JP2008092657A - Inductance identifying controller of permanent magnet motor and inverter module - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a controller and an inverter module for achieving a high-accuracy and fast-response torque control by identifying an inductance value of a magnet motor. <P>SOLUTION: The invention superimposes a high frequency rectangular wave signal on a current command value or a voltage command value. (1) If the high frequency rectangular wave signal is superimposed on the current instruction value, a rectangular wave signal component is extracted from a second current command value calculated so as to match a current detection value with a d-axis or q-axis current command value on which the rectangular wave signal is superimposed, a d-axis or q-axis inductance setting value is multiplied by a ratio of an absolute value of the extracted rectangular wave signal and an absolute value of the superimposed rectangular wave signal, and the inductance value of the motor is identified. (2) If the high frequency rectangular wave signal is superimposed on the voltage command value, the rectangular wave signal component is extracted from an absolute value of the rectangular wave signal superimposed on a d-axis or q-axis voltage command value and the current detection value, the absolute value of the extracted rectangular wave signal is divided, and the inductance value of the motor is identified. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an inductance identification control device and an inverter module for a permanent magnet motor.

  As a technique for identifying an inductance value of a permanent magnet motor, Japanese Patent Laid-Open No. 9-285198 discloses that a high-frequency sine wave current is superimposed on d-axis and q-axis current command values, and these current command values, current detection values, In accordance with the current deviation, the d-axis and q-axis voltage command values are calculated, the sine wave identification signal component is extracted from the current deviation by a bandpass filter, and the d-axis and q-axis inductance estimated values are calculated. A technique for correcting an inductance value used for vector control calculation by using this is described.

JP-A-9-285198

  In the method described in Japanese Patent Laid-Open No. 9-285198, since a high-frequency sine wave current is used for the inductance identification signal, it is impossible to realize a high-frequency sine current using a low-performance general-purpose microcomputer or the like. There is a concern that the identification calculation accuracy of the inductance value deteriorates.

  An object of the present invention is to provide an inductance identification control device for a permanent magnet motor that can realize torque control with high accuracy and high response even when a low-performance general-purpose microcomputer is used.

The present invention superimposes a high-frequency square wave signal (± 2 levels) on the current command value or voltage command value,
(1) When a square wave signal is superimposed on the current command value From the second current command value calculated so that the d-axis or q-axis current command value superimposed with the square wave signal matches the current detection value, a square wave is obtained. The signal component is extracted, and the ratio between the absolute value of the extracted square wave signal and the absolute value of the superimposed square wave signal is multiplied by the d-axis or q-axis inductance setting value to identify the inductance value of the motor. To do.
(2) When a square wave signal is superimposed on the voltage command value The square wave signal component is extracted from the absolute value of the square wave signal superimposed on the d-axis or q-axis voltage command value and the current detection value. Is used to identify the inductance value of the motor.

  The present invention can provide a control device and an inverter module that can realize torque control with high accuracy and high response by identifying an inductance value of a permanent magnet motor.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(First embodiment)
FIG. 1 shows a configuration example of an inductance identification control device for a permanent magnet motor according to an embodiment of the present invention.

In this embodiment, before the actual operation or immediately before the operation, the inductance calculation of the d-axis and q-axis inductances is performed to create an inductance setting table.
1 is a permanent magnet motor,
2 is a power converter that outputs a voltage proportional to the voltage command values Vu ^* , Vv ^* , and Vw ^* of a three-phase alternating current;
21 is a DC power supply,
3 is a current detector that can detect three-phase alternating currents Iu, Iv, Iw,
4 is a position detector using a resolver or encoder capable of detecting the motor position θ,
5 is a frequency calculation unit that outputs a frequency calculation value ω ₁ from the position detection value θc;
6 is a coordinate conversion unit that outputs the detected values Iuc, Ivc, Iwc of the three-phase alternating currents Iu, Iv, Iw and the detected position values θc of the motor position θ to the d-axis and q-axis detected current values Idc, Iqc;
7 shows a high-frequency square wave signal ΔId ^* , ΔIq ^* as a d-axis or q-axis current command value Id ^* ,
A square wave signal generator for outputting to at least one of Iq ^* ,
8 performs proportional / integral calculation according to the difference between the added value of the first d-axis current command value Id ^* and the square wave signal ΔId ^* given from the host and the detected d-axis current value Idc, and the proportional / integral calculation is performed. A d-axis current control calculation unit that outputs two signals of a second current command value Id ^** that is an addition value of calculation and Id ^** _ i that is an integration calculation value;
9 performs a proportional / integral calculation in accordance with the deviation between the first q-axis current command value Iq ^* and the square wave signal ΔIq ^* given from the upper level and the q-axis current detection value Iqc. A q-axis current control calculation unit that outputs two signals of a second current command value Iq ^** that is an addition value of calculation and Iq ^** _ i that is an integration calculation value;
10 second electric current command value Id ^**, Iq ^**, integral calculation value Id ^** _i, Iq ^** _I and square wave signal .DELTA.Id ^*,? Iq ^* and the vector control calculating unit 12 of the d-axis and q-axis An inductance identification calculation unit for calculating identification values Ld ^, Lq ^ of the d-axis and q-axis inductances based on the inductance setting values Ld ^* , Lq ^* ;
Reference numeral 11 denotes d-axis and q-axis current detection values Idc and Iqc (or first d-axis and q-axis current command values Id ^* and Iq ^* ) and d-axis and q-axis inductance identification values Ld ^ and Lq. An inductance setting table for outputting the d axis and q axis inductance setting values Ld ^** and Lq ^** in accordance with the current detection values Idc and Iqc.
Reference numeral 12 denotes second d-axis and q-axis current command values Id ^** and Iq ^** , a frequency calculation value ω _1, and motor constant setting values R ^* , Ld ^* , Lq ^* , Ke ^* , or an inductance setting table 11. A vector control calculation unit that outputs voltage command values Vdc ^* and Vqc ^* using Ld ^** and Lq ^** which are output values of
A coordinate conversion unit 13 outputs three-phase AC voltage command values Vu ^* , Vv ^* , and Vw ^* from the voltage command values Vdc ^* and Vdc ^* and the position detection value θc.

