JP3316801B2 - Multiple power converter control method and multiple power converter - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multiplex power conversion device in which inverters each using a semiconductor switch element are multiplex-connected, and more specifically to improvement of an output waveform, uniformity of power consumption and loss of each inverter. The present invention relates to an improvement in switching control for realizing the switching.

[0002]

2. Description of the Related Art In order to increase the voltage and capacity of an inverter (orthogonal power converter), a multiplex inverter device in which the output sides of a plurality of inverters (hereinafter referred to as unit inverters) are connected in series has been widely known. Have been. As a control method for obtaining desired power by performing switching control of the switching elements of a plurality of unit inverters constituting such a multiplex inverter device, a literature "IEEE Transaction D
Sector Vol. 119-D, pp. 769-775, June 1999 ", various schemes have been proposed.

Here, a switching control method of a multiplex inverter device will be described by taking a multiplex inverter device for driving a three-phase motor as an example. As a control method of a three-phase motor, V / F control or vector control is widely known. In any case, a gate pulse for controlling a switch element of an inverter is generated by a PWM (pulse width modulation) control method. That is typical. That is, the voltage command value corresponding to the AC voltage supplied to the three-phase motor is compared with the triangular wave carrier, and based on the magnitude relation, a gate pulse (PWM pulse) of the switch element of each unit inverter is generated. The switching of each switch element is controlled to output a desired AC voltage. As a typical method of generating such a PWM pulse, a phase shift method and a bias method are known based on characteristics of a triangular wave carrier used for comparison with a voltage command value. These methods will be described by taking a double inverter device as an example.

In the phase shift method, two triangular wave carriers are prepared corresponding to two unit inverters of a double inverter device. These triangular wave carriers are set with a period of Tc, a peak value of ± 2 Vdc, and a phase shifted by 90 °, and compare these triangular wave carriers with a common voltage command value to generate a PWM pulse. The cycle Tc is sufficiently smaller than the cycle of the voltage command value (AC waveform) corresponding to the inverter output, and the peak value is set based on the DC power supply voltage Vdc of each unit inverter. As described above, since the phases of the triangular wave carriers corresponding to the two unit inverters are shifted by 90 °,
These two unit inverters are driven by relatively equal switching operations. As a result, since the loss generated by the elements is made uniform, there is an advantage that the heat balance between the unit inverters is maintained. Also in the case of multiplexing into two or more stages, the voltage command values are all shared, and the PWM pulse is created by shifting the phase of the triangular wave carrier.

However, in the case of the phase shift method, since the phase of the triangular wave carrier is shifted by 90 °, there is a case where the difference between the high and low voltage of the pulse becomes large in the pulse waveform of the two-phase line voltage. In addition, there is a problem that distortion of a line voltage waveform increases. Also, if the change width of the line voltage is large, the voltage jumps up due to reflection on the wiring between the inverter and the motor, which may degrade the insulation of the motor.

In the other bias system, the range of the voltage command value is divided into a plurality of regions for each of the positive side and the negative side in accordance with the number of unit inverters, and the bias of the same triangular wave carrier is changed for each of the divided regions. In this method, a PWM pulse is generated based on a comparison between a voltage command value and each triangular wave carrier. For example, in the case of a dual inverter device, the range of the voltage command value is -2Vdc to -Vdc, -Vdc to
Divided into four regions of 0, 0 to + Vdc, + Vdc to + 2Vdc, prepare four triangular wave carriers having the same phase and a peak value width of Vdc, change their bias and assign to each region, Generate a PWM pulse in comparison with the value.

In this bias method, in the same voltage region, the switching of the unit inverter corresponding to the triangular wave carrier in that region is controlled. The output waveform of the unit inverter to be switched has a uniform pulse waveform, and the output waveform of the unit inverter that has not been switched has a constant value of 0 or ± Vdc. Therefore, the change width of the line voltage with the multiplex inverter of another phase is Vdc, and there is no problem as in the phase shift method.

[0008]

However, the conventional bias-type PWM pulse generation method has the following problems because there is a bias in the switching operation between unit inverters and between switching elements.

In other words, the switching elements that perform the switching operation during the same period are concentrated on a specific unit inverter or switching element, so that the heat balance between the switching elements is lost and the load is concentrated on the specific element. is there.

In addition, since the power consumption of each unit inverter is different, the amount of harmonics generated on the power supply side is different for each unit inverter. Usually, the input DC power of the dual inverter device is such that the secondary winding is Δ winding and Y winding.
It is obtained by rectifying AC power supplied through a transformer composed of windings. This, harmonics caused by OPERATION on the power supply side of the (5,7,17,19-order component, etc.), in order to cancel between the two units inverter. However, if the power consumption of each unit inverter is different, there is a problem that the harmonic cancellation does not function well.

An object of the present invention is to solve the above-mentioned problems, that is, to provide a PWM control which maintains the power balance of each unit inverter constituting a multiplex inverter device and has a small output waveform distortion. .

[0012]

The above object can be attained by the following means. The present invention suppresses distortion of the line voltage waveform by using a triangular wave carrier having the same phase for a plurality of unit inverters constituting a multiplex inverter, basically in the same manner as in the conventional bias method. The feature of the present invention is to switch unit inverters to be switched by PWM control with a switching cycle shorter than the cycle of the voltage command value in order to equalize the power balance between the unit inverters, which is a problem of the bias method, An object of the present invention is to equalize the power between unit inverters so that the switching operation does not concentrate on a specific unit inverter.

That is, in the conventional bias method, first,
In the case of a multiplex inverter in which n unit inverters are connected in series, the range of the voltage command value is divided into n voltage regions for each of the positive side and the negative side in accordance with the voltage command value that changes in the range of ± nVdc. Minute. Of the equally divided positive side region and negative side region, the positive side region and the negative side region at mutually symmetric positions are sequentially associated with the positive side arm and the negative side arm of n unit inverters connected in series. . Then, a triangular wave carrier is set in each voltage region, a PWM pulse is generated by comparing with a voltage command value, and the gate of the switch element of the corresponding unit inverter is controlled based on the PWM pulse. Switching was concentrated only on the unit inverter in the voltage region to which the command value belongs.

[0014] Therefore, the present invention provides, in each voltage region,
The switching operation of a plurality of unit inverters connected in series is periodically switched and performed. The switching method of the unit inverter to be switched can be realized by one of the following methods.

(1) The P of each unit inverter that generates a gate pulse by comparing a triangular wave carrier with a voltage command value
This can be realized by making the same triangular wave carrier to be input to the WM modulation means, and periodically switching and inputting the voltage command value to the PWM modulation means of each unit inverter.

(2) The bias of the triangular wave carrier input to the PWM modulating means of each unit inverter is made variable in accordance with the above-mentioned voltage range, and the bias of the triangular wave carrier input to the PWM modulating means of each unit inverter is changed. This can be realized by periodically switching between unit inverters.

