JPH11122944A - Controller of npc converter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To stabilize a DC voltage by a method, wherein both carrier signals with a large amplitude and with a small amplitude are used for the suppression of fluctuations of a neutral potential. SOLUTION: Stationary coordinate system voltage commands Vx and Vy which are obtained by an inverse dq converting means, digital current iu -iw from an A/D converter 2 and detected voltages Vc1 and Vc2 of capacitors, are inputted to a neutral potential control means 16 which outputs an inside hexagon apices ignition optical selection signal (I/0). By making detected respective phase detection currents flow through the A/D converting means 2, a 3- phase/2-phase converting means 6, and a dq-converting means 7 successively, d and q-axis currents id and iq are obtained. Selection signals of large amplitude and small amplitude carriers are outputted, and at the same time, neutral currents corresponding to all ignition options included in those regions are estimated and a signal approaching 0V is outputted, to suppress the neutral potential fluctuations of a single power supply structure.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an NPC converter (neutral point clamp type converter), and more particularly to a control device for an NPC converter capable of suppressing a change in neutral point potential.

[0002]

2. Description of the Related Art FIG. 28 shows a main circuit configuration of a known NPC converter. In FIG. 28, P and N are DC terminals for transmitting and receiving DC power, Cp and Cn are smoothing capacitors for dividing a DC voltage into a positive DC voltage Vcp and a negative DC voltage Vcn, and 1 is a gate turn-off thyristor. NP composed of equal switching elements and diodes
The C converter, U, V, and W are 3 for transmitting and receiving AC power.
This is a phase AC terminal.

Since the configuration and operation of the NPC converter 1 shown in FIG. 28 are well known, detailed description thereof will be omitted.

The NPC converter 1 has DC terminals P and N connected to a DC power supply (not shown), an NPC inverter for supplying AC power from AC terminals U, V and W to a load, and AC terminals U, V and W. Is connected to an AC power supply such as a commercial power supply and supplies DC power to the load from DC terminals P and N. NPC inverters and NPC converters often have the same operation and technical issues, but differ only in their names. Since the problems to be solved by the present invention are also common, they are treated as NPC converters without distinction.

FIG. 28 shows two sets of smoothing capacitors Cp and Cn.
There is also a configuration in which the neutral point O, which is a connection point, is connected as a terminal to a DC power supply or a DC load divided into two parts, but the present invention also includes such a configuration.

Incidentally, one of the problems in the control aspect of the NPC converter is a technique for suppressing a neutral point potential fluctuation. The NPC converter 1 in FIG. 28 has a basic property that the neutral point current Io in FIG. 28 fluctuates positively and negatively at a frequency three times as high as the AC side frequency even during steady operation.
The neutral point current Io is shunted to both capacitors Cp and Cn. For example, the neutral point current Io of the polarity in FIG. 28 increases the positive side DC voltage Vcp of the positive side capacitor Cp and the negative side capacitor Cn. Of the negative side DC voltage Vcn is reduced. That is, a current flows to unbalance the voltages of the capacitors Cp and Cn, and the potential at the neutral point fluctuates at a frequency three times the AC-side frequency. This potential variation becomes a disturbance element to the potential on the AC side, and causes waveform distortion of the AC output voltage and current.

In order to realize high-performance control, a change in the potential at the neutral point, that is, a voltage difference (V
cp-Vcn).

[0008]

Generally, by increasing the capacitance of the smoothing capacitors Cp and Cn of the NPC inverter, fluctuations in the neutral point potential can be suppressed. However, in this case, it is disadvantageous in reducing the size and weight of the system, and also causes a problem in cost.

As a method of positively suppressing the fluctuation of the neutral point potential, a method of adding a common compensation value to the modulation rate of each phase by a triangular wave comparison PWM, and a method of controlling by selecting a space voltage vector PWM. And the like, each of which has a suppressing effect.

However, when the power factor on the AC side is close to 0, the suppression effect is not sufficient and there remains a problem that an uncontrollable region exists. If the converter operation or the commercial power supply fluctuates, the compensation is delayed or the harmonic content of the output waveform increases, so that the DC voltage becomes unstable.

An object of the present invention is to provide an NPC having a single power supply configuration.
An object of the present invention is to provide a control device of an NPC converter that can suppress the inverter neutral point potential fluctuation and stabilize the DC voltage even when the converter operation or the commercial power supply fluctuates.

[0012]

In order to achieve the above object, an invention according to claim 1 comprises a bridge connection of a plurality of switching elements and converts AC power to DC power or DC power to AC power. Then, the DC voltage applied between the DC terminals is divided into a positive DC voltage and a negative DC voltage by two sets of smoothing capacitors, and the ignition signal obtained by comparing the modulation signal with the carrier signal is used as the switching signal. A controller for a pulse width modulation type NPC converter applied to an element, wherein a large amplitude carrier and a small amplitude carrier are used in combination as said carrier signal, and a neutral point potential fluctuation is suppressed. It is.

In order to achieve the above object, a second aspect of the present invention is to cross each phase modulated signal with a large amplitude carrier or with a small amplitude carrier according to the electrical angle and magnitude of a voltage command vector. 2. The NPC according to claim 1, further comprising: a carrier designating unit for designating the NPC.
It is a control device of the converter.

According to a third aspect of the present invention, a plurality of switching elements are bridge-connected to convert AC power into DC power or DC power into AC power, and to connect the DC power between DC terminals. In a control device of a pulse width modulation type NPC converter for dividing an applied DC voltage into a positive DC voltage and a negative DC voltage by two sets of smoothing capacitors and supplying an ignition signal to each switching element, Means for generating a system voltage command vector, two-phase to three-phase conversion means for converting the static coordinate system voltage command vector into a three-phase coordinate system voltage command signal, and the static coordinate system voltage command vector as input, In the space vector diagram subjected to the predetermined area division, each phase modulation signal is crossed with a large amplitude carrier according to the area where the voltage command vector exists,
A spatial vector region determining means for outputting a carrier selection signal and a spatial vector region number for designating whether to cross the small-amplitude carrier; a voltage command signal of the three-phase coordinate system and the region number; Modulation signal processing means for processing a coordinate system voltage command signal and outputting a modulation signal, and carrier generation means for outputting one of a large amplitude carrier and a small amplitude carrier for each phase according to the carrier selection signal. 2. The NPC converter according to claim 1, further comprising: comparison means for comparing the processed modulated signal with the carrier signal and outputting a firing signal to be applied to an NPC converter based on the comparison result. It is a control device.

In order to achieve the above object, the invention according to claim 4 is characterized in that the modulation signal processing means does not output a vector on a regular hexagonal side outside the space vector diagram of the NPC converter. Characterized by processing
A control device for an NPC converter according to claim 3.

According to a fifth aspect of the present invention, there is provided an NPC converter space vector diagram which receives the stationary coordinate system voltage command signal, the NPC converter phase output current, and the smoothing capacitor voltage as inputs. Neutral point potential control means for outputting a firing option selection signal for specifying which of the inner hexagonal vertices to select from among the output possible vectors in Comparing the processed modulated signal with the carrier signal and outputting a provisional firing signal based on the comparison result; and inputting the provisional firing signal and the firing option selection signal to the NPC conversion. 4. The control device for an NPC converter according to claim 3, further comprising: an ignition state determining means for determining an ignition signal to be given to the converter.

According to a sixth aspect of the present invention, a plurality of switching elements are connected in a bridge to convert AC power into DC power or DC power into AC power, and connect the DC power between DC terminals. In a control device of a pulse width modulation type NPC converter for dividing an applied DC voltage into a positive DC voltage and a negative DC voltage by two sets of smoothing capacitors and supplying an ignition signal to each switching element, Voltage command vector generating means for outputting a voltage command vector of the system, two-phase to three-phase conversion means for converting the static coordinate system voltage command vector into a three-phase coordinate system voltage command signal,
A carrier generating means for generating either a large vibration carrier or a small amplitude carrier for each phase, and an NPC conversion which receives the stationary coordinate system voltage command signal, the NPC converter phase output current, and the smoothing capacitor voltage as inputs, Neutral point potential control means for outputting a firing option selection signal for specifying which one of the vectors having two firing options is to be selected from among the output possible vectors in the vessel space vector diagram, Comparing means for comparing the three-phase coordinate system voltage command signal and the carrier signal and outputting a temporary firing signal of an NPC converter; and an NPC converter based on the temporary firing signal and the firing option selection signal. An ignition signal determining means for determining an ignition signal to be applied.
It is a control device of the converter.