  First, the basic voltage and phase control method will be described.

The basic operation of voltage control is
In the d-axis and q-axis current control calculation units 8 and 9, the second current used for the vector control calculation using the first current command values Id ^* and Iq ^* and the detected current values Idc and Iqc given from the host. Command values Id ^** and Iq ^** are calculated.

The vector control calculation unit 12 uses the second current command values Id ^** and Iq ^** , the frequency calculation value ω ₁ and the set value of the motor constant, and the voltage command values Vdc ^* and Vqc ^* shown in the equation (1) ^. To control the voltage command values Vu ^* , Vv ^* and Vw ^* of the power converter.

Where R: resistance value, Ld: d-axis inductance value, Lq: q-axis inductance value,
Ke: Induced voltage coefficient, *: Set value On the other hand, in phase control, the position detector 4 such as a resolver or encoder detects the motor position θ to obtain a position detection value θc.

The frequency calculation unit 5 uses the position detection value θc to obtain the frequency command value ω _{1 according} to (Equation 2).

  The above is the basic operation of voltage control and phase control.

Next, when the “square wave signal generation unit 7”, the “inductance identification calculation unit 10”, and the “inductance setting table 11” which are features of the present invention are not provided (the motor constant setting value is R ^* ,
The control characteristics in Ld ^* , Lq ^* , Ke ^* ) will be described.

  FIG. 2 and FIG. 3 show current step response characteristics when there is an d-axis and q-axis inductance setting error in the control device of FIG.

FIG. 2 shows a case where the d-axis and q-axis inductance values Ld and Lq of the permanent magnet motor 1 coincide with the set values Ld ^* and Lq ^* set in the vector control calculation unit 12 (Ld ^* = Ld,
Lq ^* = Lq).

When the motor is operating at a constant speed, if the q-axis current command value Iq ^* is step-changed as indicated by the broken line at point A at time 0.1 [s], the q-axis current detection value Iqc is a response. It can be seen that at time 1.6 [ms], it follows quickly without overshoot.

However, in FIG. 3, which is a characteristic when the inductance values Ld and Lq of the motor do not match the set values Ld ^* and Lq ^* (Ld ^* = 0.5 × Ld, Lq ^* = 0.5 × Lq), It can be seen that the follow-up characteristic of the q-axis current detection value Iqc is deteriorated with respect to the command value Iq ^* , and overshoot is also generated.

  That is, if there is a setting error in the inductance value of the motor, there is a problem that the follow-up characteristic to the current command value is deteriorated and overshoot occurs, eventually leading to transient torque response and accuracy deterioration.

  From here, the “inductance identification principle” in which a high-frequency square wave signal is superimposed on the current command value will be described.

The square wave signal generator 7 sets the frequency of the square wave in the d-axis and q-axis current control arithmetic units 8 and 9 in order to make the motor current follow the current command value on which the high-frequency square wave signal is superimposed. The current control response frequency facr that determines the current response characteristic is smaller than several tens Hz to several hundred Hz.

  Here, the current control response frequency facr is a time for the detected current value to reach 63.2% of the command value change width when the current command value is step-changed. If this time is Tacr, facr and Tacr are in the relationship of number (3).