More specifically, in order to convert a direct current into an alternating current by performing switching control on a switch element of a multiplex inverter formed by connecting the outputs of a plurality of unit inverters in series based on an input voltage command value, divided into a plurality of voltages regions of positive and negative symmetrical combined possible range of the voltage command value to the number of said unit inverters, set the triangular wave carrier for each said voltage region, the said voltage command value each triangle
It said unit inverter by comparing the magnitude of the wave carrier
Generate multiple gate pulses to drive the
This can be realized by switching the unit inverter driven by each gate pulse at a switching cycle shorter than the cycle of the voltage command value.

Further, the multiplex power converter according to the present invention comprises:
By connecting the outputs of multiple unit inverters in series
A multiplex inverter and a switch element of the multiplex inverter
Switching control based on input voltage command value
A control device, the control device comprising: a triangular wave carrier;
The plurality of units based on the magnitude relationship with the voltage command value.
Generate gate pulses to drive each inverter
The unit to be driven by the gate pulse.
Switch the inverter shorter than the cycle of the voltage command value.
The configuration can be switched periodically. in this case,
The control device divides a range in which the voltage command value can be taken into a plurality of positive / negative symmetric voltage regions according to the number of the unit inverters, associates the unit inverter with each of the plurality of voltage regions, and Can be switched at a switching cycle shorter than the cycle of the voltage command value.

Further, the control device sets the triangular wave carrier in common for the plurality of unit inverters, and adjusts a range in which the voltage command value can be taken according to the number of the unit inverters. Area, and the unit inverter is associated with each of a pair of positive and negative voltage areas of the plurality of voltage areas, a voltage area to which the input voltage command value belongs is determined, and a zero side of the determined voltage area is determined. A value obtained by subtracting a boundary voltage from a voltage region from the voltage command value is set as a modulation voltage command value of the unit inverter related to the determined voltage region, and a magnitude relationship between the modulation voltage command value and the triangular wave carrier. A gate pulse of the switch element is generated based on the voltage, and the unit input related to a voltage region having an absolute value smaller than the determined voltage region. A gate pulse of full voltage output is generated for the inverter, and an output short-circuit gate pulse is generated for the unit inverter in the voltage region having an absolute value larger than the determined voltage region. This can be realized by switching the correspondence between the area and the unit inverter at a switching cycle shorter than the cycle of the voltage command value.

Alternatively, the control device may divide the range in which the voltage command value can be taken into a plurality of positive / negative symmetric voltage regions in accordance with the number of the unit inverters, and correspond to the plurality of voltage regions. A triangular wave carrier having a peak width corresponding to the voltage width of the voltage region is set, and a different bias amount based on the voltage width of the voltage region can be added to the triangular wave carrier. A gate pulse of a switch element of each unit inverter may be generated based on the magnitude relationship between the voltage command value and the triangular carrier while switching the bias amount in a short switching cycle.

Here, it is preferable to set the switching cycle short. If the switching period is set to be short, there is an advantage that the power balance is equalized. However, if the switching period is too short, the number of times of switching increases, and the switching loss may increase. Conversely, if the setting is long, the fluctuation of power consumption between the unit inverters increases and the power balance fluctuates, but the number of times of switching can be reduced. Practically, it is preferable to set the value to about several cycles of the triangular wave carrier cycle Tc.

The switching cycle does not have to be synchronized with the cycle of the triangular wave carrier. However, if the switching cycle is not synchronized with the cycle of the triangular wave carrier, it is impossible to completely equalize the load balance between the unit inverters. difficult. Also, depending on the switching timing, a very narrow PWM
In some cases, a pulse may be generated and the switching element may not be able to follow up. As a result, a load on the switching element may be increased, and waveform distortion may be increased.

Therefore, it is most preferable to synchronize the switching cycle with the cycle of the triangular wave carrier. In this case, it is preferable to provide a counter that repeatedly counts the half cycle of the triangular wave carrier, and set the switching cycle at a timing when the output value of this counter is an integral multiple of the triangular wave carrier cycle. Thereby, the number of times of switching can be made more uniform, and the distortion of the waveform can be reduced.

[0024]

Embodiments of the present invention will be described below with reference to the drawings. First Embodiment FIGS. 1 and 2 show a configuration of a multiplex power conversion device according to an embodiment of the present invention, and FIGS. 3 and 4 show operation waveform diagrams of respective units. FIG.
As shown in (1), this embodiment is a multiplex inverter device configured by connecting two unit inverters per phase in series. The AC power supplied from the three-phase AC power supply 1 is converted into a predetermined voltage in the transformer 2 and is also converted into a plurality of three-phase ACs having different phases, and the multiplex inverters 3 u, 3 v , 3 . 3w. The multiplex inverters 3u , 3v , and 3w each output a phase voltage of each phase, and have basically the same configuration. Therefore, the detailed configuration will be described below based on the u-phase multiplex inverter 3u. The multiplex inverter 3u is configured by connecting two unit inverters 31u and 32u in series. Unit inverter 31u,
32u has four switch elements (Qu11 to Qu14 and Qu
21-Qu24) is a full-bridge inverter using
It comprises a diode rectifier 4 and a smoothing capacitor 5. The outputs of the multiplex inverters 3u, 3v , 3w are:
It is supplied to an AC motor 6 such as an induction motor.

The multiplex inverter 3 configured as described above
The switch elements u, 3v and 3w are controlled by the inverter control device 7. The inverter control device 7 generates a triangular wave carrier generator 71, a voltage command calculator 72 for giving a voltage command value to be supplied to the AC motor, a gate pulse (PWM pulse) for each unit inverter, and a PWM output to each switch element. Modulators 73u, 73V, 73w,
Switching signal generator 77, voltage command distributors 78u, 78
V, 78w.

The voltage command values Vu *, Vv *, Vw * of each phase output from the voltage command calculator 72 are input to voltage command distributors 78u, 78V, 78w, respectively. The switching signal CNT is input from the switching signal generator 77 to the voltage command distributors 78u, 78V, 78w. The voltage command distributors 78u, 78V, 78w have the same configuration, and generate voltage command values for each unit inverter based on the input voltage command values Vu *, Vv *, Vw * and the switching signal CNT.

Further, the PWM modulators 73u, 73V, 73w
Have the same configuration, and the PWM modulator 73u includes four comparators 74 and four inverters 75.
The comparator 74 compares the values of the signals input to the + input terminal and the − input terminal, and outputs “1” when “+” is larger, and outputs “0” when the value is opposite. The comparator 75 is
Inverts and outputs the "1" and "0" signals. The triangular wave carrier et1 is input from the triangular wave generator 71 to the + input terminals of the two comparators 74, and the voltage command distributor 78 to the-input terminals.
u, voltage command values Vu1p * and Vu1np * are input, and the remaining 2
The triangular wave carrier et2 is input from the triangular wave generator 71 to the positive input terminals of the two comparators 74, and the voltage command values Vu2p * and Vu2np * are input from the voltage command distributor 78u to the negative input terminals. The outputs of the comparator 74 and the inverter 75 are
Gate pulse Gu of switch element Qu11 ~ Qu14, Qu21 ~ Qu24
11 to Gu14 and Gu21 to Gu24.

