JPH08182347A - Current control type pwm inverter - Google Patents

Abstract

PURPOSE: To accurately compensate the distortion of the output voltage of a current control type PWM inverter caused by dead time through simple arithmetic operation. CONSTITUTION: Three-phase voltage commands in U-, V-, and W-phases are calculated from current command signals and output current signals. At the time of generating a compensating voltage for compensating the distortion caused by dead time against the voltage command, the phase compensating value θ' corresponding to dead control time is calculated in step 600 and, after the polarity, positive or negative, of a phase-compensated current command is discriminated in step 602, a distorted voltage vf is used as a compensating voltage Δvu when the discriminated polarity is positive in step 603. When the discriminated voltage is negative, another distorted voltage-vf is used as the compensating voltage Δvu in step 604. In steps 605 to 607, a V-phase compensating voltage Δvv is found in the same way as that used for U-phase and, in steps 608 to 610, a W-phase compensating voltage Δvw is found and outputted. Therefore, the compensating voltages can be easily found and the distortion of the output voltage of a current control type PWM inverter can be compensated accurately.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a current control type PWM inverter, and more particularly to a current control type PWM inverter for compensating for waveform distortion of output current when dead time is provided.

[0002]

2. Description of the Related Art In a PWM inverter, a positive side and a negative side switching elements forming the inverter are alternately conductively controlled to control an output voltage. However, since the switching element has a delay in switching due to the turn-off time, a short circuit prevention period (hereinafter referred to as dead time) is provided in actual control so that the positive side and the negative side do not conduct simultaneously. There is a problem that the output voltage of the inverter is distorted due to the influence of this dead time, and as a result, the waveform of the output current is distorted.

The current control system in the current control type inverter is usually composed of an analog circuit, and since the feedback gain can be set relatively high, distortion of the current waveform due to dead time is small. However, in recent years, with the development of microcomputers, it has become possible to perform digital processing including current control and PWM pulse generation.
The cost reduction of the inverter control circuit by digital processing is in progress. Distortion compensation of the output voltage is essential for digital current control in which the feedback gain is not too large and for analog current control aiming for higher accuracy.

Conventionally, as this type of technique, for example, there is one disclosed in Japanese Patent Publication No. 6-62579.
As shown in FIG. 20, the output current of the inverter (PWM calculator) is detected by the current sensors 81 to 83, and the phase of the detected current signal is advanced by the phase advance compensator 9 by the control dead time. According to the polarity of the compensation current signal obtained by this phase lead compensation, the voltage command is given by the compensation voltage calculators 11-13.
Is to be corrected by.

However, in this conventional technique, the polarity of the output current is obtained by detecting the actual value of the output current, but the actual current is pulsating in the case of the PWM inverter, and there is also the influence of noise. There is a problem that stable detection is difficult and correct correction cannot be performed. Further, in the above-mentioned conventional technique, since attention is paid only to the polarity of the output current at the voltage output timing, there is a problem that the voltage is not accurately corrected by the polarity of the output current.

The publication also discloses a technique as shown in FIG. Here, an operation for advancing the phase near the zero cross of the output current detection signal is performed. In FIG. 21, calculation of a current phase in the current phase calculator 14, calculation of phase lead of each phase current in the current detection signal corrector 17, calculation of current detection correction signal including calculation of amplitude of current detection correction signal, adder 18 The phase of the detection current is advanced by the sum of the detection current and the bias current (correction current) at 20. These are means newly required to perform voltage distortion compensation in consideration of control delay. As described above, the calculation of the current phase and the calculation of the amplitude of the correction signal are newly required, and the calculation is complicated. Further, the compensation voltage calculators 11 to 13 determine the detected current as three modes of positive, negative, and intermittent, and calculate the voltage command correction signal. If it is determined that the voltage is intermittent, the voltage correction signal is set to zero.

[0007]

However, in obtaining the compensation voltage as described above, the calculation for advancing the phase near the output current zero cross is complicated, and there is a problem that the calculation amount increases or the system becomes large. there were. When the current is near zero, the compensation voltage is set to zero.However, even if the current is near zero, the voltage to be compensated differs depending on the temporal relationship between the current inversion timing and the output voltage rise timing and fall timing. However, there is a problem that distortion may remain when the compensation voltage is set to zero. Therefore, the present invention corrects the voltage waveform distortion of the PWM inverter so that the output current waveform approaches a sine wave.
The purpose is to provide an inverter.

[0008]

Therefore, the invention according to claim 1 is a current control type PWM inverter in which direct current is converted into alternating current and a short-circuit prevention time is provided, a phase correction for advancing the phase of the current command signal by a predetermined value. Means and voltage compensating means for correcting the command voltage according to the polarity of the phase-corrected current command signal.

According to a second aspect of the present invention, in a current control type PWM inverter that converts direct current into alternating current and provides a short circuit prevention time, a phase correction means for advancing the phase of the current command signal by a predetermined value, and the corrected current. Polarity change predicting means for predicting the time when the polarity of the command signal changes, and voltage compensating means for correcting the voltage command by the predicted change time of the polarity and the polarity of the phase-corrected current command signal. The voltage compensating means corrects the voltage command according to the positional relationship with the PWM pulse at the time when the polarity of the current command signal changes.