Further, the peak value of the square wave is ± several% of the motor rated current, and ΔId ^* and ΔIq ^* satisfying these conditions are generated.

To the first current command values Id ^* and Iq ^* given from the host, this square wave signal ΔId ^* ,
Using the new current command value on which ΔIq ^* is superimposed (added) and the current detection values Idc and Iqc, the second current command values Id ^** and Iq ^** used for the vector control calculation are calculated to the number (4). To do.

The vector control calculation unit 12 calculates voltage command values Vdc ^* and Vqc ^* represented by the number (1). Where * is the set value.

  The applied voltages Vd and Vq of the motor are expressed using motor currents Id and Iq and motor constants.

Here, ω _r is an actual rotation frequency, and ω _r and ω ₁ are consistently matched.

In addition, by the operation of the current control calculation units 8 and 9, the relationship in which the motor currents Id and Iq match the new current command values Id ^* + ΔId ^* and Iq ^* + ΔIq ^* and the right sides of the numbers (1) and (5) From the matching relationship, the output values Id ^** and Iq ^** of the current control calculation units 8 and 9 can be expressed by the number (6).

Here, the parameter sensitivity appearing in the second current command values Id ^** and Iq ^** from the medium speed to the high speed area excluding the low speed area is examined.

  From the relationship of number (7), number (8) is established.

  Then, the number (6) becomes the number (9).

Here, from that omega ₁ is large (ω ₁ ² »ω _1), it is possible to obtain the number (10).

From equation (10), it is possible to identify the d-axis inductance value Ld from the second d-axis current command value Id ^** and the q-axis inductance value Lq from the second q-axis current command value Iq ^**. It can be seen that it is.

When the d-axis inductance value Ld included in the second d-axis current command value Id ^** of the number (10) is arranged, the number (11) is obtained.

In the d-axis current control calculation 8 in FIG. 1, proportional / integral control calculation is performed, and an added value Id ^** of the proportional calculation and the integral calculation and an output value Id ^** _ i of the integral calculation are output. -(Ke-Ke ^* ) / Ld ^* or-(Ld / Ld ^* ) Id ^* included in the molecular term of the number (11) is the second d-axis current command value Id ^* shown in the number (4) ^. It becomes the steady value of the integral calculation value Id ^** _ i of ^* .

Therefore, the d-axis inductance of the motor is calculated using the value obtained by subtracting the steady value of the integral calculation Id ^** _ i from the second current command value Id ^** , the square wave signal ΔId ^*, and the d-axis inductance setting value Ld ^*. Ld is identified.

The inductance identification calculation unit 10 calculates Ld ^ from the equation (12) for the identification value of the d-axis inductance.

In the equation (12), in order to obtain the steady value of the output value Id ^** _ i of the integral operation, a first-order lag filter having a first-order lag filter time constant of Td is inserted.

Now, paying attention to the second q-axis current command value Iq ^**, and rearranging the q-axis inductance value Lq included in the second q-axis current command value Iq ^** number (10), the number (13) Get.

Here, − (Lq / Lq ^* ) Iq ^* included in the molecular term of the number (13) is a steady value of the integral term of the second q-axis current command value Iq ^** shown in the number (4).

Therefore, using the value obtained by subtracting the steady value of the integral term Iq ^** _ i from the second current command value Iq ^** , the square wave signal ΔIq ^* and the set value Lq ^* of the q-axis inductance, the q-axis inductance Lq of the motor is used. Is identified. Inductance identification calculation unit 10 calculates Lq ^ from the identification value of q-axis inductance from Equation (14).

  Here, Tq is a first-order lag filter time constant.

  This completes the explanation of the “inductance identification principle”.

  Here, the configuration of the square wave signal generator 7 which is a characteristic of the present invention will be described with reference to FIG.

In the square wave signal generator 7,
The signal ΔId ^* , where the frequency of the square wave signal is from several tens of Hz to several hundreds Hz and the peak value is ± several A
ΔIq ^* is output. There is no problem even if the frequency values, peak values, and phase differences of the signals ΔId ^* and ΔIq ^* are different.

  Next, the configuration of the inductance identification calculation unit 10 will be described with reference to FIGS.

  The “identification calculation of the d-axis inductance Ld” will be described with reference to FIG.

In d-axis current control calculation section 8 in FIG. 1, performs the operation of the proportional-integral control, Id ^** _i is an output value of the Id ^** and integral calculation is the sum of the proportional computation and integral computation, Input to FIG.

In FIG. 5, the signal Id ^** _ i is input to the first-order lag filter 101a of the time constant Td, the steady value of the output value of the integral operation is calculated, and after subtraction from the signal Id ^** , the absolute value calculation unit 102a input.