[0028] Here, an outline of the operation of the embodiment of FIG. When power is supplied from the three-phase AC power supply 1 to each of the unit inverters 31u and 32u via the transformer 2, the power is converted into a DC voltage Vdc by the diode rectifier 4 and the smoothing capacitor 5. The transformer 2 has a function of insulating between unit inverters and suppressing harmonics flowing out to a power supply. In other words, as is well known, the secondary winding of the transformer 2 has two windings of Δ and Y for each phase. Harmonic 6m ± 1st order (m
= 1, 3, 5, ...).

The unit inverters 31u and 32u output voltages Vu1 and Vu2, respectively. Switch element Q11, Q12
And the arms of switch elements Q21 and Q22 are referred to as positive arms, and the arms of switch elements Q13 and Q14 and the arms of switch elements Q23 and Q24 are referred to as negative arms. The output voltage Vu of the unit inverter 3u is the sum of the output voltages of the unit inverters 31u and 32u. Voltage command calculator 7
In 2, the voltage command value Vu * for controlling the AC motor 6
Vw * is calculated and output based on V / F control, vector control, or the like.

The voltage command distributors 78u to 78w respectively control the unit inverters 31u, 31u based on the voltage command values Vu * to Vw *.
A voltage command value to be distributed to 32u is created. The switching signal CNT output from the switching signal generator 77 switches the distribution of the voltage command value. For these,
This is a characteristic part of the present invention and will be described later in detail.

The triangular wave carriers et1 and et2 are common to each phase and each unit inverter, and both et1 and et2 change at a value from 0 to Vdc. et1 is a triangular wave carrier for the positive arm of each unit inverter, and et2 is a triangular wave carrier for the negative arm. et2 changes with a phase difference of 180 ° compared to et1.

Next, the configuration of the voltage command distributor, which is one of the features of the present invention, will be described together with the operation, taking the voltage command distributor 78u as an example. As shown in FIG. 2, the voltage command distributor 78u includes a voltage command converter 8 and a unit inverter command calculator 9. The voltage command converter 8 has a function of separating a voltage region K in which a voltage command value exists and a pulse width modulation amount x which is a component that substantially performs PWM, and includes a region discriminator 81 and an x calculator 82. Have. The area discriminator 81 has a function of calculating the value of “K” representing the voltage area of the voltage command value. That is, the input voltage command value Vu * is compared with Vdc and -Vdc respectively set in the setting unit 84 by the comparator 74, and
The value of K corresponding to the relationship of Expression 3 is output. x calculator 82
Calculates the pulse width modulation amount x, and uses an adder (subtractor) unit 83 that performs two addition (subtraction) operations on the input voltage command value Vu * and Vdc set in the setting unit 84. , And outputs the pulse width modulation amount x to the unit inverter command calculator 9 via the switch 85 which is switched according to the value of K.

When Vdc ≦ Vu *, K = 1 (Equation 1) When Vdc <Vu * <Vdc, K = 0 (Equation 2) When −Vdc ≧ Vu *, K = −1 (Equation 3) Here, the concept of K will be further described. In the case of the conventional bias method, in the case of a multiplex inverter in which two unit inverters 31u and 32u are connected in series as in the present embodiment, the voltage command value is changed in the range of ± 2 Vdc. The value range is equally divided into two voltage regions for each of the positive side and the negative side, and the positive side region and the negative side region symmetrical to each other are respectively unit inverters 31u,
The PWM pulse is generated by setting a triangular wave carrier in each voltage region and comparing the voltage with a voltage command value, in correspondence with the positive side arm and the negative side arm of 32u. Then, in order to set a triangular wave carrier in each voltage region, a bias is applied to a common triangular wave carrier corresponding to the voltage region. By the way, applying a bias to the triangular wave carrier is substantially equivalent to applying a bias to the voltage command value. Therefore, in the present embodiment, the triangular wave carrier is used in common for each unit inverter, the voltage region to which the voltage command value belongs is determined, and the pulse width modulation amount x corresponding to the voltage command value is used in accordance with the voltage region. The command value is biased, which is substantially equivalent to biasing the triangular wave carrier.

The pulse modulation amount x is a value for converting the magnitude into a pulse width by comparison with a triangular wave carrier, and can be considered as a substantial pulse modulation amount.
x is calculated by the following equation using the value of K. This calculation is realized by the x calculator 82 using the adder 83, the setter 84, and the switch 85.

When K = 1, x = Vu * −Vdc (Equation 4) When K = 0, x = Vu * (Equation 5) When K = −1 , x = Vu * + Vdc (Equation 6) Here, FIGS. 3 and 4 (a) to (c) show the relationship between K and x with respect to Vu *. K is a quantity representing the voltage region where Vu * exists. FIG. 3 shows a waveform when K = 0 and FIG. 4 shows a waveform when K = 1.

The K obtained in this manner is multiplied by Vdc by the gain 91 of the unit inverter command calculator 9, and the switching terminal "0" side of the switching unit 92a and the switching terminal "1" side of the switching unit 92c. Is input to Similarly, x is a switch 92c
Is input to the switching terminal “1” of the switching device 92d and the switching terminal “0” of the switching device 92d. "0" is input from the setting device 93 to the other terminals of the switching devices 92a to 92d. These switches 92a-
The switch 92d is switched according to “0” and “1” of the switching signal CNT output from the switching signal generator 77. The switching signal CNT is a signal that repeats the values “1” and “0” at a cycle Tcnt, as shown in FIG. Switches 92a and 92
b and the outputs of the switches 92c and 92d are added by an adder 83 to obtain voltage command values for the positive arm of the unit inverters 32u and 32u, respectively, and inverted by the inverter 76 to output a voltage command for the negative arm. As PWM respectively
The signal is input to the modulator 73u.

Thus, each unit inverter 3
The voltage command values Vu1p * and Vu2p * for 1u and 32u are as follows.

When CNT = 1, Vu1p * = x, Vu2p * = K · Vdc (Equation 7) When CNT = 0, Vu1p * = K · Vdc, Vu2p * = x (Equation 8) Note that Vu1n * and Vu2n * Are obtained by inverting Vu1p * and Vu2p *, respectively. As a result, the gate pulse becomes as shown in FIG.
~ (L). Here, since the waveforms of Gu12, Gu14, Gu22, and Gu24 are inverted of Gu11, Gu13, Gu21, and Gu23, respectively, the illustration of the waveforms is omitted.