According to a third aspect of the present invention, in a current control type PWM inverter that converts direct current into alternating current and provides a short circuit prevention time, current detecting means for detecting the output current and the direction of the output current are determined. Reference value calculation means for calculating a reference value, current direction determination means for determining the direction of the output current based on the output current and the reference value, and calculation of a compensation voltage for the command voltage based on the determined direction of the output current. The reference value for determining the direction of the output current is calculated using at least one of the amplitude, frequency and control cycle of the current command value, and the determination result of the current direction is the predetermined time. Shall be maintained within.

According to a fifth aspect of the present invention, in a current control type PWM inverter that converts direct current into alternating current and provides a short circuit prevention time, current detecting means for detecting the output current and the direction of the output current are determined. Reference value calculation means for calculating a reference value, current direction determination means for determining the direction of the output current based on the output current and the reference value, the direction of the determined output current, the reference value and the output current Compensation voltage calculation means for calculating the compensation voltage of the command voltage based on the compensation voltage calculation means calculates the difference between the reference value and the magnitude of the output current, and calculates the magnitude of the output current (comparison value 1 ) Is less than the magnitude of the reference value (comparison value 2), the timing at which the output current direction reverses is detected from the ratio of comparison value 1 and comparison value 2, and the compensation voltage for the command voltage is calculated based on this Shall do .

[0012]

According to the first aspect of the invention, the phase correcting means corrects the phase of the current command signal having a phase delay with respect to the output voltage, and the voltage compensating means detects the compensation voltage from the change in polarity of the phase corrected current command signal. To correct the command voltage. Thereby, the output current distortion due to the dead time is corrected. In the invention according to claim 2, the phase correcting means corrects the phase of the current command signal having a phase delay with respect to the output voltage,
The polarity predicting means predicts the timing with respect to the PWM pulse at the point where the polarity of the current command value changes, and the voltage compensating means calculates the correction voltage based on the positional relationship with respect to the PWM pulse at the time when the polarity changes to calculate the command voltage. Fix it.
As a result, the timing of generating the compensation voltage is accurate,
Output current distortion due to dead time is accurately corrected.

According to another aspect of the present invention, the reference value calculating means calculates the value of the output current which changes with the control delay time by using at least one of the amplitude, frequency and control cycle of the command current value, The value whose sign is changed is used as a reference value for determining the direction of the compensation voltage, and the compensation voltage calculation means gives the direction of the compensation current based on the result of comparison between the reference value and the detected current to calculate the compensation voltage. . As a result, the distortion voltage can be correctly compensated by a simple calculation. Compensation can be performed without the effects of current ripple, noise, and control dead time, and output current distortion due to dead time can be compensated for, as well as the calculation being extremely simple.

According to another aspect of the present invention, the reference value calculating means calculates the value of the output current that changes with the control delay time by using at least one of the amplitude, frequency and control cycle of the command current value, The value whose sign is changed is used as a reference value for determining the direction of the compensation voltage, and the compensation voltage calculation means calculates the ratio between the amount of change in the current and the detected current within the control cycle where the current command crosses zero, and the calculation is performed. The timing at which the direction of the current command changes is calculated from the value. As a result, the voltage due to the dead time can be correctly compensated even when the current direction is reversed within the control cycle. If this timing is earlier than the rise and fall of the output voltage, the voltage compensation value is calculated based on the current direction determined at the sampling point, and the voltage compensation is set to zero between the rise and fall of the output voltage, and the output If it is later than the rising and falling edges of the voltage, the current direction determined at the sampling point is reversed and the voltage compensation value is calculated.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment in which the present invention is applied to control of an induction motor which is a drive source of an electric vehicle will be described below with reference to the drawings. FIG. 1 shows the configuration of the first embodiment of the present invention.
The driver operates the accelerator pedal 30, and the operation amount is
It is detected by the position sensor 31. The detected values are output to the induction motor 43 as a torque command signal Te ^* and a magnetic flux command signal φ ^* . The vector control calculator 32 calculates an exciting current command id ^* , a torque current command iq ^* , and a slip frequency ωse of the induction motor 43 by a predetermined calculation from the torque command signal Te ^* and the magnetic flux command signal φ ^* .

The slip frequency ωse at the adder 34 and the rotation speed ωre of the induction motor 43 detected by the rotation speed sensor 44.
And the primary frequency ω1 is calculated. The primary frequency ω1 is further integrated by the integrator 50 to calculate the phase reference θ. The coordinate converter 51 converts the inverter output currents iu, iv, and iw detected by the current sensors 81, 82, and 83 into the exciting current signal i of the rotating magnetic field coordinate system according to the conversion formula (1).
d and the torque current signal iq.

[Equation 1] The current control calculator 52 uses the exciting current signal i of the same rotating coordinate system.
Induction motor 43 according to d and the exciting current command signal id ^*
The voltage command vd ^* is output. The current control calculator 53 outputs the voltage command vq ^* for the induction motor 43 according to the torque current signal iq and the torque current command signal iq ^* in the same rotating coordinate system. The coordinate converter 54 calculates vd based on the phase reference θ.
^* , Vq ^* is the voltage command v of the stator coordinates from the rotating magnetic field coordinates
Convert to u1 ^* , vv1 ^* , vw1 ^* . This conversion formula is represented by formula (2).