In order to achieve the above object, the invention according to claim 7 is characterized in that, in the neutral point potential control means, all of the possible combinations of ignition options in the current control cycle are set at the present sampling time. Neutral point potential,
Predict the neutral point potential at the next sampling from the value obtained by integrating the neutral point current in the control section and the capacitor capacity,
7. The NPC converter control device according to claim 6, wherein a combination of the ignition options in which the neutral point potential at the next sampling time is closest to zero (V) is selected.

In order to achieve the above object, an invention according to claim 8 includes a bridge connection of a plurality of switching elements, wherein AC power is converted to DC power or DC power is converted to AC power, and between the DC terminals. In a control device of a pulse width modulation type NPC converter for dividing an applied DC voltage into a positive DC voltage and a negative DC voltage by two sets of smoothing capacitors and supplying an ignition signal to each switching element, Means for generating a system voltage command vector; two-phase to three-phase conversion means for converting the stationary coordinate system voltage command vector into a three-phase coordinate system voltage command signal; In the space vector diagram where the area division has been performed, whether each phase modulated signal crosses with the large amplitude carrier according to the area where the voltage command vector exists,
A space vector region determination means for outputting a carrier selection signal and a space vector region number for specifying whether to cross the small-amplitude carrier, and a voltage command signal of the three-phase coordinate system and the region number, Modulation signal processing means for processing a coordinate system voltage command signal and outputting a modulation signal, carrier generation means for outputting one of a large amplitude carrier and a small amplitude carrier for each phase according to the carrier selection signal, The static coordinate system voltage command signal, the output current of each phase of the NPC converter, and the smoothing capacitor voltage are input and NP
Neutral point potential control means for outputting a firing option selection signal for designating which of the vectors having two firing options is selected from among the output possible vectors in the C converter space vector diagram Comparing means for comparing the processed modulated signal with the carrier signal and outputting a temporary firing signal of an NPC converter; and a NPC converter based on the temporary firing signal and the firing option selection signal. A control device for an NPC converter, comprising: a firing signal determining means for determining a firing signal to be given.

To achieve the above object, the invention according to claim 9 is characterized in that in the neutral point potential control means, all the possible combinations of the ignition options in the current control cycle are set at the present sampling time. Neutral point potential,
Predict the neutral point potential at the next sampling from the value obtained by integrating the neutral point current in the control section and the capacitor capacity,
9. The NPC converter control device according to claim 8, wherein a combination of ignition options whose neutral point potential is closest to zero (V) at the next sampling time is selected.

In order to achieve the above object, the invention according to claim 10 is characterized in that either the actual current value of each phase or the current reference value of each phase is used for the neutral point potential prediction control. The control device for an NPC converter according to claim 6 or 8, wherein

In order to achieve the above object, the invention according to claim 11 is characterized in that the action of the neutral point potential control is always performed, and the ignition state is always determined based on the temporary ignition signal and the ignition option selection signal. It is determined or performed when the value of the neutral point potential at the time of the current sampling exceeds a certain allowable range. If the predicted value of the neutral point potential is likely to exceed a certain allowable level, the determination is performed, and if the estimated value is not likely to exceed the allowable range, the temporary ignition signal is set to the ignition state as it is, or the determination is performed. A control device for an NPC converter according to any one of claims 8 to 10.

In order to achieve the above object, the invention according to claim 12 is based on the principle that the carrier generation means always uses small vibration carriers in all three phases, and the neutral point at the time of the current sampling is used. If the potential value exceeds a certain allowable radiation or the value of the neutral point potential at the next sampling time is likely to exceed a certain allowable width, a large amplitude carrier and a small amplitude carrier are used together. A control device for an NPC converter according to any one of claims 8 to 10.

According to any one of the first to twelfth aspects, it is possible to suppress the fluctuation of the neutral point potential of the NPC inverter due to the single power supply configuration.

In order to achieve the above object, an invention according to claim 13 is an NPC converter in which a plurality of switching elements are bridge-connected to convert AC power to DC power, and a plurality of switching elements are bridge-connected. An NPC inverter that converts DC power into AC power, connects the NPC converter and the positive and negative DC terminals of the NPC inverter, and divides a DC voltage applied between the DC terminals. In the control device of the NPC converter for connecting two sets of smoothing capacitors for performing the pulse width modulation on each of the switching elements,
The positive DC power of the PC inverter is input to the positive input terminal of the first subtractor, and the negative DC power of the NPC inverter is input to the negative input terminal of the first subtractor. The power deviation is obtained, and the positive DC power of the NPC converter is supplied to a positive input terminal of a second subtractor, and the negative DC power of the NPC converter is supplied to a negative input terminal of the second subtractor. The converter-side DC power deviation is obtained and the inverter-side DC power deviation is supplied to a positive input terminal of a third subtractor, and the converter-side DC power deviation is supplied to a negative input terminal of the third subtractor. A power deviation is obtained by inputting the power deviation, and a compensation output obtained from the power deviation through a second gain is input to a neutral point compensation circuit of the NPC inverter. It is.

In order to achieve the above object, an invention according to claim 14 is an invention in which a plurality of switching elements are bridge-connected to convert an AC power into a DC power and a plurality of switching elements are bridge-connected. An NPC inverter that converts DC power into AC power, connects the NPC converter and the positive and negative DC terminals of the NPC inverter, and divides a DC voltage applied between the DC terminals. In the control device of the NPC converter for connecting two sets of smoothing capacitors for performing the pulse width modulation on each of the switching elements,
The positive DC power of the PC inverter is applied to the positive input terminal of a first subtractor, and the positive DC power of the NPC converter is applied to the negative input terminal of the first subtractor via a first gain. Input to calculate the positive side DC power deviation,
The negative DC power of the inverter is supplied to the positive input terminal of the second subtractor, and the negative DC power of the NPC converter is supplied to the first input terminal of the second subtractor.
Are input to the negative input terminal of the second subtractor via the gain of the second subtractor to determine the negative DC power deviation, and the positive DC power deviation is input to the positive input terminal of the third subtractor, Side DC power deviation is input to the negative input terminal of the third subtractor to obtain a power deviation, and the power deviation is obtained by using a compensation output obtained through a second gain as a neutral point compensation of the NPC inverter. NPC characterized in that it is inputted to a circuit
It is a control device of the converter.

In order to achieve the above object, an invention according to claim 15 is an invention in which a plurality of switching elements are bridge-connected to convert an AC power to a DC power and a plurality of switching elements are bridge-connected. An NPC inverter that converts DC power into AC power, connects the NPC converter and the positive and negative DC terminals of the NPC inverter, and divides a DC voltage applied between the DC terminals. In the control device of the NPC converter for connecting two sets of smoothing capacitors for performing the pulse width modulation on each of the switching elements,
The positive DC power of the PC inverter is input to the positive input terminal of the first subtractor, and the negative DC power of the NPC inverter is input to the negative input terminal of the first subtractor. The power deviation is obtained, and the positive DC power of the NPC converter is supplied to a positive input terminal of a second subtractor, and the negative DC power of the NPC converter is supplied to a negative input terminal of the second subtractor. The DC power deviation of the converter is obtained by inputting the signal to the function generator which generates an output when the power deviation, which is the output of the second subtractor, exceeds a predetermined range. To the positive side input terminal of the subtractor, and to the output side of the function generator via a first gain to the negative side input terminal of the third subtractor to obtain a power deviation. The power deviation, which is the output of the subtractor 3 A compensation output obtained through the gain control unit of the NPC converter, characterized in that entered in the neutral point compensation circuit of the NPC inverter.