As described above, according to this embodiment, the range ± 2 Vdc in which the voltage command value can be taken is divided into two pairs of positive / negative symmetrical voltage regions in accordance with the number of multiplexed unit inverters. The unit inverters are respectively associated with the voltage regions, and the correspondence is switched at a switching cycle shorter than the cycle of the voltage command value.

That is, the triangular wave carrier is set in common for the two unit inverters, and the range in which the voltage command value can be taken is divided into two pairs of positive and negative symmetrical voltage regions in accordance with the two unit inverters. Each unit inverter is associated, a voltage region to which the input voltage command value belongs is determined, and a value obtained by subtracting a boundary voltage between the determined voltage region and a voltage region on the zero side from the voltage command value is determined. A modulation voltage command value of the unit inverter pertaining to the voltage region is generated, and a gate pulse of the switch element is generated based on the magnitude relationship between the modulation voltage command value and the triangular wave carrier, and the absolute value is smaller than the determined voltage region. A gate pulse of full voltage output is generated for the unit inverter in the voltage range, and a gate pulse having an absolute value larger than that of the determined voltage range For unit inverter according to the region to produce a gate pulse output short circuit, you're switched in a short switching period than the period of the voltage command value corresponding relationship between the voltage region and the unit inverter.

With this configuration, according to the present embodiment, the output waveform Vu (FIG. 3 (o)) becomes equal to that of the conventional bias system, but the two unit inverters 31 are used.
Both u and 32u perform switching almost equally, and the power balance is maintained.

FIG. 4 shows the case where the voltage command value is in the region of K = 1, but comparing Vu of FIG. 3 with FIG.
Although the magnitude of the voltage command value is different, the phase of the pulse has the same relationship as that of the conventional bias method, so that the distortion of the line voltage is suppressed. This will be further described with reference to FIGS. FIGS. 5A and 5B schematically show line voltage waveforms between the u and v phases of the present embodiment. FIG. 5A shows a u-phase and FIG. 5B shows a v-phase unit inverter. The output waveform is shown. FIG. 6 is a diagram for explaining a line voltage waveform between the u and v phases in the conventional example. As is clear from the comparison of these figures, the output waveforms Vu1, Vu2, Vv1, and Vv2 of each unit inverter according to the present embodiment are different from the conventional example of FIG. 6 in that the power balance of the unit inverter is kept uniform. In addition, every inverter is performing a switching operation. The voltage waveforms Vu and Vv obtained by adding the outputs are the same as those in FIG. Therefore, the waveform of the line voltage Vuv becomes the same as that of the conventional example of FIG. 6, and the distortion of the line voltage is suppressed.

Here, when the cycle Tcnt of the switching signal CNT is set to be short, the power balance is equalized. However, when the cycle Tcnt is too short, the number of times of switching increases and the switching loss may increase. Conversely, if you set it longer,
Fluctuations in power consumption between unit inverters increase,
The power balance will fluctuate, but the number of switching will not increase. Both have merits and demerits, but as a practical value, it may be set to a value of about several cycles of Tc.

In this embodiment, two triangular wave carriers et1 and et2 are used for the positive arm and the negative arm, and their phases are shifted by 180 °. However, even if they are in phase, the same effect as in the present embodiment can be obtained. That is, as shown in FIG. 7, a product obtained by adding the bias of −Vdc to the triangular wave carrier et1 can be used as et2. In this case, the voltage command values Vu1p *, Vu2 in FIG.
Only p * is used, and Vu1n * and Vu2n * are not used. Further, the gate pulses Gu13, Gu14, Gu23, and Gu24 are obtained by inverting the gate pulses in the case of FIG. In this case, the waveforms of the respective parts are as shown in FIG. 8, and as is apparent from comparison with FIG. 3, exactly the same gate pulse is obtained.

As described above, by adding a bias to one triangular wave carrier, one triangular wave carrier can be commonly used. Alternatively, the triangular wave carrier may be completely the same, and one of the voltage command values (for example,
The same effect can be obtained by applying a bias to Vu1n * and Vu2n *). The point is that a common triangular wave carrier is used for each unit inverter.
As a result, the phase and the bias of the triangular wave carrier, the bias applied to the voltage command value, and the like may be changed so that the same waveform as that of the conventional bias method is obtained. This point is exactly the same in the other embodiments described below.

(Second Embodiment) FIGS. 9 to 11 show a second embodiment of the present invention. This embodiment is an improvement of the inverter control device 7 in FIG. 1, and can be realized by replacing the inverter control device 7A in FIG. 9 with 7 in FIG. FIG.
In FIG. 2, components denoted by the same reference numerals as those in FIG. 2 have the same functional configuration, and therefore description thereof will be omitted. This embodiment is different from the embodiment shown in FIGS.
Instead of equalizing the number of switching times by changing the distribution of the voltage command value in accordance with the value of, the bias amount of the triangular wave carrier is changed to equalize the number of switching times.

The bias adder 79 outputs triangular wave carriers et1 and et2 whose phases are different from the triangular wave carrier generator 71 by 180 °.
, And a bias is added to et1 and et2 in accordance with the state of the switching signal CNT of the switching signal generator 77 to generate four new triangular wave carriers et1p, et2p, et1n, et2n.
The waveforms of the triangular carrier generated by the bias adder 79 are as shown in (e) to (h) of FIG. When the switching signal CNT of the switching signal generator 77 is “1”, et1p and et1n are Vdc
When CNT is “0”, et2p and et2n
Has a bias of Vdc. Also, the voltage command value V
u * is directly input to the PWM modulator 73u as Vu1p * and Vu2p *, and Vu1n * and Vu2n * are obtained by inverting the sign of Vu *. Similar processing is performed for the V and W phases.

On the other hand, in the PWM modulators 73Au to 73Aw, comparators
Using the inverter 74 and the inverter 75, gate pulses Gu11 to Gu23 are created. FIGS. 9 and 11 show that the voltage command value Vu * is 0 ≦ Vu * ≦ Vdc
, And operation waveforms when Vdc ≦ Vu * ≦ 2Vdc, respectively. In each figure, (i) to (l) show the gate pulses of Gu11 to Gu23, and (m) to (o) show the output voltage waveforms.

As shown in the figure, it can be seen that the gate pulses Gu11 to Gu23 to each unit inverter switch almost uniformly, and the power balance between the unit inverters is kept constant. In addition, the phase of the pulse of the output voltage waveform Vu is constant without depending on the magnitude of the voltage command value, and voltage distortion can be reduced.

In other words, according to the present embodiment, the range ± 2 Vdc in which the voltage command value can be taken is divided into a pair of positive / negative symmetrical voltage regions in accordance with the number of unit inverters, and the voltage region Triangular wave carriers et having a peak width corresponding to the voltage width of
Create 1p, et2p, and invert them, et1n,
Create et2n, switch the amount of bias between triangular wave carriers et1p and et2p at a switching cycle shorter than the period of the voltage command value, and switch the bias amount between et1n and et2n based on the magnitude relationship between the voltage command value and the triangular wave carrier. By generating the gate pulse of the switch element of each unit inverter in this way, the power balance can be kept uniform with a simple configuration.