[Equation 2]

On the other hand, the phase advance calculator 55 outputs a phase compensation value θ'advancing the phase reference θ by the control dead time Δt. The current polarity determiner 56 determines the polarity of the output current from the phase compensation value θ ′, the exciting current command value id ^* , and the torque current command iq ^* . Based on the determination result, the compensation voltage calculator 57 outputs compensation voltages (compensation amounts) Δvu, Δvv, and Δvw for compensating the output voltage distorted by the dead time. The voltage commands vu1 ^* , vv1 ^* , vw1 ^* are compensated with the compensation voltages Δvu, Δv for compensating for the distortion voltage due to dead time.
v and Δvw are added by adders 37, 38 and 39. The compensated voltage commands vu ^* , vv ^* , vw ^* are input to the PWM calculator 41 through the dead time generator 40 that adds dead time. Here, three-phase complementary PWM pulse P
u, Pu ', Pv, Pv', Pw, Pw 'are created. The power of each pulse is converted by the power converter 42 and output to the induction motor 43.

Next, a method of compensating for a distortion voltage due to dead time will be described. The output distortion voltage due to dead time is determined by the polarity of the output current.When the output current is positive, the distortion voltage of the corresponding phase is positive, when the output current is negative, the distortion voltage of the corresponding phase is negative, The absolute value of the strain voltage is almost constant regardless of the magnitude of the output current. Therefore, by compensating for the distortion voltage based on the polarity of the current command correction value that advances the phase by the amount of control dead time for the output current with a phase delay, compensation that eliminates the effects of current ripple, noise, and control dead time can be achieved. It becomes possible and the output current waveform can be approximated to a sine wave.

When determining the compensation voltage, the polarity of the current is determined by
A current command value can be used. In this, first, the phase advance calculator 55 calculates the phase compensation value θ ′ shown in the equation (3) from the phase reference θ and the primary frequency ω1.

(Equation 3) Here, Δt is the control dead time.

The current polarity determiner 56 obtains the current phase θi from the phase compensation value θ ′, the exciting current signal id ^* , and the torque current command signal iq ^* by the equation (4).

[Equation 4] And, in the case of −90 ° ≦ θi <90 °, the U-phase current is positive, otherwise the U-phase current is negative, and 30 ° ≦ θi <210 °.
, The V-phase current is positive, otherwise the V-phase current is negative, 15
When 0 ° ≦ θi <330 °, the W-phase current is positive, and otherwise the W-phase current is negative. The compensation voltage calculator 57 receives the compensation voltages Δvu, Δvv, Δ according to the output of the current polarity determiner 56.
Output vw. For example, when the U-phase current is positive, the distortion voltage + vf is the compensation voltage Δvu, and when the U-phase current is negative,
The distortion voltage −vf is the compensation voltage Δvu.

FIG. 2 shows a flow chart when the current control and PWM pulse generation in the configuration of FIG. 1 are processed by a microcomputer. First, in step 500, when a signal for updating the output of the PWM pulse is input, step 5
In 01, the output current signals iu, iv, iw detected by the current sensors 81, 82, 83 are read, and step 502
, The output current signals iu, iv,
Coordinate conversion of iw into an exciting current signal id and a torque current signal iq is performed.

In steps 503 to 505, the exciting current command signal id ^* and the torque current command signal iq ^* are read and i
Voltage commands vu1 ^* , vv1 ^* , vw1 according to d and iq
^* Is calculated. In step 506, id ^* , iq ^*
And the phase reference θ ′ of the phase lead from the compensation voltages Δvu, Δv
v and Δvw are calculated. In step 507, v
u1 ^* , vv1 ^* , and vw1 ^* are Δvu, Δvv, and Δvw
Voltage command signals vu ^* , vv ^*
vw ^* is calculated.

In step 508, the voltage command signal vu ^* ,
U, V including dead time based on vv ^* , vw ^*
W3 phase PWM signals Pu, Pu ', Pv, Pv', P
w and Pw ′ are calculated. Steps 501 and 502 correspond to the coordinate converter 51, steps 503 and 504 correspond to the current control calculators 52 and 53, and step 505 corresponds to the coordinate converter 54. Step 507 corresponds to the adders 37 to 39. Step 508 corresponds to the dead time generator 40 and the PWM calculator 41.

FIG. 3 shows details of step 506. That is, first, the phase compensation value θ ′ is calculated in step 600, and the polarity of the current command is determined in step 601, and if the determined polarity is positive in step 602, the distortion voltage vf is compensated for in step 603. Let Δvu. If the determined polarity is negative, the distortion voltage −vf is set to the compensation voltage Δvu in step 604. After this, in steps 605 to 607, the compensation voltage Δv for the V phase is the same as for the U phase.
v, in steps 608 to 610, the W-phase compensation voltage Δvw
Is requested and output. Step 600 corresponds to the phase advance calculator 55 and constitutes the phase correction means of the invention.
Step 601 corresponds to the current polarity determiner 56, and steps 602 to 601 correspond to the correction voltage calculator 57. The above 601 to 610 constitute the voltage compensating means of the invention.