In order to achieve the above object, the invention according to claim 16 is an invention in which a plurality of switching elements are bridge-connected to convert an AC power into a DC power, and a plurality of switching elements are bridge-connected. An NPC inverter that converts DC power into AC power, connects the NPC converter and the positive and negative DC terminals of the NPC inverter, and divides a DC voltage applied between the DC terminals. In the control device of the NPC converter for connecting two sets of smoothing capacitors for performing the pulse width modulation on each of the switching elements,
The positive DC power of the PC inverter is input to the positive input terminal of the first subtractor, and the negative DC power of the NPC inverter is input to the negative input terminal of the first subtractor. The power deviation is obtained, and the positive DC power of the NPC converter is supplied to a positive input terminal of a second subtractor, and the negative DC power of the NPC converter is supplied to a negative input terminal of the second subtractor. The converter-side DC power deviation is obtained to input the inverter-side DC power deviation to a positive input terminal of a third subtractor, and a multiplier for multiplying the converter-side DC power deviation by a constant and a first gain are obtained. Input to the negative side input terminals of the third subtractor via a second gain to obtain a power deviation, and input the compensation output obtained through the second gain to a neutral point compensation circuit of the NPC inverter. And the first An input terminal of a function generator that generates an output when the inverter-side DC power deviation exceeds a predetermined range is connected between an output terminal of the subtractor and a positive-side input terminal of the third subtractor; When an output is generated from the output terminal of the generator, a multiplication start command is given to the multiplier, and the first gain is connected to a negative input terminal of the third subtractor. It is a control device of the NPC converter which is a feature. In order to achieve the above object, an invention according to claim 17 is an NP for converting an AC power to a DC power by bridging a plurality of switching elements.
A C converter, and an NPC inverter that bridges a plurality of switching elements and converts DC power into AC power. The NPC converter connects the positive and negative DC terminals of the NPC inverter to each other, and In a control device of an NPC converter for connecting two sets of smoothing capacitors for dividing a DC voltage applied between both DC terminals and performing pulse width modulation on each of the switching elements, a DC voltage on the positive side of the NPC inverter is provided. The power is input to the positive input terminal of a first subtractor, and the negative DC power of the NPC inverter is input to the negative input terminal of the first subtractor to determine the inverter-side DC power deviation. The positive DC power of the converter is supplied to the positive input terminal of a second subtractor, and the negative DC power of the NPC converter is supplied to the second subtractor.
Are input to the negative input terminal of the subtractor to determine the converter-side DC power deviation, the inverter-side DC power deviation is input to the positive input terminal of the third subtractor, the converter-side DC power deviation, The third link is connected via a parallel circuit consisting of two normally-closed and normally-open interlocking contacts and first and third gains.
The DC power deviation is obtained by inputting to each of the negative input terminals of the subtractor, and the power deviation is input to a neutral point compensation circuit of the NPC inverter through a compensation output obtained through a second gain. 3 is connected between the output terminal of the subtractor and the input terminal of the second gain, the input terminal of a function generator that generates an output when the DC power deviation exceeds a predetermined range,
When an output is generated from an output terminal of the function generator, a switching command is given to the two interlocking contacts, and the first
A control device for an NPC converter, characterized in that the gain is switched from the third gain to the third gain.

To achieve the above object, an invention according to claim 18 is an invention in which a plurality of switching elements are bridge-connected to convert one AC power into a DC power, and one NPC converter is used to bridge the plurality of switching elements. NPC inverters connected to each other to convert DC power into AC power (n = an integer of 2 or more), and a positive DC terminal of the NPC converter is connected to each positive DC terminal of each NPC inverter. A DC voltage applied between the DC terminals of the NPC converter and the NPC inverters, wherein the DC terminals are connected in common, and the negative DC terminals of the NPC converter are connected to the respective negative DC terminals of the NPC inverters. In the control device of the NPC converter which connects two sets of smoothing capacitors for dividing One of the positive DC power of the NPC inverter is applied to a positive input terminal of a first subtractor, and the positive DC power of the NPC converter is applied to a product of a first gain and an n-th gain. To the negative input terminal of the first subtractor to determine the positive DC power deviation, and to apply any negative DC power of the NPC inverter to the positive terminal of the second subtractor. The NP input terminal
The negative DC power of the C converter is input to the negative input terminal of the second subtractor via the product of the first gain and the nth gain to obtain a negative DC power deviation. The positive DC power deviation is input to the positive input terminal of a third subtractor, and the negative DC power deviation is input to the negative input terminal of the third subtractor to determine a power deviation. A control device for an NPC converter, wherein a compensation output obtained via a second gain is input to a neutral point compensation circuit of the NPC inverter.

According to the inventions corresponding to the thirteenth to eighteenth aspects, the DC voltage is stabilized even when the converter operation or the commercial power supply fluctuates.

[0031]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the drawings. Before that, a conventional PWM system, a neutral point potential fluctuation suppressing technology, and an outline of the present invention will be described. FIG. 1 shows output phase voltages and element states of one phase of the NPC inverter. The NPC inverter can obtain three-valued phase voltages independently for three phases by the switching shown in FIG. Accordingly, 19 kinds of discrete voltage vectors can be output in 3 ³ = 27 kinds of ignition states. FIG. 2 shows the output voltage vector of the NPC inverter and the corresponding firing state (S
u, Sv, Sw).

In FIG. 2, for example, V12 (1, 0,-
1) U-phase firing state Su = 1, V-phase firing state Sv = 0,
This shows that the W-phase firing state Sw = −1.

If the DC voltage is equal, the NPC inverter can output a voltage vector having half the size of a normal two-level voltage type inverter, so that a significant reduction in the harmonic content can be expected.

However, the neutral point potential of the NPC inverter having one power supply varies depending on the current flowing into or out of the neutral point. Neutral current flows only when the phase voltage is clamped to neutral. When a current flows into the neutral point, the neutral point potential increases, and when the current flows out, the neutral point potential decreases.

Here, a conventional triangular wave comparison PWM will be described.

One of the methods for determining the ignition state of each switch of the inverter and outputting a target voltage command value is a triangular wave comparison PWM. Conventional triangular wave comparison PWM
Here, the firing state of the four switches is determined using the carrier (T1) reciprocating between 0 and +1 and the carrier (T2) reciprocating between -1 and 0. FIG. 3 shows an example.

In FIG. 3, when T1 <Vu ^* , Su1 = ON, Su2 = ON, Su3 = OFF, Su4
= OFF When T2 <Vu ^* <T1, Su1 = OFF, Su2 = ON, Su3 = ON, Su4
= OFF Vu ^* <T2, Su1 = OFF, Su2 = OFF, Su3 = ON, Su
4 = ON, which corresponds to the firing state in FIG.

By switching in this manner,
As a result, a pulse train in which the average voltage is equal to the voltage command value can be output.

Next, a conventional method for suppressing the neutral point potential fluctuation in the space vector PWM will be described.

In the space vector PWM using the space vector diagram shown in FIG. 2, when a certain command V is given, the sampling period is divided into three or three or more predetermined time sections, and the vector surrounding the command V is Pulses are output such that the average output within the sampling period is equal to the command V by allocating to each time section and outputting in order. A set of voltage vectors that can be generated by the same vector set is a small equilateral triangle on the space vector diagram, and the space vector diagram of FIG. 2 can be divided into regions by these equilateral triangles.
FIG. 4 shows the area division of the vector diagram of the section from 0 ° to 60 ° when the entire section of 360 ° is divided into sections of 60 °. FIG. 5 shows combinations of vectors used in each area.

For example, when the command V is at the position shown in FIG.
The region in which V exists is the following. If the firing time distribution ratios of the three vectors surrounding V are αi, αj, and αk, respectively, the respective firing time distribution ratios are determined so as to satisfy the following equation.