In particular, since the determination of the voltage region K and the calculation of x, which are required in the first embodiment, are not required, the calculation load can be reduced.

(Third Embodiment) FIG. 12 shows a third embodiment of the present invention.
2 shows a configuration of a characteristic unit according to the embodiment. In the first and second embodiments, the distribution of the voltage command value or the bias amount of the triangular wave carrier is switched to equalize the number of times of switching for each unit inverter. Here, if the switching cycle Tcnt is about several times the cycle of the triangular wave carrier, a certain degree of balance is ensured, but it is difficult to strictly maintain the power balance. Also, depending on the switching timing, a very narrow pulse is generated, and the switching element may not be able to follow such a pulse. As a result, the load on the element may be increased,
The waveform distortion may be large.

The present embodiment is an improvement on these points, in which the inverter control devices 78u to 78w in FIG. 1 are replaced with those in FIG. In FIG. 12, 78Bu is an improved voltage command distributor, 8B is a voltage command converter, 81
B is a region discriminator that discriminates the region R of the voltage command value, 82B is an x calculator, 85B is a switch that switches based on the output R of the region discriminator 81B, and 9B is a voltage command to all unit inverters. A unit inverter command calculator for calculating a value; 93, a calculation unit for calculating a voltage command value for each unit inverter; 94, a signal S for outputting a signal S based on the outputs of the area discriminator 81B and the switching signal generator 77B;
A data table 95 is a C data table for outputting a signal C based on the outputs of the area discriminator 81B and the switching signal generator 77B, and 96 is a multiplier. Other parts are shown in Fig. 1,
They are the same as those with the same reference numerals 2 and 9.

The voltage command value Vu * is separated into a value R indicating a region and a pulse width modulation amount x in the voltage command converter 8B. R is a quantity representing the voltage region where Vu * exists, and is determined by the region discriminator 81B as shown in the following Expressions 9 to 12.

When Vdc ≦ Vu *, R = 1 (Equation 9) When 0 <Vu * <Vdc, R = 0 (Equation 10) When Vdc <Vu * <0, R = 2 (Equation 11) When −Vdc ≧ Vu *, R = 3 (Equation 12) The area discriminator 81B tabulates the value of R with respect to the voltage command value Vu *, and performs the above-described determination. The pulse modulation amount x is calculated by the following equation 1 using the value of R in the x calculator 82B.
Calculate as 3 to Formula 16.

When R = 1, x = Vu * -Vdc (Equation 13) When R = 0, x = Vu * (Equation 14) When R = 2, x = Vu * (Equation 15) R = 3 In this case, x = Vu * + Vdc (Equation 16) The switching signal generator 77B outputs the values of the switching signals TWC1 and TWC2 as shown in FIG. Switching signal TWC
1, TWC2 takes an integer value from 0 to 3, and repeats the up-count every half cycle Tc / 2 of the triangular wave carrier. The values of the switching signals TWC1 and TWC2 change in synchronization with the peak of the triangular wave carrier. The values of TWC1 and TWC2 are shifted by "2". Thus, the switching signals TWC1, TW
By synchronizing C2 with the triangular wave carrier, the number of switching times can be accurately equalized, and at the same time, the generation of narrow pulses can be prevented.

The S data table and the C data table output signals S and C based on the values of R and TWC1 (TWC2). C data table and S
FIGS. 14 and 15 show the contents of the data table. TW in the figure
C means TWC1 or TWC2 (the table is shared between unit inverters). According to the embodiment of FIG. 12, the voltage command value Vu1p * to the unit inverter 31u is provided.
Is as shown in Expression 17 below.

Vu1p * = S · Vdc + C · x (Equation 17) As shown in FIGS. 14 and 15, C is a value of “0” and “1”.
S takes values of “0” and “± 1”. Vu1n * is generated by inverting the sign of Vu1p *. Unit inverter 32u
, The same processing is performed, but TWC2 is used instead of TWC1 as the switching signal. Also, for the other phases, the same one as the voltage command distributor 78Bu may be used as 78Bv and 78Bw.

FIGS. 16 and 17 show operation waveforms of the voltage command distributor 78Bu when the value of Vu * is in a different voltage range.
FIG. 16 shows a waveform when Vu * exists in the region of R = 0. R = 0
In FIG. 14, it can be seen from FIG. 14 that TWC1 = 0, C = 1 when TWC1 = 0, and that S is always 0 from FIG. Therefore, from Equation 17, Vu1p * becomes “x” when TWC1 is 0 or 1, and becomes zero when TWC1 = 2 or 3 (FIG. 16 (d),
(e)). In the case of Vu2p *, the switching signal TWC2 is shifted by “2” with respect to TWC1, so that it becomes “x” when TWC1 = 2 and 3, and otherwise becomes “0” (FIG. 16 (d) ) And (g)). As a result, the gate pulse becomes as shown in FIGS.
Thus, the symmetry of the waveform to the unit inverters 31u and 32u is maintained, and the power balance is completely equalized. Further, a narrow pulse due to the switching signal does not occur.

When voltage command value Vu * is in another area (R
= 1 to 3) are shown in FIGS. From these figures, it can be seen that the use of the voltage command distributor of FIG. 12 makes it possible to equalize the power balance and prevent the generation of narrow pulses. Also, as can be seen from the output waveform Vu in each figure, the phases of the pulses are all equal (in the Tc period,
It is understood that distortion of the waveform can be suppressed.

In the present embodiment, the data table used for each unit inverter is shared and the timing of the switching signal is shifted, but the switching signal is shared and a data table is prepared for each unit inverter. The same effect can be obtained.

In the present embodiment, two stages of unit inverters are used for one phase. However, when the number of unit inverters is further increased, it can be applied by changing a part of the configuration. When the unit inverter has N stages, the voltage command value area is divided into N, and the voltage command value area R is determined. Further, N switching signals (TWC) may be provided, the maximum value may be set to 2N-1, and the relationship between TWC, R, S, and C may be newly tabulated.

As described above, by using the present embodiment, it becomes possible to realize a multiplex converter that suppresses distortion of the output waveform and maintains the power balance between the unit inverters completely completely.

(Fourth Embodiment) FIG. 16 shows a fourth embodiment of the present invention.
An embodiment will be described. It has been described that the power balance between the unit inverters is completely maintained by using the third embodiment. However, the unit inverter according to the third embodiment is similar to that shown in FIG.
And the positive arm consisting of Qu12 (or Qu21 and Qu22) and Qu13
And Qu14 (or Qu23 and Qu23) on the negative side. Taking the unit inverter 31u as an example, the relationship between the switch state (mode) and the output voltage is as follows.