The present embodiment is configured as described above, and corrects the distortion voltage due to the dead time by using the polarity of the current command value whose phase is advanced by the control dead time Δt with respect to the current command value. The output voltage distortion can be corrected without being affected by the current ripple and noise at the time of current detection, the current detection delay and the output voltage control delay. In the present embodiment, the current polarity is obtained from the phase of the current command value, but in addition to this, it is also possible to obtain it from the sign of the current command value by coordinate conversion from the rotating magnetic field coordinates to the stator coordinates. That is, according to the equation (5), the current command values iu ′ ^* , iv ′ ^* , i in which the phase delay is compensated are
Ask for w ' ^* . In this method, the respective polarities become the polarities of the current command value and the distortion of the output voltage is compensated.

(Equation 5)

FIG. 4 shows a second embodiment of the present invention. In this embodiment, a compensation voltage calculator 57A and a zero-cross prediction calculator 58 are provided in place of the compensation voltage calculator 57 in the first embodiment shown in FIG. Other configurations are first
This is the same as the embodiment. The distortion voltage due to the dead time is determined by the polarity of the output current during the dead time period, that is, the rising or falling of the PWM pulse. Therefore, by predicting the time of the zero-cross point of the current command value and compensating for the distortion voltage based on the polarity of the PWM pulse at that time, the output current waveform of the PWM pulse can be made closer to a sine wave.

The zero-cross prediction calculator 58 calculates the current phase θ.
It is determined from the i and the primary frequency ω1 whether each phase current zero crossing point occurs, and if so, the time until the zero crossing point occurs Tzu, Tzv, T based on the following equation.
Find zw. Here, 0 ≦ θi <360 °, and in the case of U phase, if 0 ≦ 270 ° −θi <ω1 · Δt, a zero cross point occurs,

(Equation 6) If 0 ≦ 90 ° −θi <ω1 · Δt, a zero cross point occurs,

(Equation 7) Becomes In the case of V phase, if 0 ≦ 30 ° −θi <ω1 · Δt, a zero cross point occurs,

(Equation 8) If 0 ≦ 210 ° −θi <ω1 · Δt, a zero cross point occurs,

[Equation 9] Becomes In case of W phase, 0 ≦ 150 ° −θi <ω1 · Δt
Then, a zero cross point occurs,

[Equation 10] If 0 ≦ 330 ° −θi <ω1 · Δt, a zero cross point occurs,

[Equation 11] Becomes

The compensation calculator 57A is a current polarity determiner 56A.
Output and the time until the zero cross point occurs Tzu, T
Compensation voltages Δvu, Δvv, Δvw depending on zv and Tzw
Is calculated and output. When the PWM pulse for one cycle is as shown in FIG. 5 and the zero-cross point is within the PWM cycle, the distortion voltage due to the dead time depends on whether the position of the zero-cross point is a, b, or c. The amounts are different as shown in FIG. The determination of a, b, and c is based on the following equation. Here, the DC power supply voltage is vdc. In case of U phase,

(Equation 12) Then, a section,

(Equation 13) Then, section b,

[Equation 14] Then, it is section c. For V phase,

(Equation 15) Then, a section,

[Equation 16] Then, section b,

[Equation 17] Then, it is section c. For W phase

(Equation 18) Then, a section,

[Formula 19] Then, section b,

(Equation 20) Then, it is section c. According to the above determination result, for example, the compensation voltage Δvu for one cycle for the U-phase current is shown in FIG.
Is determined as shown in.

FIG. 8 shows a flow chart when the output voltage distortion due to the dead time is processed by the microcomputer.
That is, in step 700, the phase compensation value θ ′ is calculated, and in step 701 the command current polarity determination and the current phase θi are performed.
When the determined polarity is positive in step 702, the process proceeds to step 703, where it is determined whether or not a zero-cross point occurs. If it occurs, in step 705, the distortion voltage −vf is set to the compensation voltage Δvu at the position a in FIG. 5 according to the position of the zero class point, 0 is set to Δvu at the position b, and 0 is set at the position c in the case of the position b. The distortion voltage + vf is Δvu. If the zero-cross point does not occur, in step 707 the distortion voltage + vf
Is the compensation voltage Δvu.

If the polarity is negative in step 702, the process proceeds to step 704, and if it is determined that a zero cross has occurred, then in step 706, if the position is the a position according to the position of the zero cross point, the distortion voltage + vf. Is Δvu, 0 at the b position is Δvu, and at the c position,
The distortion voltage −vf is Δvu. If it is determined that the distortion voltage −vf is not generated, the distortion voltage −vf is set to the compensation voltage Δv.
Let u. After that, also for the V phase and the W phase, the distortion voltage can be accurately obtained by the same method as described above, and the compensation voltage can be appropriately determined. Steps 702, 703, 704 correspond to the zero-cross prediction calculator 58, and steps 705, 70
6, 707, 708, ... Correspond to the compensation voltage calculator 57A. In this embodiment, steps 701 to 70 described above are performed.
.. constitute the voltage compensation means of the invention.