On the other hand, the high-frequency signal ΔId ^{* of} FIG. 4 is also input to the absolute value calculation unit 103a, and a constant 104a that is a set value Ld ^* of the d-axis inductance is set as the ratio between the output value of the absolute value calculation unit 102a and the output value of 103a. Multiplication is performed to calculate the d-axis inductance identification value Ld ^.

  Similarly, “identification calculation of q-axis inductance Lq” will be described with reference to FIG.

In the q-axis current control calculation unit 9 in FIG. 1, proportional / integral control calculation is performed, and Iq ^** , which is an addition value of proportional calculation and integral calculation, and Iq ^** _ i, which is an output value of integral calculation, Input to FIG. In FIG. 6, the signal Iq ^** _ i is input to the first-order lag filter 101b of the time constant Tq, the steady value of the output value of the integral operation is calculated, and after subtraction from the signal Iq ^** , the absolute value calculation unit 102b input.

On the other hand, the high-frequency signal ΔIq ^{* of} FIG. 4 is also input to the absolute value calculation unit 103b, and a constant 104b that is a set value Lq ^* of the q-axis inductance is set as a ratio of the output value of the absolute value calculation unit 102b and the output value of 103b. Multiplication is performed to calculate the q-axis inductance identification value Lq ^.

  Further, the configuration of the inductance setting table 11 will be described with reference to FIGS.

FIG. 7 shows a table of the relationship between the d-axis current detection value Idc and the identification value Ld ^ obtained by the inductance identification calculation unit 10.

During actual operation, the d-axis inductance setting value Ld ^** is referred to by the signal Idc (or signal Id ^* ).

  Similarly, FIG. 8 shows a table of the relationship between the q-axis current detection value Iqc and the identification value Lq ^ obtained by the inductance identification calculation unit 10.

During actual operation, the q-axis inductance set value Lq ^** is referred to by the signal Iqc (or signal Iq ^* ).

By replacing these referenced values (Ld ^** , Lq ^** ) with the set values (Ld ^* , Lq ^* ) of the vector control calculation unit 12, a highly accurate and highly responsive torque control system can be realized.

(Second embodiment)
In the first embodiment, the motor constants of the vector control calculation unit 12 are set using the output values Ld ^** and Lq ^** of the inductance setting table 11, but Ld ^** and Lq ^** are The control gains of the second d-axis and q-axis current command calculation units 8 'and 9' can be applied to the calculation.

Also in this embodiment, before the actual operation or immediately before the operation, the identification calculation of the d-axis and q-axis inductances is performed to create the inductance setting table.
FIG. 9 shows this embodiment.

  In the figure, 1-7, 10-13 and 21 are the same as those in FIG.

As shown in the equation (15), if the control gains (Kpd, Kid, Kpq, Kiq) of the current control calculation units 8 ′, 9 ′ are corrected using the inductance setting values (Ld ^** , Lq ^** ). Thus, a torque control system with higher response can be realized.

Here, Kpd: second d-axis current command value Id ^** proportional gain for calculation, Kid: integral gain, Kpq: second q-axis current command value Iq ^** proportional gain for calculation, Kiq: integral gain ,
facr: current control response frequency [Hz]

(Third embodiment)
In the first embodiment, the high-frequency square wave signal is superimposed on the first current command value. However, the high-frequency square wave signal can be superimposed on the voltage command value and applied. In this embodiment, before the actual operation or immediately before the operation, the inductance calculation of the d-axis and q-axis inductances is performed to create an inductance setting table.
FIG. 10 shows this embodiment.

In the figure, 1 to 6, 8, 9, 11 to 13, 21 are the same as those in FIG. Reference numeral 7 'denotes a square wave signal generator for outputting high-frequency square wave signals ΔVd ^* and ΔVq ^* to at least one of the d-axis or q-axis voltage command values Vd ^* and Vq ^* .

The signal ΔVd ^* , where the frequency of the square wave signal is several hundred Hz to several thousand Hz and the peak value is ± several V
ΔVq ^* is output. There is no problem even if the signals ΔVd ^* and ΔVq ^* have different frequency values, peak values, and phase differences.

Reference numeral 10 'denotes an inductance identification calculation unit that calculates the identification values Ld ^^ and Lq ^^ of the d-axis and q-axis inductances based on the square wave signals ΔVd ^* and ΔVq ^* and the current detection values Idc and Iqc.

  From here, the “inductance identification principle” in which a high-frequency square wave signal is superimposed on the voltage command value will be described.

  Since the identification of the inductance uses a high-frequency square wave signal component included in the motor current, the frequency of the square wave determines the current response characteristic set in the d-axis and q-axis current control calculation units 8 and 9. The control response frequency is greater than facr and needs to be several hundred Hz to several thousand Hz.

