angle may be given between these two. .

Also, after giving three different angular frequencies to ■, and finding the gain from each vl to i,
) equation to form a three-dimensional simultaneous equation, the unknown quantity may be determined by a solution method other than the above embodiment, and it is my opinion that the same effect as the above embodiment can be obtained.

(Effects of the Invention) As described above, according to the first invention, an output voltage is generated from the inverter serving as the drive power source of the induction motor so as to be in a single-phase power supply state, and the output voltage and the current flowing through the conductor are connected. Since the constants of the induction motor are identified based on the model norm adaptation system using the Lr, even if the constant is unknown, automatic measurement of the stopped state of the motor can be reliably and easily performed, and self-tuning of automatically setting the constant can be easily performed. It has the effect of making

Further, according to the second invention, an output voltage is generated from an inverter serving as a drive power source for an induction motor so as to be in a single-phase power supply state, and the output voltage is generated from the voltage using the voltage, frequency, and current flowing through the induction motor. The Lr induction motor constant is obtained by calculating the gain characteristic of the current with respect to the frequency and substituting it into the equation representing the transfer function of the induction motor, so even if the constant is unknown, automatic measurement of the stopped state of the motor can be reliably and easily performed.
Furthermore, it has the effect of facilitating self-tuning in which the constants are automatically set.

[Brief explanation of drawings]

FIG. 1 is a circuit diagram showing a constant measuring device for an induction motor according to an embodiment of the first invention, FIG. 2 is a configuration diagram of an identifier, and FIG.
The figure is a block diagram of an equivalent nonlinear feedback system, Figure 4 is a single-phase equivalent circuit diagram of an induction motor, Figure 5 is a block diagram of a mathematical model, and Figure 6 is a block diagram of a constant identifier.
FIG. 7 is a block diagram of an integrator for beams, FIG. 8 is a circuit diagram showing a constant measuring device for an induction motor according to an embodiment of the second invention, FIG. 9 is a gain curve diagram of G (s), and FIG. The 1θ diagram is a diagram showing how to obtain the corner angular frequency. (1) is an induction motor, (2) is a PVIM inverter, (
3a), (3b), and (3c) are U and V, respectively.
, W-phase current detector, (4) is 3-phase/d@q@ coordinate converter, (5) is function generator, (6) is doq''/3-phase coordinate converter, (7) C constant tuning The device (8) is a constant calculation device. In the figures, the same reference numerals indicate the same or corresponding parts.

Claims (4)

[Claims] (1) In an induction motor that uses an inverter as a drive power source, generate an inverter output voltage so as to achieve single-phase power supply or an equivalent power supply state, and use that voltage and the current flowing through the induction motor to create a model norm adaptation system. 1. A method for measuring the constants of an induction motor, the method comprising identifying the constants of the induction motor based on the above criteria, and determining the primary resistance R_s, secondary resistance R_r, primary self-inductance L_s, secondary self-inductance L_r, and mutual inductance M. (2) The method for measuring constants of an induction motor according to claim 1, wherein the output voltage of the inverter used for the measurement includes two or more frequency components. (3) In an induction motor that uses an inverter as a drive power source, generate the inverter output voltage so that it is in a single-phase power supply state or an equivalent power supply state, and calculate the voltage, frequency, and current flowing through the induction motor. By substituting and calculating into the transfer function formula, the primary resistance R_s, the secondary resistance R_r,
A method for measuring constants of an induction motor, characterized by determining a primary self-inductance L_s, a secondary self-inductance L_r, and a mutual inductance M. (4) The frequency of the inverter output voltage used for the measurement is determined by determining the ratio of voltage and current at a plurality of frequencies to find a frequency that minimizes the calculation error, and using that frequency. 3. A method for measuring constants of an induction motor according to item 3.