Αi · V1 + αj · V11 + αk · V12 = V (1) αi + αj + αk = 1 (2) The command V is output by switching the firing state for each of these firing time distribution ratios.

When the PWM for the command V in this example is executed, the neutral point current flows when the ignition state is (1, 1).
0,0), (0, -1, -1) and (1,0, -1)
It is time. First, consider the neutral point potential fluctuation in the case of (1, 0, 0) and (0, -1, -1), which are the firing state of the vector V1. FIG. 6 shows these two firing states.

When the direction of the phase current of each phase is as shown in the figure, the neutral point currents of FIGS. 6A and 6B have the same magnitude but different values for positive and negative. Therefore, when outputting the vector V1, if the neutral point potential is too high, the ignition state in which the neutral point current flows in accordance with the direction of the AC side current is selected, and if the neutral point potential is too low. , The ignition state in the direction in which the neutral point current flows is selected. In this manner, the neutral point current can be positively controlled.

On the other hand, the vector V12 shows the ignition state as (1,
In this ignition state, the neutral point current continues to flow in one direction, and the neutral point current cannot be controlled. That is, the command V is the vector V12
, The time distribution for outputting V12 increases, and the neutral point potential fluctuation becomes uncontrollable.

To avoid this, vectors (V12, V23, V34, V45, V56, V56) on the outer hexagonal side shown in FIG.
A method for realizing a voltage command given in combination with another vector without using V61) is known.

For example, in order to realize the command V of FIG. 4 without using V12, there may be a set of (V1, V11, V2).
There can be a set of (V1, V11, V22). As described above, there are several types of region division for PWM that does not use the vector on the side. As one of them, there is one divided into regions as shown in FIG. 7, and vectors used in each region in FIG. 7 are shown in FIG.

The switching of the vector described above cannot be dealt with by the triangular wave comparison method of FIG. 3 because the transition of the ignition state includes +1 to -1 or -1 to +1. For this reason, as described below, the embodiment of the present invention uses the triangular wave comparison PWM using both the large amplitude carrier and the small amplitude carrier.

In the conventional triangular wave PWM of the NPC inverter, as shown in FIG.
A carrier T1 reciprocating between 1 and a carrier T2 reciprocating between -1 and 0 have been used.

In the present invention, these carriers are called "small amplitude carriers". Further, the carrier reciprocating between -1 and +1 as in T0 of FIG.
In the embodiment of the present invention, the small-amplitude carrier and the large-amplitude carrier are appropriately used in combination.

In FIG. 9, when Vu ^* > T0, Su1 = ON, Su2 = ON, Su3 = 0FF, Su4
= OFF Vu ^* <T0, Su1 = OFF, Su2 = OFF, Su3 = ON, Su
4 = ON. Where the carrier and the modulation signal intersect, the output voltage switches from + E to -E or -E to + E. In other words, 0 is skipped, and the operation becomes the same as that of the two-level inverter.

FIG. 9 shows an example of the waveform when the U-phase modulated signal crosses the large amplitude carrier and the V phase crosses the small amplitude carrier.

In general, in an induction motor drive using a PWM inverter, a phase voltage may be arbitrarily set to output a certain voltage command vector as long as the line voltage is the same (harmonics change, but Let's focus only on the fundamental wave). Therefore, the three-phase modulation signals Vu ^* , Vv ^* , Vw ^*
Shifting the whole up or down within the ± 1 limit,
The fundamental wave of the output voltage does not change. That is, the modulation signal can be processed so as to maintain the line voltage.

FIG. 9 shows a simple waveform in which the processing of the modulation signal is not particularly performed and no consideration is given to the suppression of harmonics and neutral point potential fluctuation.

In this embodiment, the modulation signal of a certain phase intersects with the large amplitude carrier according to the angle of the command vector,
The switching of the vector is realized by intersecting the modulated signal of another phase with the small amplitude carrier. Further, each phase modulation signal is processed so as to distribute the ignition time to each vector so that the same average voltage as the command vector is output in the sampling period.

Hereinafter, a method of processing a modulation signal in the case where a command vector exists in each of the regions,... In FIG. 7 will be described.

(1) Region in FIG. 10A Three firing states are output during one control cycle. 3
There are three ways to switch between the three firing states depending on which vector to start from, two ways clockwise or counterclockwise, and two firing options for the vector V1.
× 2 × 2 = 12 patterns.

During one sampling period, each modulated signal is held at a constant value. When crossing the held modulation signal with the carrier signal rising from the valley to the peak, the ignition state (Su, Sv, Sw) of each phase is determined by the modulation signal sticking to the maximum value +1 or the minimum value -1. It changes only once, except when it is In addition, the direction of the change is always a downward direction, 1 → 0 or 0 → −1 when crossing a small amplitude carrier, and 1 → −1 when crossing a large amplitude carrier. Of the above 12 patterns, the pattern on the upper side of FIG. 10B is adopted as a pattern that meets these conditions.

Hereinafter, the relationship between the ignition switching pattern and the modulation signal for realizing the pattern will be considered with reference to FIG. From FIG. 10B, the U-phase switches between 1.0 and 1.0 using the small-amplitude carrier T1, and the V-phase switches between 1 and -1 using the large-amplitude carrier T0. Since the W phase is solidly attached to -1, the carrier may be either large or small, but is considered as a small amplitude carrier in consideration of unification with other regions. Further, the modulation signal is processed as described below under the condition that the output of the W phase is always −1 and the line voltage is constant. Where Vuo, Vvo, Vwo
Denotes a modulation signal of each phase before processing, and Vu ^* , Vv ^* , V
w ^* indicates a modulated signal of each phase after processing.

Vu ^* = Vuo−Vwo−1 Vv ^* = Vvo−Vwo−1 Vw ^* = − 1 (3) After all, this modulated signal is converted to each large amplitude / small as shown in the lower part of FIG. By intersecting with the amplitude carrier, ignition switching as shown in the upper part of FIG. 10B can be realized, and by this switching, a pulse train of the same average voltage as the command vector V in the area can be output.

(2) Area of FIG. 11 (a) Twelve orders can be considered for switching the three ignition states as in the case of the area. Among them, the ignition switching pattern shown in FIG. 11 (b) is used. adopt.

As shown in FIG. 11B, the ignition state is switched by using the small-amplitude carrier for the W phase and using the large-amplitude carrier for the V phase. Since the U-phase is +1 solid, the carrier may be large or small, but is considered as a small-amplitude carrier in consideration of unification with other regions. Further, the modulation signal is processed as follows from the condition that the output of the U phase is always 1 and the line voltage is constant.

Vu ^* = 1 Vv ^* = 1−Vuo + Vvo Vw ^* = 1−Vuo + Vwo (4) After all, this modulated signal is crossed with each large / small amplitude carrier as shown in the lower side of FIG. 11B. If this is the case, the ignition switching as shown in the upper part of FIG. 11B can be realized, and a pulse train having the same average voltage as the command vector V in the region of FIG.

(3) Area of FIG. 12 (a) To switch between the three firing states, there are three ways depending on which vector to start from, two ways whether clockwise or counterclockwise, and the points of vectors V1 and V2. Since there are two types of arc options, there are a total of 3 × 2 × 2 × 2 = 24 patterns. Among them, a firing switching pattern as shown in FIG.

As shown in FIG. 12B, the U-phase and the W-phase use small-amplitude carriers, and the V-phase uses large-amplitude carriers to switch the ignition state. Further, the modulation signal is processed as follows from the condition for simultaneously switching the U-phase and the V-phase and the condition for keeping the line voltage constant.

Vu ^* = 1−Vuo + Vvo Vv ^* = 2 · (−Vuo + Vvo) +1 Vw ^* = − 3 · Vu + 1 + 1 (5) Eventually, this modulated signal is converted to each large amplitude as shown in the lower part of FIG. By intersecting with the / small-amplitude carrier, it is possible to realize the ignition switching as shown in the upper part of FIG.