Further, the square wave peak value generates square wave signals ΔVd ^* and ΔVq ^* which are ± several% of the motor rated voltage.

The square voltage signals ΔVd ^* and ΔVq ^* are superimposed (added) to calculate new voltage command values Vdc ^** and Vqc ^** shown in Equation (16).

Here, since the frequency of the superimposed square wave signals ΔVd ^* and ΔVq ^* is higher than the current control response frequency facr set in the current control calculation units 8 and 9, the square wave signal component is included in the detected current values Idc and Iqc. .

  Therefore, ΔId and ΔIq of the square wave signal component are extracted using a band pass filter.

Then, from the relationship between the signal .DELTA.Vd ^* and the signal ΔId the d-axis inductance value Ld, from the relationship of the signal? Vq ^* and the signal ΔIq can be a q-axis inductance value Lq, identified.

  Inductance identification calculation section 10 'performs identification calculations shown in equations (17) and (18).

Where fcd: square wave frequency [Hz]
Next, the configuration of the inductance identification calculation unit 10 ′ will be described with reference to FIGS.

  The “identification calculation of d-axis inductance Ld” will be described with reference to FIG.

The signal ΔVd ^* is input to the absolute value calculation unit 101′a, and the signal Idc is input to the bandpass filter 102′a that detects the frequency fdc component of the square wave.

  The identification value Ld ^^ of the d-axis inductance is calculated by multiplying the ratio between the output value of the absolute value calculation unit 101'a and the output value of the bandpass filter 102'a by the detection gain Kc shown in Equation (19). .

Similarly, “identification calculation of q-axis inductance Lq” will be described with reference to FIG. The signal ΔVq ^* is input to the absolute value calculation unit 101′b, and the signal Iqc is input to the bandpass filter 102′b that detects the frequency fdc component of the square wave.

  The identification value Lq ^^ of the q-axis inductance is calculated by multiplying the ratio between the output value of the absolute value calculation unit 101'b and the output value of the bandpass filter 102'b by the detection gain Kc shown in equation (19). .

As in the first embodiment, the relationship between the current detection values Idc and Iqc and the identification values Ld ^^ and Lq ^^ obtained by the inductance identification calculation unit 10 is used as a table, and the signal Idc ( Alternatively, the d-axis inductance setting value Ld ^{** is referred} to by the signal Id ^* ), and the q-axis inductance setting value Lq ^** is referred to by the signal Iqc (or the signal Iq ^* ).

By replacing these referenced values (Ld ^** , Lq ^** ) with the set values (Ld ^* , Lq ^* ) of the vector control calculation unit 12, a highly accurate and highly responsive torque control system can be realized.

  Even in such an embodiment, it is obvious that the operation is the same as that of the above embodiment and the same effect can be obtained.

Further, if the control gains (Kpd, Kid, Kpq, Kiq) of the current control calculation units 8, 9 are corrected using the inductance setting values (Ld ^** , Lq ^** ), a torque control system with higher response can be obtained. It is also clear to realize.

(Fourth embodiment)
From the first to the third embodiment, the identification calculation of the d-axis and q-axis inductances is performed before actual operation or immediately before operation, and a table of current detection values and inductance values is created. Although the inductance value is referred to, the inductance identification calculation can be performed even during actual operation, and the vector control calculation can be performed using the identification value.
FIG. 13 shows this embodiment.

  In the figure, 1 to 10, 12, 13, and 21 are the same as those in FIG.

  In this embodiment, the vector control calculation is performed using the output signals (Ld ^^, Lq ^^) of the inductance identification calculation unit 10 directly as shown in the equation (20).

  Thus, it is clear that even in the actual operation, the inductance identification calculation is performed and the system that performs the vector control calculation using the identification value operates in the same manner as in the above-described embodiment, and the same effect can be obtained. It is.

  In this embodiment, the high-frequency square wave signal is superimposed on the first current command value. Instead, the high-frequency square wave signal is superimposed on the voltage command value used in the third embodiment. A method may be used.

  Furthermore, the control gains (Kpd, Kid, Kpq, Kiq) of the current control calculation units 8 and 9 are corrected using the inductance setting values (Ld ^, Lq ^) and (Ld ^^, Lq ^^). It is also clear that a torque control system with higher response can be realized.

(Fifth embodiment)
From the first embodiment to the fourth embodiment, an expensive current detector 3 is used to generate a three-phase alternating current Iu˜
Although it is a method for detecting Iw, it can also be applied to a control device that performs inexpensive current detection.
FIG. 14 shows this embodiment.

  In the figure, 1, 2, 4 to 13, 21 are the same as those in FIG.