(Mode 1) When Qu11 = ON, Qu12 = OFF, Qu13 = OFF, Qu14 = ON, Vu1 = V
dc (Mode 2) When Qu11 = OFF, Qu12 = ON, Qu13 = ON, Qu14 = OFF, Vu1 =-
Vdc (Mode 3) When Qu11 = OFF, Qu12 = ON, Qu13 = OFF, Qu14 = ON, Vu1 = 0 (Mode 4) When Qu11 = ON, Qu12 = OFF, Qu13 = ON, Qu14 = OFF, Vu1 = 0 Here, in both mode 3 and mode 4, the output voltage Vu1 = 0
However, the switching pattern is different. When an inductive load such as an electric motor is connected as a load of the inverter, the current continues to flow even when Vu1 = 0, so that it is desirable to use mode 3 and mode 4 as evenly as possible. If the state of Vu1 = 0 is output only in mode 3, the current flowing through Qu12 and Qu14 will be larger than the current flowing through Qu11 and Qu13 as a whole, and the loss generated by the elements will be lower two elements ( Qu
12 and Qu14). In particular, switching devices such as IGBTs often have one arm formed of one module, which causes a problem of imbalance between upper and lower losses. That is, in the first to third embodiments, only the mode 3 is used to set the output of the unit inverter to zero voltage (short circuit). For example, looking at Gu11 and Gu13 in FIGS. 16 to 19, it can be seen that the output voltage Vu1 is set to zero by setting both values to zero. The mode in which Gu11 and Gu13 are “0” is mode 3, and the above-described problem occurs.

The fourth embodiment solves such a problem. This embodiment is obtained by improving the inverter control devices 78Bu to 78Bw in the third embodiment as shown in FIG. In FIG. 20, the voltage command distributor 78Cu
Is a unit inverter command calculator 9C that calculates voltage command values for all unit inverters, and a calculation unit 93C that calculates voltage command values for each unit inverter
And a data table 94 for outputting signals Sp and Sn based on the outputs of voltage command converter 8B and switching signal generator 77C.
C, 94D, voltage command converter 8B and switching signal generator 77C
, And data tables 95C and 95D that output signals Cp and Cn based on the output of the above, and a switching signal generator 77C that outputs switching signals TWC1C and TWC2C. Other parts having the same reference numerals as those shown in FIGS. 1, 2, 9, and 12 have the same reference numerals.

Next, the operation principle of the fourth embodiment will be described. The voltage command value Vu * is separated into R and x in the voltage command converter 8B as in the third embodiment. The voltage command value of each unit inverter is calculated based on R, x, and the values of the switching signals TWC1C and TWC2C. At this time, the arithmetic unit 93C calculates an individual voltage command value for each arm for each unit inverter. In the embodiments described above, the voltage command value Vu1n * of the negative arm simply reverses the sign of the command Vu1p * on the positive side. An individual data table is used for calculating the voltage command value. DATA-Sp, DATA-Cp, DATA-Sn, DATA-Cn
Are shown in FIGS. The switching signal output from the switching signal generator 77c is shown in FIG.
become that way. The period from zero to the peak of the triangular wave carrier is counted in synchronization with the triangular wave carrier.
Repeat the up count up to. In addition, TWC2C
Is counting by 2 with respect to TWC1C.

The voltage command value to each unit inverter is:
From the values of TWC1C or TWC2C and the value of R, select the values of Sp and Cp or Sn and Cn, and calculate the voltage command value according to the following equation.

Vu1p * = Vdc · Sp + x · Cp (Equation 18) Vu1n * = Vdc · Sn + x · Cn (Equation 19) That is, the maximum value of the counter is set to 7 in comparison with the third embodiment. The difference is that the voltage command value for the negative arm is created separately from that for the positive arm. Thereby, the degree of freedom of pulse width modulation (PWM) is increased, and the number of switching modes can be increased.

FIGS. 26 to 29 show operation waveforms when the voltage command distributor 8B of FIG. 20 is used. Output voltage waveform Vu
16 correspond to FIGS. 16 to 19 in the case of the third embodiment, but it can be seen that the waveforms of the gate pulses are significantly different. In particular, focusing on Gu11 and Gu13 during the period when Vu1 is zero, mode 3 and mode 4 are alternately output and Vu1 =
It turns out that it is set to 0. This relationship does not change regardless of the area of the voltage command value.

In the present embodiment, two stages of unit inverters are used for one phase. However, even when the number of unit inverters is further increased, the present embodiment can be applied with a slight modification. In the case of N stages, the voltage command value area is divided into N to determine the voltage command value area R. Also, N switching signals (TWC) are prepared, the maximum value is set to 4N-1,
The relationship between TWC, R and Sp, Sn, Cp, Cn may be newly tabulated.

As described above, by using this embodiment, the multiplex conversion that can suppress the distortion of the output waveform, keep the power balance between the unit inverters completely uniform, and further equalize the element utilization rate inside the unit inverters. Vessel can be realized.

(Fifth Embodiment) In the third and fourth embodiments, the voltage command value is recalculated for each unit inverter to equalize the power balance. In order to realize these methods, since the entire control device is configured using a microprocessor, it can be realized on software. However, when the number of stages of the multiplexing system is increased, the load of these processes increases, which may affect the responsiveness of the motor control.

Therefore, in the fifth embodiment of the present invention, the voltage command value is used as it is, and the bias amount of the triangular wave carrier is changed to obtain the same effect as in the third and fourth embodiments.
This is a WM control method. In this case, it is necessary to provide a triangular wave carrier for each unit inverter, which complicates the hardware, but simplifies the software processing, so that a high-speed response of the motor can be realized.

FIG. 30 shows the configuration of the characteristic portion of this embodiment. The illustrated inverter control device 7E is obtained by improving the inverter control device 7A of the second embodiment.
In FIG. 30, the bias adders 79Eu to 79Ew are the improved parts, and the other components are the same as those of the previous embodiments with the same reference numerals.

The configuration of the embodiment of FIG. 30 will be described together with the operation. The triangular wave generator 71 generates two triangular wave carriers et1 and et2 whose phases are different by 180 °. In synchronization with the triangular wave carrier, the switching signal generator 77B generates switching signals TWC1C and TWC2C. These relationships are the same as in the fourth embodiment, and the relationship between the phase of the triangular wave carrier and the switching signal is as shown in FIG. In the bias adder 79Eu (79Ev, 79Ew), the voltage command value Vu * (Vv *,
Based on the magnitude of Vw *) and the value of the switching signal, a bias to be applied to the triangular wave carrier is calculated, added to the triangular wave carrier, and output. The triangular wave carrier output from the bias adder 79Eu is generated individually for each phase and for each arm of each unit inverter. As the voltage command value, the same value is used for the positive arm and inverted for the negative arm, and the voltage command distributor as in the fourth embodiment is not provided.