According to this embodiment, when the current detection value is detected by correcting the distortion voltage due to the dead time by using the polarity of the current command value which is advanced in phase with respect to the current command value by the control dead time. It is possible to correct the output voltage distortion that is not affected by the current ripple and noise, and the current detection delay and the voltage control delay. In addition, correct the output by correcting the current zero crossing point timing. Voltage distortion correction can be performed.

FIG. 9 shows a third embodiment of the present invention. The driver operates the accelerator pedal 30, and the operation amount is detected by the position sensor 31. The detected value is the induction motor 4
3 is output as a torque command signal Te ^* and a magnetic flux command signal φ ^* . The vector control calculator 32 calculates the exciting current command id ^* , the torque current command iq ^* , and the slip frequency ωse of the induction motor 43 by a predetermined calculation from the torque command signal Te ^* and the magnetic flux command signal φ ^* . In the adder 34, the sum of the slip frequency ωse and the rotation speed ωre of the induction motor 43 detected by the rotation speed sensor 44 is calculated to obtain 1
The next frequency ω1 is calculated.

The primary frequency ω1 is further integrated by the integrator 35 to calculate the phase reference θ. The coordinate converter 33 performs coordinate conversion on the exciting current command id ^* and the torque current iq ^* of the rotating coordinate system based on the phase reference θ, and the three-phase current command i.
Output u ^* , iv ^* , and iw ^* . The current controller 36 uses the current commands iu ^* , iv ^* , iw ^* and the current sensors 81, 8
The voltage commands vu1 ^* , vv1 ^* , vw1 ^* are calculated from the three-phase drive currents iu, iv, iw detected by 2, 83.

On the other hand, in the current direction discriminator 45, current reference values iuth, ivth, which compensate the control delay based on the exciting current command id ^* , the torque current command iq ^* , and the slip frequency ω1,
iwth is calculated. This current reference value is necessary to obtain the current direction required to create the compensation voltage. The current polarity discriminator 46 determines the current reference values iuth and ivth.
, Iwth and the detected output currents iu, iv, iw, the differences Δiu, Δiv, Δiw are calculated. The compensation voltage calculator 47 calculates compensation voltages Δuu, Δuv, Δuw to be compensated from Δiu, Δiv, Δiw.

Voltage command vu calculated by the current controller 36
Adders 37, 38, 39 to 1 ^* , vv1 ^* , vw1 ^*
Then, the compensation voltages Δvu, Δvv, and Δvw are added. The compensated voltage commands vu ^* , vv ^* , vw ^* are input to the PWM calculator 41 through the dead time generator 40 that adds dead time. Here, three-phase complementary PWM pulse P
u, Pu ', Pv, Pv', Pw, Pw 'are created. The power of each pulse is converted by the power converter 42 and output to the induction motor 43.

FIG. 10 shows one phase of the output section of the power converter 42. In the figure, the signal vupper applied to the gate of the power element A on the power source side and the signal vlower applied to the power element B on the ground side have a simultaneous off time tdt as shown in FIG. Therefore, the output voltage v of the inverter
0 (here, the output of the power converter 42) becomes a voltage that differs by the dead time depending on the direction of the output current i0 as shown in FIG. (A) shows the case where i0 is positive, and (b) shows the case where i0 is negative. To compensate for this, the magnitude of the output current or the phase at any time is not needed, only the direction of the current is sufficient. Even when the control delay is taken into consideration, only the direction of the current in which the delay is taken into consideration needs to be known.

Therefore, the value of the output current that changes with the control delay time is calculated, and the value with its sign changed is used as the reference value for determining the direction. The result of comparing the reference value with the detected current is used as a signal that gives the direction of the current. Further, in order to prevent the current direction from being erroneously determined immediately after the sign of the reference value changes, a delay is provided in the change of the reference value. As described above, it is possible to correctly perform dead time compensation in consideration of control delay with a simple calculation.

Next, the calculation of the compensation voltages Δvu, Δvv and Δvw will be described. Since the same calculation is performed for all U, V and W3 phases, only the U phase will be described here. The U-phase current command iu ^* can be expressed by equation (21).

[Equation 21] Differentiating equation (21) gives equation (22).

[Equation 22] The zero crossing of the current iu ^* is caused by ωt = nπ, (n =
..., -1, 0, 1, ...).

Therefore, the change in current in the vicinity where the current becomes zero can be expressed by equation (23).

(Equation 23) Here, if the control delay is Ts, the current that changes with the control delay time is given by equation (24).

[Equation 24] Therefore, if the magnitude relation with the detected current is discriminated using this as a reference for the current direction discrimination, the current direction reversal timing can be discriminated earlier by the control delay Ts, and the compensation voltage is obtained.

FIG. 12 is a flow chart showing the operation flow in the configuration of FIG. That is, step 100
In step 101, the detected value of the position sensor 31 is read as the torque command signal Te ^* , and in step 101, the magnetic flux command φ
^In step 102, which is read as ^* , vector control calculation is executed for the torque command signal Te ^* and the magnetic flux command φ ^*, and the exciting current command value id ^* and the torque current command value i are calculated.
q ^* and the slip frequency ωse are obtained. Step 10
3, the rotation speed ω of the induction motor 43 from the rotation speed sensor 44
Re is read, and in step 104, the primary frequency ω1 is calculated by taking the sum of the slip frequency ωse and the rotation speed ωre of the induction motor 43, and the phase reference θ is obtained by integral calculation. In step 105, the excitation current command value id ^* and the torque current command value iq ^* are coordinate-converted by the phase reference θ, and the current command values iu ^* , iv ^* ,
Calculate iw ^* . In step 106, the output currents iu, iv, iw of the induction motor 43 detected by the current sensors 81, 82, 83 are read, and step 107
Then, the voltage commands vu1 ^* , vv1 ^* , vw1 ^* are calculated from the current commands iu ^* , iv ^* , iw ^* .