14 is a current estimating unit for estimating a DC current I _DC flowing through the input bus of the power converter, the AC currents Iu 3 phase flowing to the motor 1, Iv, and Iw.

  Using the estimated current values Iu ^, Iv ^, Iw ^, the coordinate conversion unit 6 calculates current detection values Idc, Iqc for the d-axis and the q-axis. It is obvious that such a current sensorless system operates in the same manner as in the above embodiment, and the same effect can be obtained.

  In this embodiment, the high-frequency square wave signal is superimposed on the first current command value. Instead, a high-frequency square wave signal may be superimposed on the voltage command value.

Furthermore, the control gains of the current control calculation units 8 and 9 are set using the inductance set values (Ld ^** , Lq ^** ), or the inductance identification values (Ld ^, Lq ^) and (Ld ^^, Lq ^^). It is also clear that if (Kpd, Kid, Kpq, Kiq) is corrected, a torque control system with higher response can be realized.

(Sixth embodiment)
In the first embodiment, the position detector 4 detects the position of the permanent magnet motor 1, but the present invention can also be applied to a control device in which the position sensor is omitted.

  FIG. 15 shows this embodiment.

  In the figure, 1 to 3, 6 to 13 and 21 are the same as those in FIG.

15 is an axial error Δθ () between the control axis θc ^* and the magnetic flux axis θ of the motor, using the voltage command values Vdc ^* and Vqc ^* of the d axis and q axis and the detected current values Idc and Iqc of the d axis and q axis. = Θc ^* −θ) is calculated from Equation (21), and the phase error estimation unit outputs a phase error estimated value Δθc.

Reference numeral 16 denotes a frequency estimation unit that calculates the frequency estimated value ω ₁ c so that the phase error estimated value Δθc becomes zero.

  It is obvious that such a position sensorless system operates in the same manner as in the above embodiment, and the same effect can be obtained.

  In this embodiment, the high-frequency square wave signal is superimposed on the first current command value. Instead, a high-frequency square wave signal may be superimposed on the voltage command value.

Furthermore, the control gains of the current control calculation units 8 and 9 are set using the inductance set values (Ld ^** , Lq ^** ), or the inductance identification values (Ld ^, Lq ^) and (Ld ^^, Lq ^^). It is also clear that if (Kpd, Kid, Kpq, Kiq) is corrected, a torque control system with higher response can be realized.

(Seventh embodiment)
In the sixth embodiment, the method is applied to a control device in which the position sensor is omitted.

Using the output values Ld ^** and Lq ^** of the inductance setting table 11, the motor constant of the phase error estimating unit that outputs the phase error estimated value Δθc can be corrected.
FIG. 16 shows this embodiment.

  In the figure, 1 to 3, 6 to 13, 16, 17, and 21 are the same as those in FIG.

15 ′ uses the d-axis and q-axis voltage command values Vdc ^* and Vqc ^* , the d-axis and q-axis current detection values Idc and Iqc, and the output values Ld ^** and Lq ^** of the inductance setting table 11, This is a phase error estimator that estimates and calculates an axis error Δθ (= θc ^* −θ) between the control axis θc ^* and the magnetic flux axis θ of the motor from Equation (22), and outputs a phase error estimated value Δθc ′.

  It is obvious that such a position sensorless system operates in the same manner as in the above embodiment, and the same effect can be obtained.

  In this embodiment, the high-frequency square wave signal is superimposed on the first current command value. Instead, a high-frequency square wave signal may be superimposed on the voltage command value.

Furthermore, the control gains of the current control calculation units 8 and 9 are set using the inductance set values (Ld ^** , Lq ^** ), or the inductance identification values (Ld ^, Lq ^) and (Ld ^^, Lq ^^). It is also clear that if (Kpd, Kid, Kpq, Kiq) is corrected, a torque control system with higher response can be realized.

  Alternatively, instead of the method of detecting the three-phase alternating currents Iu to Iw with the expensive current detector 3, the present invention can also be applied to a control device that performs inexpensive current detection.

(Eighth embodiment)
An example in which the present invention is applied to a module will be described with reference to FIG.

  This example shows an embodiment of the first example.

  Here, the frequency calculation unit 5, coordinate conversion unit 6, square wave signal generation unit 7, d-axis current control calculation unit 8, q-axis current control calculation unit 9, inductance identification calculation unit 10, inductance setting table 11, vector control calculation The unit 12 and the coordinate conversion unit 13 are configured using a one-chip microcomputer.

Further, the one-chip microcomputer and the power converter are housed in one module configured on the same substrate. The module here means "standardized unit",
It consists of separable hardware / software components. In addition, although it is preferable that it is comprised on the same board | substrate on manufacture, it is not limited to the same board | substrate. From this, it may be configured on a plurality of circuit boards built in the same housing.