Here, the bias adder 79Eu, which is a feature of the present embodiment, will be described with reference to FIG. In the figure, reference numeral 97 denotes a triangular wave carrier corrector, which adds a bias necessary for the triangular wave carrier.
The triangular wave carrier corrector 97 has two data tables 94
E and 94F, and a setter 91 and an adder 83 provided correspondingly. Data table 94E (94F)
Is a data table in which the bias amounts for the switching signal TWC1C (TWC2C) and the signal R representing the area of the voltage command value are recorded. In the drawings, components denoted by the same reference numerals in other embodiments are the same components.

The bias adder 79Eu configured as described above
Then, a voltage region in which the value of Vu * exists is determined by the region discriminator 81B, and a signal R representing the result of the determination is output. This is the same as the third and fourth embodiments (see FIG. 12). The triangular wave carrier corrector 97 receives the input R and the switching signal TWC.
Based on 1C (TWC2C), signals SpE and SnE are output with reference to data tables 94E and 94F. These data tables
94E and 94F are set as shown in FIGS.
E and SnE are set to take values of “1”, “0”, “−1”, and “−2”. Therefore, the bias adder 79
According to Eu, the output triangular wave carriers et1pu and et1nu
Is et1pu = et1 + SpE · Vdc (Equation 20) et1nu = et1 + SnE · Vdc (Equation 21) 34 to 37 show operation waveform diagrams of the present embodiment. FIG. 34 shows the case where Vu * is in the region of R = 0. Same figure
Looking at (e) to (h), the voltage command value (Vu1p *, Vu1n *, etc.) of each unit inverter is the original Vu * or −Vu *, which is different from the fourth embodiment (FIG. 26). are doing. Instead,
Triangular wave carrier et1pu, et1nu, et2pu, et2nu, TWC1C
The amount of bias changes according to the value of. However, the waveform of the gate pulse and the output waveform of the inverter are the same as those in FIG.

Under the condition that Vu * shown in FIGS. 35 to 37 is in another voltage range, the voltage command value to each unit inverter is the original Vu * or -Vu *, but the output of FIGS. The same waveform as the waveform is obtained.

As described above, according to the inverter control apparatus of the present embodiment, the signal processing of the voltage command value is simplified, the operation delay can be reduced, and a control system with high response can be realized.

In the present embodiment, an embodiment having performance equivalent to that of the fourth embodiment has been described. However, if the switching signal generator is changed and the data table is rewritten, it is possible to obtain performance equivalent to the third embodiment.

Further, in the first to fifth embodiments described above, an example is described in which the unit inverter per phase is two, but the present invention is applicable to any number of stages. . That is, when the number of stages is N, in the case of the first embodiment, the value of the switching signal is changed from the binary value “0” and “1” to the N value, and the switch 92 is an N-contact switch. The content of the area discrimination and the x operation of the voltage command converter 8 is set to N
What is necessary is just to change to what matched the step. In the second embodiment, the switching signal generator 77 and the bias adder 79 may be associated with each other to generate and output a newly added triangular wave carrier for the unit inverter. In the third to fifth embodiments, the switching signal generator 77
Change the maximum value of B and 77C, the number of outputs, and the discrimination of the voltage area R according to the number of stages, and change the data table (S, C, Sp, C
This can be realized by rewriting p, Sn, Cn, etc.).

[0083]

As described above, according to the present invention, the distortion of the output waveform of the multiplex inverter device can be suppressed, and the power balance between the unit inverters can be kept uniform.

[Brief description of the drawings]

FIG. 1 shows a configuration diagram of a multiplex inverter device according to a first embodiment of the present invention.

FIG. 2 is a detailed configuration diagram of a voltage command distributor according to the first embodiment.

FIG. 3 is an example of an operation waveform diagram of the first embodiment.

FIG. 4 is another example of an operation waveform diagram of the first embodiment.

FIG. 5 is an operation waveform diagram illustrating an effect of the first exemplary embodiment of the present invention.

FIG. 6 is an operation waveform diagram of a conventional example shown corresponding to FIG.

FIG. 7 is a configuration diagram of a control device according to a modified example of the first embodiment.

FIG. 8 is an example of an operation waveform of a modification of FIG. 7;

FIG. 9 is a configuration diagram of a main part of a second embodiment of the present invention.

FIG. 10 is a diagram illustrating an example of an operation waveform according to the second embodiment.

FIG. 11 is a diagram showing another example of the operation waveform of the second embodiment.

FIG. 12 is a configuration diagram of a voltage command distributor according to a third embodiment of the present invention.

FIG. 13 is a waveform diagram illustrating a switching signal according to the third embodiment.

FIG. 14 is a diagram illustrating the contents of a data table DATA-C according to the third embodiment.

FIG. 15 is a diagram illustrating the contents of a data table DATA-S according to the third embodiment.

FIG. 16 is a diagram illustrating an example of an operation waveform according to the third embodiment.

FIG. 17 is a diagram showing another example of the operation waveform of the third embodiment.

FIG. 18 is a diagram showing still another example of the operation waveform of the third embodiment.

FIG. 19 is a diagram showing still another example of the operation waveform of the third embodiment.

FIG. 20 is a configuration diagram of a voltage command distributor according to a fourth embodiment of the present invention.

FIG. 21 is a waveform diagram illustrating a switching signal according to the fourth embodiment.

FIG. 22 is a diagram showing the contents of a data table DATA-Cp of the fourth embodiment.

FIG. 23 is a diagram showing the contents of a data table DATA-Cn of the fourth embodiment.

FIG. 24 is a diagram showing the contents of a data table DATA-Sp of the fourth embodiment.

FIG. 25 is a diagram showing the contents of a data table DATA-Sn according to the fourth embodiment.

FIG. 26 is a diagram illustrating an example of an operation waveform according to the fourth embodiment.

FIG. 27 is a diagram showing another example of the operation waveform of the fourth embodiment.

FIG. 28 is a diagram showing still another example of the operation waveform of the fourth embodiment.

FIG. 29 is a diagram showing still another example of the operation waveform of the fourth embodiment.

FIG. 30 is a configuration diagram of a main part of a control device according to a fifth embodiment of the present invention.

FIG. 31 is a detailed configuration diagram of a bias adder according to a fifth embodiment.

FIG. 32 is a diagram illustrating the contents of a data table DATA-SpE of the fifth embodiment.

FIG. 33 is a diagram illustrating the contents of a data table DATA-SnE according to the fifth embodiment.

FIG. 34 is a diagram illustrating an example of an operation waveform according to the fifth embodiment.

FIG. 35 is a diagram showing another example of the operation waveform of the fifth embodiment.

FIG. 36 is a diagram showing still another example of the operation waveform of the fifth embodiment.