In step 108, the output current direction determination reference value is calculated, in step 109 the direction of the output current is determined, and in step 110 the voltage command compensation voltages Δvu, Δvv, Δvw are calculated according to the determination result. In step 111, voltage commands vu1 ^* , vv1 ^* , vw1 ^*
Are added to the compensation voltages Δvu, Δvv, and Δvw to calculate the compensation voltage commands vu ^* , vv ^* , and vw ^* . In step 112, dead time addition calculation is performed, and step 11
In 3, PWM pulse calculation is performed based on the compensation voltage command signals vu ^* , vv ^* , vw ^* . In step 114, a drive pulse is output to the gate of the power element that performs power conversion.

FIG. 13 shows details of steps 108 to 110. First, in step 1000, the current change amount Δiumog at the current zero crossing point is obtained from the equation (25). Next, at step 1001, the delay time k · Ts until the reference value is changed after it is determined that the direction of the current is reversed from the current change amount Δiumog is calculated. Although k may be a fixed value, a value proportional to the carrier frequency ωc / 1st-order output frequency ω1 is preferable.

(Equation 25)

In step 1002, the current direction discrimination reference value iuth (n) = Δiumog (n) -2Δiumog (n−k).
-Calculate sgn. In step 1003, the difference Δiu between the current iu and the reference value iuth is obtained. From the difference, the positive / negative polarity of the current iu is determined in step 1004. If the determination result is positive, the flag indicating the current direction is set in step 1005, and sgn becomes 1. In step 1006, the compensation voltage Δvu (n) = − vdc · Ts
-The calculation of tdt is performed. If the result of step 1004 is negative, the flag is cleared and sgn becomes 0 in step 1008. In step 1009, the compensation voltage Δvu (n) = vdc · Ts · tdt is calculated. Thereafter, in step 1007, the calculated compensation voltage Δvu is output. Note that vdc is a DC power supply voltage and tdt is a dead time.

FIG. 14 shows the compensation voltage Δvu, the output current iu, and the current direction discrimination reference value iuth obtained by the above calculation.
Shows the relationship. The step 106 is the invention current detecting means, and the steps 1000 to 1002 are the reference value calculating means.
Steps 1003, 1004, 1005 and 1008 form a current direction determining means, and steps 1006 and 1009 form a compensation amount calculating means.

The present embodiment is configured as described above, and the current direction reversal timing can be estimated earlier by the control delay time Ts by obtaining the current direction determination value iuth and the compensation voltage Δvu. As a result, the voltage distortion due to the dead time can be easily and correctly compensated. In addition, the current direction reversal timing when the current changes within the control cycle is calculated, and the timing and the rise of the output voltage are calculated.
With the configuration in which the compensation voltage has an appropriate value that differs depending on the context of the fall timing, it is possible to eliminate voltage distortion due to dead time near the current zero cross. In addition, the change delay time k is added to the current direction determination reference value iuth.
Since Ts is provided, it is possible to prevent erroneous determination that the magnitude relationship between the new reference value and the detected current is reversed immediately after the current zero crossing.

Next, the current direction discriminator 45, the current polarity discriminator 46, and the compensation voltage calculator 47 of the third actual excitation shown in FIG.
Will be replaced with 45 ', 46', 47 'to explain as a fourth embodiment. In the discrete control system, only the current at the sampling point is known. In FIG. 15, the sampling points of the output current are shown by dots, but the timing (t) at which the direction of the current changes from the ratio of the amount of change Δi0 of the current and the detected current i0 within the control cycle in which the output current crosses zero
0 / Ts) can be obtained. In addition, Ts is a sampling time and t0 is an output current zero crossing time. As shown in FIG. 16, when this timing is i01 which is earlier than the rising tup and the falling tdown of the output voltage, the compensation voltage is calculated based on the current direction determined at the sampling point, and in the case of t02 between tup and tdown, Compensation voltage Δvu
Is zero and i03 is later than tup and tdown, the current direction determined at the sampling point is reversed and the compensation voltage is calculated. As a result, voltage distortion due to dead time can be correctly compensated even when the current direction is reversed within the control cycle.

FIG. 17 shows the relationship between the output voltage and the output current in the control period section in which the current direction is reversed. Output current is CA
In the case of SE1 (t01 ≧ tdown), the compensation voltage is a value obtained from the sampled current value. In the case of CASE2 (tup≤t02 <tdown), the compensation voltage is zero. This may be the case where the magnitude of the compensation voltage to be applied at the rise and fall of the output voltage is the same and the sign is opposite, or both are zero. This is because the compensation voltage may be zero when considering one control cycle. In case of CASE3 (t03 <tup),
The compensation voltage is a value obtained by inverting the sign of the compensation voltage obtained from the sampled current value. This is because the current has already been inverted when the output voltage rises and when it falls.