  In other embodiments, the same configuration can be adopted.

  As described above, according to the present invention, in the permanent magnet motor vector control system, the inductance value of the motor is identified and set in the control system before the actual operation, immediately before the operation, or during the actual operation. By creating, referring to, and automatically correcting, torque control with high accuracy and high response can be realized. In addition, it is possible to provide an inductance identification control device for a permanent magnet motor that can be applied in common to a system that performs inexpensive current detection and a system that omits a position sensor.

  Moreover, since the inductance value of the motor is identified by superimposing a two-level square wave, the present invention also has an effect that it can be realized by a low-performance microcomputer as compared with the case of superimposing a sine wave. .

The block diagram of the inductance identification control apparatus of the permanent magnet motor which shows one Example of this invention. An example of the current control characteristic (Ld ^* = Ld, Lq ^* = Lq) when the inductance value of a motor and the inductance setting value set to vector calculation correspond. An example of the current control characteristic (Ld ^* = 0.5 × Ld, Lq ^* = 0.5 × Lq) when the inductance value of the motor and the inductance setting value set in the vector calculation do not match. Explanatory drawing of the square wave signal generation part 7 in the control apparatus of FIG. Explanatory drawing of the inductance identification calculating part 10 in the control apparatus of FIG. Explanatory drawing of the inductance identification calculating part 10 in the control apparatus of FIG. Explanatory drawing of the inductance setting table 11 in the control apparatus of FIG. Explanatory drawing of the inductance setting table 11 in the control apparatus of FIG. The block diagram of the inductance identification control apparatus of the permanent magnet motor which shows the other Example of this invention. The block diagram of the inductance identification control apparatus of the permanent magnet motor which shows the other Example of this invention. Explanatory drawing of the inductance identification calculating part 10 'in the control apparatus of FIG. Explanatory drawing of the inductance identification calculating part 10 'in the control apparatus of FIG. The block diagram of the inductance identification control apparatus of the permanent magnet motor which shows the other Example of this invention. The block diagram of the inductance identification control apparatus of the permanent magnet motor which shows the other Example of this invention. The block diagram of the inductance identification control apparatus of the permanent magnet motor which shows the other Example of this invention. The block diagram of the inductance identification control apparatus of the permanent magnet motor which shows the other Example of this invention. An example of the block diagram which shows embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Permanent magnet motor 2 ... Power converter 3 ... Current detector 4 ... Position detector 5 ... Frequency calculation part 6, 13 ... Coordinate conversion part 7, 7 '... Square wave signal generation part 8, 8' ... d-axis current Control calculation unit 9, 9 '... q-axis current control calculation unit 10, 10' ... Inductance identification calculation unit 11 ... Inductance setting table 12 ... Vector control calculation unit 14 ... Current estimation unit 15, 15 '... Phase error estimation unit 16 ... superimposed on the voltage command value of the square wave signal .DELTA.Vd ^* ... d-axis to be superimposed on the current command value of the square-wave signal? Iq ^* ... q axis to be superimposed on the current command value of the frequency estimation unit 17 ... phase calculating unit .DELTA.Id ^* ... d-axis Square wave signal ΔVq ^* ... Square wave signal Id ^* superimposed on q-axis voltage command value First d-axis current command value Id ^** ... Second d-axis current command value Iq ^* ... First q-axis current command value Iq ^** ... second q-axis current command value Id ^** _i ... second d Current command value Id ^** integral calculation value Iq ^** _i ... second q-axis current command value Iq ^** integral calculation value Ld ^, Ld ^^ ... d-axis inductance of the identified value Lq ^, Lq ^^ ... q-axis inductance identification values Ld ^** , Lq ^** ... d-axis and q-axis inductance values Vdc ^** ... d-axis voltage command value Vqc ^** ... q-axis voltage command value Idc ... d Axis current detection value Iqc ... q-axis current detection value θc ^* ... rotational phase command value ω ₁ ... frequency command value ω ₁ c ... frequency estimation value Δθ ... phase error values Δθc, Δθc '... phase error estimation value I _DC ... Input DC bus current detection value

Claims (19)