(4) Region of FIG. 13 (a) In order to switch the three ignition states, 24 orders can be considered as in the case of the region. Among them, the ignition switching pattern as shown in FIG. adopt.

From FIG. 13B, the U-phase and W-phase use small-amplitude carriers, and the V-phase uses large-amplitude carriers to switch the firing state. Further, the modulation signal is processed as follows from the condition of simultaneously switching the V phase and the W phase and the condition of the constant line voltage.

Vu ^* = − 3 · Vw0-1 Vv ^* = 2 · (Vv0−Vw0) −1 Vw ^* = Vv0−Vw0−1 (6) After all, this modulated signal is converted to the lower side of FIG. As shown in Fig. 13 (b), by intersecting each large-amplitude / small-amplitude carrier, the ignition switching as shown in the upper part of Fig. 13 (b) can be realized. it can.

(5) Area of FIG. 14 (a) Switching between the three ignition states is performed in three ways depending on which vector to start from, two ways in the clockwise or counterclockwise direction, and in the points of the vectors V1 and V2. Two arc options each,
Since there are three firing options for V0, a total of 3 × 2
× 2 × 2 × 3 = 72 patterns. Among them, a firing switching pattern as shown in FIG.

As shown in FIG. 14B, the ignition state is switched using the small-amplitude carrier in all phases. Further, the modulation signal is processed as follows from the condition that the W phase is always 0 and the line voltage is constant.

Vu ^* = Vu0−Vw0 Vv ^* = Vv0−Vw0 Vw ^* = 0 (7) After all, if this modulated signal intersects with the small amplitude carrier as shown in the lower part of FIG. 14 (b), it is possible to realize the ignition switching as described above, and by this switching, it is possible to output a pulse train having the same average voltage as the command vector V in the area.

Although the range of 0 ° to 60 ° has been described above, other ranges can be similarly derived.

<First Embodiment> FIG.
It is a block diagram showing a 1st embodiment of a control device of a PC converter. The phase current on the AC side of the inverter is detected by a current detector (not shown), and this phase current is detected by the A / D converter 2.
, Where it is converted to digital currents iu, iv, iw. The digital currents iu, iv, iw are input to the three-phase / two-phase conversion means 6 and are changed into two-phase currents ix, iy.

The two-phase currents ix and iy converted by the three-phase / two-phase conversion means 6 are converted into dq along with the phase angle (electrical angle) θ.
The d-axis current id and the q-axis current i
is converted to q. The d-axis current id and the q-axis current iq converted by the dq conversion means 7 are subtracted by subtracters 8 and 9, respectively.
, And the d-axis current command value idref and the q-axis current command value iqref are input to the other input terminals of the subtracters 8 and 9, respectively, where the current deviation is obtained. The respective current deviations are respectively proportionally integrated (PI)
10 and 11 to output a d-axis voltage command Vd and a q-axis voltage command Vq.

The voltage command Vd and the voltage command Vq are input to the inverse dq converter 12 together with the phase angle (electrical angle) θ.
Here, voltage commands Vx and Vy of the stationary coordinate system are output.

The voltage commands Vx and Vy of the stationary coordinate system are input to the space vector area determining means 13, where the large or small carrier selection (1/0) signals Cu, Cv and Cw and the area numbers are output. Large or small carrier selection (1 /
0) The signals Cu, Cv, and Cw are input to the carrier generator 3, where the carriers CAu, CAv, and CAw are generated.

The voltage commands Vx and Vy of the stationary coordinate system obtained by the inverse dq conversion means 12 are converted by the two-phase / three-phase conversion means 14 into voltage commands Vu0, Vv0 and Vw of the three-phase coordinate system, respectively.
Outputs 0. Voltage commands Vu0, Vv of the three-phase coordinate system
The modulation signals Vu ^* , Vv ^* , and Vw ^* are output by inputting 0, Vw0 and the region number output from the space vector region determination unit 13 to the modulation signal processing unit 15.

The modulation signals Vu ^* ,
Vv ^* , Vw ^* and carrier CAu of carrier generator 3,
CAv and CAw are input to the comparator 4, and temporary firing signals Su0, Sv0 and Sw0 are output, respectively.

The neutral point potential control means 16 calculates the voltage commands Vx and Vy of the stationary coordinate system obtained by the inverse dq conversion means 12 and
The digital currents iu, iv, iw of the A / D converter 2 and the detection voltages Vc1, Vc2 of the capacitors are input,
Here, the firing option selection signal for the inner hexagonal vertex (1
/ 0) is output. The ignition state determination means 5 includes an ignition option selection signal (1/0) of the neutral point potential control means 16 and provisional ignition signals Su0, Sv0, Sw0 of the comparator 4.
And outputs the firing signals Su, Sv, Sw to the drive circuit of the switching device, respectively.

In the configuration described above, each phase detection current detected from the hardware side in FIG.
The d-axis current id and q-axis current iq are obtained by sequentially passing through the conversion means 2, three-phase to two-phase conversion means 6, and dq conversion means 7, and these currents id and iq and speed control and vector control (not shown) are obtained. The deviation between the d-axis current command value idref and the q-axis current command value iqref calculated by the means is input to the proportional integration means 10 and 11, and the voltage command V for the stationary coordinate system is input.
x and Vy are obtained, and the voltage commands Vx and Vy are input to the space vector area determination means 13, so that the command vector is obtained and any of the areas in FIG. , A vector to be further used and its output pattern are determined, and a carrier selection signal for selecting a large-amplitude carrier / small-amplitude carrier is output, and at the same time, for all the firing options included in the area, Neutral current is predicted,
A selection signal of an ignition option such that the neutral point potential approaches 0 V is output.

The voltage commands Vx, Vy are input to the modulation signal processing means 15 via the two-phase to three-phase conversion means 14, where the modulation signals are processed, and the modulation signals Vu ^* , Vv ^* , V
w ^* is output.

Then, the modulated signals Vu ^* , Vv ^* , Vw
^* And a carrier generated from the carrier generator 3 corresponding to the carrier selection signal are compared by a comparator 4, where a provisional firing signal is output, and a firing state determination circuit is provided for the provisional firing signal. 5, a firing option is selected such that the neutral point potential approaches 0 V, and a firing signal is output to the drive circuit of each switching device.

As a result, according to the first embodiment described above, it is possible to suppress the fluctuation of the neutral point potential of the NPC inverter due to the single power supply configuration.

<Second Embodiment> Next, a large-amplitude carrier / small-amplitude carrier combination P according to a second embodiment of the present invention will be described.
The neutral point potential fluctuation suppression method (neutral point potential prediction control) in the WM will be described with reference to FIG. 27 for describing the above-described conventional problem.

In FIG. 27, it has already been described that the neutral point potential rises when a current flows into the neutral point and falls when the current flows out, but this relationship will be quantitatively described below.

From FIG. 27, the equation for entering and exiting the current at the neutral point is as follows.

Io = iCP−iCN (8) The element characteristics of the P-side and N-side capacitors are as follows: C · (d / dt) · VCP = iCP (9) C · (d / dt) · VCN = iCN From the formulas (8), (9) and (10), C · (d / dt) · (VCP−VCN) = iO (11) where ΔVC = VCP−VCN (12) In other words, equation (11) is as follows: C · (d / dt) ΔVC = iO (13)

The temporal change of ΔVC in equation (12) is a so-called “neutral potential fluctuation”, and equation (13) expresses the relationship between the neutral point current and the neutral point potential.

Now, the control section [kTs, (k + 1) T
s] is considered. (Ts is a sampling cycle) The value of ΔVC at time t = kTs, that is, ΔVC (kTs) can be measured, and the interval [kTs, (k
+1) Ts], the value of ΔVC at time t = (k + 1) Ts can be obtained by the equation (1)
From 3), prediction can be made by the following equations (14) and (15).

[0091]

(Equation 1)

For each control cycle, the control device sets the interval [kT
s, (k + 1) Ts], because the vector order and the time ratio given to each vector are determined by oneself, so if there are vectors with multiple firing options in the vector set,
It can be predicted with considerable accuracy how much neutral current will flow when any ignition option is chosen.