The control flow will be described below with reference to FIGS. The steps 2000 to 2003 have the same contents as the steps 1000 to 1003 in the third embodiment shown in FIG. That is, the current change amount Δiumog at the current zero crossing point is calculated from the equation (25), and the current change amount Δiumog is calculated.
The delay time k · Ts until the reference value is changed after it is determined that the current direction is reversed by iumog is calculated. The current direction discrimination reference value iuth (n) is calculated to obtain the difference Δiu from the output current iu.

In step 2004, the ratio Riu between the current direction determination reference value and the current output current value Δiu is calculated,
In steps 2005 and 2006, it is judged from Riu whether or not the output current is reversed within the current control cycle. If not reversed, the process proceeds to step 2011.

If it is reversed, in step 2007,
The time t0 when the output current iu becomes zero is calculated by using the equation of t0 = Rin · Ts. In step 2008, the rising timing tup and the falling timing tdown of the output are obtained using the equation (26).

(Equation 26) In steps 2009 and 2010, the temporal front-rear relationship between t0 and the rising timing tup and the falling timing tdown is obtained. When t0 is smaller than tdown and tup, the process proceeds to step 2016. When t0 is smaller than tdown and greater than tup, the process proceeds to step 2019 and t0.
When it is larger than down, the process proceeds to step 2011.

In step 2011, whether the current Δiu is positive or negative is judged. If the judgment result is positive, step 2
At 012, a flag indicating the current direction is set, and sgn becomes 1. In step 2013, the compensation voltage Δvu
(N) = vdc.Ts.tdt is calculated and the calculated value is output. If the result at step 2011 is negative, at step 2014 the flag is cleared and sg
n becomes 0. In step 2015, the compensation voltage Δvu
(N) =-vdc.Ts.tdt is calculated,
The calculated value is output.

In step 2016, the positive or negative polarity of the current Δiu is judged. If the judgment result is positive, step 20
17, the compensation voltage Δvu (n) = − vdc · Ts · t
dt is calculated and the calculated value is output. If the determination result is negative, the compensation voltage Δvu (n) = vdc · Ts
-The calculation of tdt is performed and the calculated value is output. In step 2019, the positive / negative polarity of the current Δiu is not judged,
Only the compensation voltage Δvu (n) = vdc · Ts · tdt is calculated and the calculated value is output. Here, steps 2000 to 2002 are the reference value calculation means, and step 2
003 to 2014 are current direction discriminating means, and step 20
Reference numerals 15 to 2019 make up the compensation amount calculation means.

The present embodiment is constructed as described above, and the timing at which the direction of the current changes is obtained from the ratio Δiu of the amount of change in the current and the detected current within the control cycle in which the current command crosses zero, and this timing is the output voltage. If it is earlier than the rising and falling edges of, the compensation voltage is calculated based on the current direction determined at the sampling point, the voltage compensation is set to zero between the rising and falling edges of the output voltage, and the If it is slow, the current direction determined at the sample point is reversed and the compensation voltage is calculated. As a result, the voltage due to the dead time can be correctly compensated even when the current direction is reversed within the control cycle.

Also in a discrete time system, the output time of H and L of the output voltage may be corrected. In the correction, when the current is positive, the H output time is set to the dead time t
If the current is negative, the output time of L may be shortened by tdt. Although the present embodiment has been described by using the example applied to the vector control of the induction motor, the present invention is not limited to this and can be easily applied to the control of the synchronous motor. That is, when controlling the synchronous motor, only the vector control calculator 32 that calculates the current command values id ^* , iq ^* and the slip frequency ωse is different.
This is because the other configurations are common.

[0055]

The present invention is constructed as described above, and corrects the phase of the current command signal which is phase-lagging with respect to the output voltage, and calculates the compensation voltage from the change in the polarity of the current command signal to correct the voltage command. To do. Thereby, the output current distortion due to the dead time is corrected. Output current with less noise can be obtained. Then, by correcting the voltage command by calculating the voltage correction value in consideration of the timing of the PWM pulse at the point where the polarity of the current command correction value changes, the output current distortion due to the dead time is accurately corrected and a more sinusoidal waveform is obtained. An output current close to is obtained.

Further, the value of the current that changes with the control delay time is calculated, and the value whose sign is changed is used as the reference value for determining the direction. The result of comparing this reference value with the detected current is the compensation current. By using a signal that gives a direction, the distortion voltage can be correctly compensated by a simple calculation. Compensation can be performed without the effects of current ripple, noise, and control dead time, and output current distortion due to dead time can be compensated for, as well as the calculation being extremely simple.

The timing at which the direction of the current changes is obtained from the ratio of the amount of change in the current and the detected current within the control cycle in which the current command is zero-crossed. If this timing is earlier than the fall of the output voltage, Calculate the compensation voltage based on the current direction determined at the sample point,
Within the control cycle, voltage compensation is set to zero between the rising and falling edges of the output voltage, and if it is later than the rising and falling edges of the output voltage, the compensation voltage is calculated by reversing the current direction determined at the sample point. The voltage due to the dead time can be correctly compensated even when the current direction is reversed by.