Vector control calculation is performed according to the current control output value, frequency calculation value, and motor constant setting value calculated so that the current detection value matches the d-axis and q-axis current command values given from the host. In the control device of the permanent magnet motor that calculates and controls the output voltage value of the power converter,
A magnet characterized by superimposing a square wave signal on the current command value and identifying the inductance value of the motor from the relationship between the output value of the current control and the square wave signal or the relationship between the square wave signal and the current detection value Motor inductance identification control device. In claim 1,
A second current command value calculated by superimposing the square wave signal on at least one of the first d-axis or q-axis current command value given from the host and matching the current command value with the detected current value The square wave signal component is extracted from the signal, and the ratio between the absolute value of the extracted square wave signal and the absolute value of the superimposed square wave signal is multiplied by the d-axis or q-axis inductance setting value to obtain the d-axis or q-axis. A magnet motor inductance identification control device for identifying an inductance value of a shaft motor. In claim 1,
An inductance identification control device for a magnet motor, wherein a frequency value of the square wave signal is smaller than a control response frequency value of current control. In claim 1,
Before the actual operation, the inductance value identification calculation is performed to create a table of the relationship between the detected current value or the current command value and the identified inductance value,
During actual operation, at least one of d-axis and q-axis inductance values set in the vector control calculation is referred to from a table in accordance with the current detection value or current command value, and the inductance identification control of the magnet motor is characterized. apparatus. In claim 1,
Before the actual operation, the inductance value identification calculation is performed to create a table of the relationship between the detected current value or the current command value and the identified inductance value,
A magnet motor inductance identification control device that corrects at least one of the d-axis and q-axis current control gains with a value referenced from an inductance table in accordance with a detected current value or a current command value during actual operation. . In claim 1,
Using the inductance value identified during actual operation,
An inductance identification control device for a magnet motor, wherein at least one of d-axis and q-axis inductance values set in vector control calculation is directly corrected. In claim 1,
Using the inductance value identified during actual operation,
A magnet motor inductance identification control device for correcting a control gain of the current control. In claim 1,
The current detection value is
An inductance identification control device for a magnet motor, which is a current that reproduces a motor current from a DC current detection value of an input bus of a power converter. In claim 1,
The frequency calculation value is
A magnet motor characterized in that a phase error value, which is a deviation between a rotational phase command value of a power converter and a rotational phase value of a permanent magnet motor, is obtained by estimation calculation, and the estimated phase error value is calculated to be zero. Inductance identification control device. Vector control calculation is performed according to the current control output value, frequency calculation value, and motor constant setting value calculated so that the current detection value matches the d-axis and q-axis current command values given from the host. In the control device of the permanent magnet motor that calculates and controls the output voltage value of the power converter,
By superimposing a square wave signal on the voltage command value, from the relationship between the output value of the current control and the square wave signal, or from the relationship between the square wave signal and the current detection value,
An inductance identification control device for a magnet motor, characterized by identifying an inductance value of the motor. In claim 10,
The square wave signal is superimposed on at least one of the d-axis and q-axis voltage command values, and the inductance value of the motor is identified using the square wave signal and the component value of the square wave signal included in the current detection value. An inductance identification control device for a magnet motor. In claim 10,
An inductance identification control device for a magnet motor, wherein a frequency value of the square wave signal is larger than a control response frequency value of current control. In claim 10,
Before the actual operation, the inductance value identification calculation is performed to create a table of the relationship between the detected current value or the current command value and the identified inductance value,
During actual operation, at least one of d-axis and q-axis inductance values set in the vector control calculation is referred to from a table in accordance with the current detection value or current command value, and the inductance identification control of the magnet motor is characterized. apparatus. In claim 10,
Before the actual operation, the inductance value identification calculation is performed to create a table of the relationship between the detected current value or the current command value and the identified inductance value,
A magnet motor inductance identification control device that corrects at least one of the d-axis and q-axis current control gains with a value referenced from an inductance table in accordance with a detected current value or a current command value during actual operation. . In claim 10,
Using the inductance value identified during actual operation,
An inductance identification control device for a magnet motor, wherein at least one of d-axis and q-axis inductance values set in vector control calculation is directly corrected. In claim 10,
Using the inductance value identified during actual operation,
A magnet motor inductance identification control device for correcting a control gain of the current control. In claim 10,
The current detection value is
An inductance identification control device for a magnet motor, which is a current that reproduces a motor current from a DC current detection value of an input bus of a power converter. In claim 10,
The frequency calculation value is
A magnet motor characterized in that a phase error value, which is a deviation between a rotational phase command value of a power converter and a rotational phase value of a permanent magnet motor, is obtained by estimation calculation, and the estimated phase error value is calculated to be zero. Inductance identification control device. A power converter that drives a permanent magnet motor;
Vector control calculation is performed according to the current control output value, frequency calculation value, and motor constant setting value calculated so that the current detection value matches the d-axis and q-axis current command values given from the host. In an inverter module having a controller for a permanent magnet motor that calculates and controls the output voltage value of the power converter,
A high-frequency square wave signal is superimposed on the current command value or voltage command value, and the motor inductance value is identified from the relationship between the output value of the current control and the square wave signal, or the relationship between the square wave signal and the current detection value. An inverter module characterized by that.

Images