In the neutral point potential prediction control method, for all the ignition options, the neutral point current when each ignition option is adopted is predicted, and the ignition option at which the neutral point potential approaches 0 V is selected. By doing so, the neutral point potential fluctuation is suppressed. As an example, a case of a region and a case of a region are shown.

(1) Neutral point potential prediction in the region of FIG. 18A (the same applies to the region) As shown in FIG. 18A, V1 having two firing states in the region is included as a vector to be used. In. V1
FIGS. 19A and 19B show the two ignition states.

Here, when the phase current flows in the direction of the arrow in the figure, that is, the neutral point current iO (1, 0, 0) flowing as shown in FIG. 27, and as shown in FIG. The flowing neutral point current i0 (0, -1, -1) is a U-phase current iu.
Can be expressed as follows using

I 0 (1,0,0) = − iu (16) i 0 (0, −1, −1) = iu (17) In other vectors V22 and V11 used in the region, the neutral point current is Since the current does not flow, the neutral point potential can be predicted to fluctuate during the control cycle Ts as shown in equations (18) and (19).

Here, the distribution time ratio for outputting the vector V1 is α1.

ΔVC [(k + 1) Ts] (1, 0, 0) = ΔVC (kTs) + i0 (1, 0, 0) · α1 · Ts / C (18) ΔVC [(k + 1) TS) (0, −1, −1) = ΔVC (kTs) + i0 (0, −1, −1) · α1 · Ts / C (19) At time t = (k + 1) Ts, the predicted value of ΔVC is closer to zero. Use the firing option. Thus, from the prediction of the fluctuation of the neutral point potential, the one that brings the neutral point potential closer to 0 V is selected from the two firing options to suppress the neutral point potential.

(2) Neutral point potential prediction in the region shown in FIG. 18B (the same applies to the region) In the region, V1 and V2 having two ignition states are included as vectors to be used. Can be considered.

The neutral point currents flowing in the respective ignition states are represented by i0 (1,1,0), i0 (0,0, -1), i0
Given (1, 0, 0) and i0 (0, -1, -1), the neutral point potential fluctuation can be predicted as follows. However, V1
Is the distribution time ratio for outputting V2, and α2 is the distribution time ratio for outputting V2.

ΔVC [(k + 1) TS) (1,0,0) (1,1,0) = ΔVC (kTs) + i0 (1,0,0) · α1 · TS / C + i0 (1,1,0) ) .Alpha.2 .Ts / C (20) .DELTA.VC [(k + 1) TS] (0, -1,1) (1,1,0) =. DELTA.VC (kTs) + i0 (0, -1,1, -1). α1 · TS / C + i0 (1,1,0) · α2 · Ts / C (21) ΔVC [(k + 1) TS] (1,0,0) (0,0, −1) = ΔVC (kTs) + I0 (1,0,0) .alpha.1.Ts / C + i0 (0,0, -1) .alpha.2.Ts / C (22) .DELTA.VC [(k + 1) TS] (0, -1, -1) ( 0,0, -1) =. DELTA.VC (kTs) + i0 (0, -1, -1) .alpha.1 Ts / C + i0 (0,0, -1) .alpha2.Ts / C (23) Neutral point potential closer to 0V Selects a combination of arc option suppresses neutral point potential fluctuation.

Next, an example of suppressing the neutral point potential fluctuation will be described.
This will be described with reference to FIGS. FIG. 16 (a)
Is the conventional triangular wave comparison PWM output (UV line voltage PW
M output waveform), and FIG. 16B shows a PWM of the large-amplitude carrier / small-amplitude carrier combined type according to the present embodiment.
5 shows an example of an output (UV line voltage PWM output waveform).

FIGS. 17A, 17B, and 17C show simulation results of the neutral point potential when the induction motor, which is a load, is subjected to vector control and speed control. FIG. 17 (a) shows the case where operation was performed at 40 Hz, and FIG.
FIG. 17B is a neutral point potential waveform diagram when the suppression control is started during acceleration (about 40 Hz), and FIG. 17C is a case where the operation is performed at 60 Hz. (When this embodiment is not used), and the latter half of the time axis shows the case with suppression control (when this embodiment is used). As is clear from this figure, it can be seen that the neutral point potential fluctuation is suppressed in any case.

The simulation conditions are PW
In this case, the M frequency is 2 kHz, the control cycle is 250 μs, the DC voltage is 400 V, the capacitor capacity is 2 mF, and the rating of the induction motor is 11 kW, 200 V, 4 P, 1500 RPM.

<Third Embodiment> FIG. 20 is a control block diagram for explaining the third embodiment. The difference from FIG. 15 is that the neutral point potential control means 16 in FIG. In this example, the actual currents iu, iv, and iw are used as the input currents, but the current reference values iuref, ivref, and iwref are used. This current reference value iuref, ivr
ef and iwref are d-axis current command values idref and q calculated by speed control / vector control means (not shown).
The axis current command value iqref is converted into static coordinate system current commands ixref and iyref by the inverse dq conversion means 27, and the converted current commands ixref and iyref are input to the two-phase to three-phase conversion means 26. The obtained current reference values iuref, ivref, iwref. The other configuration is the same as that of FIG.

Therefore, it is needless to say that the third embodiment can provide the same operation and effect as the second embodiment.

<Fourth Embodiment> FIG. 21 is a schematic diagram for explaining a fourth embodiment. This is achieved by bridging a plurality of switching elements, such as GTOs, into an NPC converter (hereinafter referred to as CNV) 30 for converting AC power from an AC power supply 60 into DC power, and a plurality of switching elements, such as GTOs. An NPC inverter (hereinafter referred to as INV) 40 that converts DC power into AC power and supplies the AC power to, for example, an electric motor 70.
And two sets of smoothing capacitors 5 for connecting the positive and negative DC terminals of the CNV 30 and INV 40 and dividing the DC voltage applied between the two DC terminals.
The control device of the NPC converter which connects 0, 51 and performs pulse width modulation of each switching element is configured as follows.

That is, as described later, IN
The DC power on the positive side of V40 is applied to the positive input terminal of the first subtractor 81, and the DC power on the negative side of the INV 40 is applied to the negative input terminal of the first subtractor 81 as described later. The DC power deviation of the inverter side is obtained by inputting each of them, and the DC power on the positive side of the CNV 30 detected as described later is input to the positive input terminal of the second subtractor 82 and the CNV 30 detected as described later. The DC power on the negative side of
To the negative input terminal of the subtractor 82 to determine the converter-side DC power deviation, the inverter-side DC power deviation to the positive input terminal of the third subtractor 83, and the converter-side DC power deviation to the third To the negative input terminal of the subtracter 83 to obtain a power deviation, and to input a compensation output obtained through the second gain K2 to a neutral point compensation circuit (not shown) of an NPC inverter. It was made.

Here, referring to FIG.
A method of detecting DC power on the positive side and the negative side of V30 and the inverter INV40 will be described. Each phase U, V, W
Each time, a multiplier 100, a coefficient unit 101, a switching command unit 10
2. A positive and negative DC power is detected by a circuit including a changeover switch 103 and an adder 104. That is, the voltage command (U-phase) and the current command (U-phase) are
Are calculated, and the product of the two is obtained. The obtained multiplied value is input to the coefficient unit 101U and is converted into a predetermined multiple. The adder 10 is connected via a changeover switch 103U consisting of two interlock switches having a pair of normally open terminals and normally closed terminals.
4a.

In this case, for example, one of the normally closed terminals of the changeover switch 103U is connected between the output side of the coefficient unit 101U and the positive power output terminal, and between the terminal “0” and the negative power output terminal. Is connected to the other of the normally-closed terminals of the changeover switch 103U.
One of the normally open terminals of the changeover switch 103U is connected between the negative power output terminal, and the other of the normally open terminals of the changeover switch 103U is connected between the “0” terminal and the positive power output terminal. .