FIG. 37 is a diagram showing still another example of the operation waveform of the fifth embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 3 phase AC power supply 2 Transformer 3u ~ 3w Inverter which outputs each phase voltage 31u, 32u Unit inverter 4 Diode rectifier 5 Smoothing capacitor 6 AC motor 7 Control device 71 Triangular wave carrier generator 72 Voltage command calculator 73u ~ 73w PWM modulation Unit 77 Switching signal generator 78u ~ 78w Voltage command distributor

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Daigo Kaneko 7-1-1, Omika-cho, Hitachi City, Ibaraki Prefecture Within Hitachi Research Laboratory, Hitachi, Ltd. (56) References JP-A-10-70886 (JP, A) JP-A-11-89242 (JP, A) JP-A-11-215840 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H02M 7/48

Claims (10)

(57) [Claims] A control of a multiplex power converter for converting a direct current to an alternating current by performing switching control of a switch element of a multiplex inverter formed by connecting outputs of a plurality of unit inverters in series based on an input voltage command value. In the method, a range in which the voltage command value can be taken is divided into a plurality of positive and negative symmetric voltage regions according to the number of the unit inverters, a triangular wave carrier is set for each of the voltage regions, and the voltage command value and each of the triangular wave carriers are set. A plurality of gate pulses for driving the unit inverter are generated by comparing the magnitude relations of the unit inverters, and the unit inverter to be driven by the generated gate pulses is switched at a switching cycle shorter than the cycle of the voltage command value. A method for controlling a multiplex power conversion device, comprising: 2. A control device, comprising: a multiplex inverter having outputs of a plurality of unit inverters connected in series; and a control device for performing switching control based on a voltage command value input to a switch element of the multiplex inverter. a gate driving respectively before <br/> SL plurality of units inverters based on the magnitude relationship between the voltage command value and the triangular wave carrier
A multiplex power converter, wherein a pulse is generated and the unit inverter to be driven by the gate pulse is switched at a switching cycle shorter than a cycle of the voltage command value. 3. The control device divides a range in which the voltage command value can be taken into a plurality of positive / negative symmetrical voltage regions in accordance with the number of the unit inverters.
3. The multiplexed power according to claim 2 , wherein a rear end is set, the unit inverter is associated with each of the plurality of voltage regions, and the correspondence is switched at a switching cycle shorter than a cycle of the voltage command value. 4. Conversion device. 4. A control device for performing switching control based on a magnitude relationship between a voltage command value input to a switch element of the multiplex inverter and a triangular wave carrier, the multiplex inverter comprising outputs of a plurality of unit inverters connected in series. The control device sets the triangular wave carrier in common for the plurality of unit inverters, and sets a range in which the voltage command value can be taken according to the number of the unit inverters. And the unit inverter is associated with each of a pair of positive and negative voltage regions of the plurality of voltage regions, a voltage region to which the input voltage command value belongs is determined, and a voltage on the zero side of the determined voltage region is determined. A value obtained by subtracting a boundary voltage from the voltage range from the voltage command value is used as a modulation voltage of the unit inverter related to the determined voltage range. A pressure command value, and generates a gate pulse of the switch element based on a magnitude relationship between the modulation voltage command value and the triangular wave carrier, and the unit according to a voltage region having an absolute value smaller than the determined voltage region. A gate pulse of full voltage output is generated for the inverter, and an output short-circuit gate pulse is generated for the unit inverter pertaining to a voltage region having an absolute value greater than the determined voltage region. And switching the correspondence between the unit inverter and the unit inverter at a switching cycle shorter than the cycle of the voltage command value. 5. A multiplex inverter in which outputs of a plurality of unit inverters are connected in series, and a control device for performing switching control based on a magnitude relationship between a voltage command value input to a switch element of the multiplex inverter and a triangular carrier. The control device divides a range in which the voltage command value can be taken into a plurality of positive / negative symmetric voltage regions in accordance with the number of the unit inverters, and adjusts the voltage of the voltage region in accordance with the plurality of voltage regions. A triangular wave carrier having a peak width corresponding to the width is set, and a different bias amount based on the voltage width of the voltage region can be added to the triangular wave carrier, and the bias is set at a switching cycle shorter than the cycle of the voltage command value. While switching the amount, the switch of each unit inverter is switched based on the magnitude relationship between the voltage command value and the triangular wave carrier. A multi-power conversion device for generating a gate pulse of a switch element. 6. The multiplex power conversion apparatus according to claim 2, wherein the switching cycle is synchronized with a cycle of the triangular wave carrier. 7. The counter according to claim 5, further comprising a counter for repeatedly counting a half cycle of the triangular wave carrier, wherein the switching value is set to a timing at which an output value of the counter is an integral multiple of the triangular wave carrier cycle. A multiplex power converter according to any of the preceding claims. 8. A multiplex inverter having outputs of a plurality of unit inverters connected in series, and a control device for performing switching control based on a magnitude relationship between a voltage command value input to a switch element of the multiplex inverter and a triangular wave carrier. The control device, the triangular wave carrier common to the plurality of unit inverters, and a triangular wave carrier 180 degrees out of phase corresponding to the positive arm and the negative arm of the unit inverter Setting, the range in which the voltage command value can be taken is divided into a plurality of positive / negative symmetric voltage regions according to the number of the unit inverters, and the unit inverter is provided in each of a pair of positive / negative symmetric voltage regions of the plurality of voltage regions. The voltage region to which the input voltage command value belongs is determined, and a voltage region on the zero side of the determined voltage region is determined. The value obtained by subtracting the boundary voltage from said voltage command value, the first and the modulation voltage command value, the first modulation voltage of the positive-side arm or negative side arms of said unit inverter according to the discriminated voltage regions A value obtained by inverting the polarity of the command value is set as a second modulation voltage command value of the other arm, and the positive arm is set based on the magnitude relationship between the first and second modulation voltage command values and the corresponding triangular wave carrier. And a gate pulse of the switch element of the negative arm, and the correspondence between the voltage region and the unit inverter is switched at a switching cycle shorter than the cycle of the voltage command value. 9. The apparatus according to claim 8, further comprising a counter for repeatedly counting a half cycle of the triangular wave carrier, wherein the switching cycle is set to a timing at which an output value of the counter is an integral multiple of the triangular wave carrier cycle. A multiplex power converter according to any of the preceding claims. 10. A multiplex inverter comprising outputs of a plurality of unit inverters connected in series, and a control device for performing switching control based on a magnitude relation between a voltage command value input to a switch element of the multiplex inverter and a triangular wave carrier. The control device divides a range in which the voltage command value can be taken into a plurality of positive / negative symmetric voltage regions in accordance with the number of the unit inverters, and adjusts the voltage of the voltage region in accordance with the plurality of voltage regions. A triangular wave carrier having a peak width corresponding to the width is set, and a different bias amount based on the voltage width of the voltage region can be added to the triangular wave carrier, and the bias is set at a switching cycle shorter than the cycle of the voltage command value. While switching the amount, the positive and negative of each unit inverter is determined based on the magnitude relationship between the voltage command value and the triangular wave carrier. A multiplex power converter, wherein gate pulses of switch elements of a side arm and a negative arm are respectively generated.