[Brief description of drawings]

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

FIG. 2 is a flow chart when a current control and PWM pulse generation in the first embodiment are processed by a microcomputer.

FIG. 3 is a flowchart showing in detail a calculation of a compensation voltage.

FIG. 4 is a block diagram showing a second embodiment of the present invention.

FIG. 5 is a diagram showing a relationship between a zero-cross point and a PWM pulse.

FIG. 6 is a diagram showing a relationship between a current polarity, a zero cross point and a strain voltage.

FIG. 7 is a diagram showing the relationship between the polarity of the U-phase current, the delay time and the strain voltage.

FIG. 8 is a flowchart in the case where distortion of output voltage due to dead time is processed by a microcomputer.

FIG. 9 is a block diagram showing a third embodiment of the present invention.

FIG. 10 shows a one-phase output section of the voltage converter.

FIG. 11 is an explanatory diagram of dead time and output voltage distortion due to dead time.

FIG. 12 is a current control and PWM in the third embodiment.
It is a flow chart at the time of processing a pulse by a microcomputer.

FIG. 13 is a flowchart showing details of calculation of a compensation voltage.

FIG. 14 is a diagram showing a relationship among a compensation voltage, an output current, and a current direction determination reference value.

FIG. 15 is an explanatory diagram of a timing when the direction of current changes.

FIG. 16 is a diagram showing a positional relationship between current direction timing and a PWM pulse.

FIG. 17 is a diagram showing a relationship between an output voltage and an output current in a control cycle section in which the current direction is reversed.

FIG. 18 is a flowchart showing the flow of calculation.

FIG. 19 is a flowchart showing the flow of calculation.

FIG. 20 is a diagram showing a conventional example.

FIG. 21 is a diagram showing a second conventional example.

[Explanation of symbols]

 1, 41 PWM calculator 2, 43 induction motor 3 frequency calculator 4, 50 integrator 5, 6, 52, 53 current control calculator 7, 33, 51, 54 coordinate converter 9 phase advance compensator 14 current phase calculator Parts 18, 19, 20, 21, 22 Adder 23, 34, 37, 38, 39 Adder 30 Accelerator pedal 31 Position sensor 32 Vector control calculator 40 Dead time generator 41 PWM calculator 42 Power converter 45 Current direction Discriminator 46 Current polarity discriminator 47 Compensation voltage calculator 55 Phase advance calculator 56 Current polarity determiner 57 Compensation voltage calculator 58 Zero-cross prediction calculator 44 Rotation speed sensor 81, 82, 83 Current sensor

Claims (6)

[Claims] 1. A current control type PWM inverter which converts direct current into alternating current and provides a short-circuit prevention time, and a phase correction means for advancing the phase of a current command signal by a predetermined value and a command based on the polarity of the phase-corrected current command signal. A current control type PWM having a voltage compensating means for correcting the voltage
Inverter. 2. In a current-controlled PWM inverter that converts direct current into alternating current and provides a short-circuit prevention time, phase correction means for advancing the phase of the current command signal by a predetermined value and the polarity of the corrected current command signal changes. And a voltage compensating means for compensating a voltage command with the predicted polarity change time and the polarity of the phase-corrected current command signal. A current control type PWM inverter, wherein the voltage command is corrected according to the positional relationship with the PWM pulse at the time when the polarity of the current command signal changes. 3. A current control type PWM inverter which converts direct current into alternating current and has a short circuit prevention time, and a current detecting means for detecting an output current, and a reference for calculating a reference value for determining the direction of the output current. Value calculation means, current direction determination means for determining the direction of the output current based on the output current and the reference value, and compensation voltage calculation means for computing the compensation voltage of the command voltage based on the determined direction of the output current. The reference value for determining the direction of the output current is calculated using at least one of the amplitude, frequency, and control cycle of the current command value, and the determination result of the current direction is maintained within a predetermined time. A current control type PWM inverter. 4. The current control type PWM inverter according to claim 3, wherein the time for which the current direction determination result is maintained is a value inversely proportional to the output frequency. 5. A current control type PWM inverter which converts direct current into alternating current and which is provided with a short circuit prevention time, a current detecting means for detecting an output current, and a reference for calculating a reference value for discriminating the direction of the output current. Value calculation means, current direction determination means for determining the direction of the output current based on the output current and the reference value, compensation voltage for the command voltage based on the determined output current direction, the reference value and the output current Compensation voltage calculation means for calculating the difference between the reference value and the magnitude of the output current, and the magnitude of the output current (comparison value 1) is equal to the reference value. When it is less than the magnitude (comparison value 2), the timing at which the output current direction is reversed is detected from the ratio of comparison value 1 and comparison value 2, and the compensation voltage of the command voltage is calculated based on this. Current control type PW M inverter. 6. The compensation voltage is a value based on the current direction at a current sampling point when the output current reversal timing is earlier than the output voltage rising timing or the falling timing within one control cycle, and the rising timing is set. 6. The value is set to zero when it is between the rising timing and the falling timing, and is set to a value determined in the direction opposite to the direction at the current sampling point when it is later than either the rising timing or the falling timing. The current control type PWM inverter described.