The changeover of the changeover switch 103U is performed according to a command from the changeover command unit 102U. The changeover is performed according to the polarity of the voltage command (U phase). The output of the switch 101U is supplied to only one of the normally closed terminals of the changeover switch 103U. Such a configuration is the same for the V-phase and the W-phase, and a description thereof will be omitted.

As described above, the deviation between the positive DC power and the negative DC power of the INV 40 is obtained, the deviation between the positive DC power and the negative DC power of the CNV 30 is obtained, and a gain is added to the power deviation between the two. 84 and the second
And the negative DC power of the CNV 30 is input to the positive input terminal of the subtractor 82 and the negative input terminal of the second subtracter 82 to determine the converter DC power deviation. To the positive input terminal of the third subtractor 83,
The converter-side DC power deviation is input to each of the negative input terminals of the third subtracter 83 to obtain a power deviation, and the power deviation is converted to a compensation output obtained via a second gain K2 by an NPC inverter (not shown). This is input to a neutral point compensation circuit (correction amount calculation means).

As described above, according to the fourth embodiment, IN
The difference between the positive DC power and the negative DC power of V40 is obtained, the difference between the positive DC power and the negative DC power of the CNV 30 is obtained, and the gain 84 is made proportional to the power difference between the two. Is input to the positive and negative DC power of the INV 40 to the neutral potential compensating circuit.
When there is a fluctuation in the power supply 30 and the commercial power supply, there is no delay in compensation or an increase in the harmonic content of the output waveform. Further, according to the fourth embodiment, frequency dispersion and
Beat suppression can be achieved.

<Fifth Embodiment> As shown in FIG.
The difference from the fourth embodiment shown in FIG.
5, 86 are newly added and configured as follows. That is, the positive DC power of the INV 40 is supplied to the positive input terminal of the first subtractor 81, and the positive DC power of the CNV 30 is supplied to the negative input of the first subtractor 81 via the first gain 85. Input to each of the terminals to determine the positive DC power deviation, and the DC power on the negative side of INV 40 to a second subtractor 8
2, the negative DC power of the CNV 30 is input to the negative input terminal of the second subtracter 82 via the first gain 86 to obtain the negative DC power deviation. The positive DC power deviation is input to the positive input terminal of a third subtractor 83, and the negative DC power deviation is input to the negative input terminal of the third subtracter 83 to obtain a power deviation. The compensation output obtained from the power deviation via the second gain 84 is input to the neutral point compensation circuit of the INV 40.

The other structure is the same as that of FIG.
The same effects as those of FIG. 21 can be obtained even when there is a difference between the device capacities of the CNV 30 and the INV 40.

<Sixth Embodiment> As shown in FIG.
The difference from the fourth embodiment shown in FIG.
6. A function generator 88 is added, and is configured as follows. That is, the DC power on the positive side of the INV 40 is
And the negative DC power of the INV 40 is input to the negative input terminal of the first subtracter 81 to obtain the inverter DC power deviation.
30 is supplied to the positive input terminal of the second subtractor 82, and the negative DC power of the CNV 30 is supplied to the second subtractor 8.
And a converter-side DC power deviation is obtained by inputting to each negative input terminal of the second subtractor 82, and via a function generator 88 which generates an output when the power deviation which is the output of the second subtractor 82 exceeds a predetermined range. The first gain 86 is connected, the inverter-side DC power deviation is input to the positive input terminal of the third subtractor 83, and the output side of the first gain 86 is connected to the negative input of the third subtractor 83. The input power is input to the respective input terminals to determine a power deviation, and the power deviation, which is the output of the subtracter 83, is input to the neutral point compensation circuit of the NPC inverter via the compensation output obtained via the second gain 84. It was done.

The other structure is the same as that of FIG.
In particular, when a neutral point compensation circuit also exists on the converter side, it has a function of preventing excessive compensation.

<Seventh Embodiment> As shown in FIG.
The difference from the fourth embodiment shown in FIG. 21 is that a multiplier 89, a first gain 86, and a function generator 88 are newly added and configured as follows. That is, the DC power on the positive side of the INV 40 is supplied to the positive side input terminal of the first
The DC power of the NV 40 is input to the negative input terminal of the first subtracter 81 to determine the inverter-side DC power deviation, and the positive DC power of the CNV 30 is converted to the second subtractor 8.
2, the DC power on the negative side of the CNV 30 is input to the negative input terminal of the second subtracter 82 to determine the DC power deviation on the converter side, and further, the DC power deviation on the inverter side is calculated. 3 is input to the positive-side input terminal of the subtracter 83, and a multiplier 89 that multiplies the converter-side DC power deviation by a constant.
And a negative input terminal of the third subtractor 83 via a first gain 86 to obtain a power deviation. The power deviation is obtained by calculating a compensation output obtained through a second gain 84 of the INV 40. The signal is input to the neutral point compensation circuit and the first subtractor 8
2 and an input terminal of a function generator 88 that generates an output when the inverter-side DC power deviation exceeds a predetermined range, between the output terminal of the second subtractor 83 and the positive input terminal of the third subtractor 83. A multiplication start command is given to the multiplier 89 when an output is generated from the output terminal of the subtractor 88, and the first gain 86 is connected to the negative input terminal of the third subtractor 83. It is.

The other structure is the same as that of FIG.
When the DC deviation of INV40 becomes excessive, CNV
It is possible to prevent the compensation circuit from reaching saturation due to the 30 DC deviation.

<Eighth Embodiment> As shown in FIG.
The difference from the fourth embodiment shown in FIG.
5. Function generator 88, third gain 90, normally open contact 91
a and two interlocking contacts 91 having a normally closed contact 91b are newly added and configured as follows. That is,
The DC power on the positive side of the INV 40 is input to the positive input terminal of the first subtractor 81, and the DC power on the negative side of the INV 40 is input to the negative input terminal of the first subtractor 81, and the DC power on the inverter side is input. The deviation is obtained, and the positive DC power of the CNV 30 is input to the positive input terminal of the second subtractor 82, and the negative DC power of the CNV 30 is input to the negative input terminal of the second subtracter 82. The inverter-side DC power deviation is input to the positive input terminal of the third subtractor 83, and the converter-side DC power deviation is calculated by the two interlocking contacts 91 and the first and third gains 85. , 90 to the negative input terminal of the third subtractor 83 to obtain a DC power deviation, and to calculate the DC power deviation by a compensation output obtained via a second gain 84. NP
The input to the neutral point compensation circuit of the C inverter, between the output terminal of the third subtractor 83 and the input terminal of the second gain 85,
An input terminal of a function generator 88 that generates an output when the DC power deviation exceeds a predetermined range is connected, and when an output is generated from the output terminal of the function generator 88, a switching command is given to the two interlocking contacts 91. At the same time, the first gain 85 is switched to the third gain 90.

The other structure is the same as that of FIG.
The operation and effect are the same as those in FIG.

<Ninth Embodiment> As shown in FIG.
One NPC converter (CN) that converts a plurality of switching elements into a bridge and converts AC power to DC power.
V) 30 N-type inverters (INV) (1) to (n), each of which is a bridge connection of a plurality of switching elements and converts DC power into AC power (n = an integer of 2 or more).
, 4n, and each INV 41, 42,.
4n, the positive DC terminal of CNV30 is commonly connected to each positive DC terminal, and the negative DC terminals of CNV30 are commonly connected to each negative DC terminal of each INV41, 42,. 4n are connected to two sets of smoothing capacitors 50 and 51 for dividing a DC voltage applied between DC terminals of INVs 41, 42,... 4n, and a control device of an NPC converter for pulse width modulation of each switching element. It is configured as follows.

.. 4n, the nth positive DC power is applied to the positive input terminal of the first subtractor 81n, and the positive DC power of the CNV 30 is applied to the first gain K1 and the first gain K1. Gain 92n obtained by multiplying the n-th gain Kn
, To input to the negative input terminal of the first subtractor 81n to obtain the positive DC power deviation.
4n, the nth negative DC power is applied to the positive input terminal of the second subtractor 82n, and the negative DC power of the CNV 30 is applied to the first gain K1 and the nth gain Kn.
Is subtracted from the second subtractor 82n via a gain 93n
, The negative DC power deviation is obtained, the positive DC power deviation is input to the positive input terminal of the third subtractor 83n, and the negative DC power deviation is input to the third subtractor 83n. , A power deviation is obtained by inputting the power deviation to a negative input terminal, and a compensation output obtained through the second gain 84n is input to a neutral point compensation circuit of the INV4n.

In the above-described compensation circuit for the n-th inverter INVn, a compensation circuit having a similar configuration is provided for each of the inverters other than the n-th inverter.

According to the above-described embodiment, by individually setting the gains Kn and K2n according to the capacity and the margin of each inverter, efficient compensation can be achieved as a whole system.

[0126]

According to the present invention described above, it is possible to suppress the fluctuation of the neutral point potential of the NPC inverter with one power supply configuration.
Further, it is possible to provide a control device for an NPC converter in which the DC voltage is stabilized even when the converter operation or the commercial power supply fluctuates.

[Brief description of the drawings]

FIG. 1 is a diagram showing a firing state of an NPC inverter for explaining a conventional PWM method and a neutral point potential fluctuation suppression technique.

FIG. 2 is a space voltage vector diagram of an NPC inverter.

FIG. 3 is a diagram showing an example of a conventional triangular wave comparison PWM method.

FIG. 4 is a diagram showing area division of the space voltage vector diagram of FIG. 2;

FIG. 5 is a diagram showing combinations of vectors used in each area in FIG. 4;

FIG. 6 is a diagram for explaining a conduction state of a vector V1 in FIG. 4;

FIG. 7 is a diagram showing an example of region division for PWM without using the on-side vector in FIG. 2;

FIG. 8 is a view for explaining vectors used in each area in FIG. 7;

FIG. 9 is a waveform diagram showing an example of a triangular wave comparison PWM system using both a large amplitude carrier and a small amplitude carrier according to the present invention.

FIG. 10 is a diagram showing a firing switching pattern in the region of FIG. 7;

FIG. 11 is a diagram showing a firing switching pattern in the region of FIG. 7;

FIG. 12 is a diagram showing a firing switching pattern in the region of FIG. 7;

FIG. 13 is a diagram showing a firing switching pattern in the region of FIG. 7;

FIG. 14 is a diagram showing a firing switching pattern in the region of FIG. 7;

FIG. 15 is a control block diagram showing the first embodiment of the present invention.

FIG. 16 shows a comparison PWM between the embodiment shown in FIG. 15 and a conventional triangular wave.
FIG. 4 is a waveform chart showing an example of a PWM output of a system.

FIG. 17 is a waveform chart showing a simulation result when vector control and speed control are performed on the induction motor for explaining the operation and effect of the embodiment of FIG. 15;

FIG. 18 is a vector diagram of a region of a space voltage vector diagram for explaining the second embodiment of the present invention.

FIG. 19 is a circuit diagram illustrating a firing state of a vector V1 and a direction of a flowing current for explaining a second embodiment of the present invention.

FIG. 20 is a control vector diagram for explaining a third embodiment of the present invention.

FIG. 21 is a circuit diagram for explaining a fourth embodiment of the present invention.

FIG. 22 is a circuit diagram illustrating a fifth embodiment of the present invention.

FIG. 23 is a circuit diagram for explaining a sixth embodiment of the present invention.

FIG. 24 is a circuit diagram for explaining a seventh embodiment of the present invention.

FIG. 25 is a circuit diagram for explaining an eighth embodiment of the present invention.

FIG. 26 is a circuit diagram illustrating a ninth embodiment of the present invention.

FIG. 27 is a view for explaining a method of detecting DC power on the positive side and negative side of the converter and the inverter in FIGS. 21 to 26;

FIG. 28 is a diagram showing a main circuit of an NPC converter according to the related art and the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... NPC converter, 2 ... A / D converter, 3 ... Carrier generator, 4 ... Comparator, 5 ... Ignition state determination means, 6 ...
Three-phase / two-phase conversion means, 7: dq conversion means, 8, 9: subtractor, 10, 11: proportional integration means, 12: inverse dq conversion means, 13: space vector area determination means, 14: two-phase / 3 Phase conversion means, 15: modulation signal processing means, 16: neutral point potential control means, 26: two-phase / three-phase conversion means, 27: inverse dq
Conversion means, 28 neutral point potential control means, 30 converter, 40 inverter, 50, 51 smoothing capacitors,
60 ... AC power supply, 70, 71, 72, ... 7n ... electric motor,
81, 82, 83 ... subtractor, 84, 85, 86, 90 ...
Gain, 87: switch, 88: function generator, 89: multiplier, 91a, 91b: two interlocking contacts.

Claims (18)

[Claims] An AC power is converted to a DC power or a DC power is converted to an AC power by connecting a plurality of switching elements in a bridge, and a DC voltage applied between DC terminals is supplied to a positive side by two sets of smoothing capacitors. In the control device of the NPC converter of the pulse width modulation system, which divides the DC signal into a DC voltage and a negative DC voltage, and applies an ignition signal obtained by comparing the modulation signal and the carrier signal to each of the switching elements, A control device for an NPC converter, wherein a large-amplitude carrier and a small-amplitude carrier are used together to suppress fluctuations in neutral point potential. 2. A method according to claim 1, wherein each phase modulation signal is crossed with a large amplitude carrier according to an electrical angle and a magnitude of a voltage command vector,
2. The NP according to claim 1, further comprising a carrier designating means for designating whether to cross the small amplitude carrier.
Control device for C converter. 3. A bridge connection of a plurality of switching elements to convert AC power to DC power or DC power to AC power, and to convert a DC voltage applied between DC terminals to a positive side by two sets of smoothing capacitors. A pulse width modulation type NPC converter control device that divides the DC voltage into a DC voltage and a negative DC voltage and supplies an ignition signal to each of the switching elements; a means for generating a static coordinate system voltage command vector; Two-phase to three-phase conversion means for converting a voltage command vector into a voltage command signal in a three-phase coordinate system, and a voltage command in a space vector diagram in which the static coordinate system voltage command vector is input and a predetermined area is divided. A carrier selection signal and space for designating whether each phase modulation signal crosses a large amplitude carrier or a small amplitude carrier according to an area where a vector exists. A spatial vector region determining means for outputting a vector region number, a voltage command signal of the three-phase coordinate system and the region number as inputs, processing the three-phase coordinate system voltage command signal, and outputting a modulation signal Modulation signal processing means, and a carrier generation means for outputting one of a large amplitude carrier and a small amplitude carrier for each phase according to the carrier selection signal, and comparing the processed modulation signal with the carrier signal,
The control device for an NPC converter according to claim 1, further comprising: comparison means for outputting a firing signal to be applied to the NPC converter based on the signal. 4. The modulation signal processing unit according to claim 3, wherein the modulation signal processing means processes the modulation signal so as not to output a vector on a side of a regular hexagon outside a space vector diagram of the NPC converter. Control device for NPC converter. 5. An input of the static coordinate system voltage command signal, the output current of each phase of the NPC converter, and the smoothing capacitor voltage, and has two firing options among the output possible vectors in the NPC converter space vector diagram. A neutral point potential control means for outputting a firing option selection signal specifying which firing option of the inner regular hexagonal vertex is to be selected, and comparing the processed modulation signal with the carrier signal And
A comparison means for outputting a temporary ignition signal based on the input signal; a ignition state determination means for receiving the temporary ignition signal and the ignition option selection signal and determining an ignition signal to be supplied to the NPC converter; The control device for an NPC converter according to claim 3, comprising: 6. A bridge connection of a plurality of switching elements to convert AC power to DC power or DC power to AC power, and to convert a DC voltage applied between DC terminals to a positive side by two sets of smoothing capacitors. A pulse width modulation type NPC converter control device that divides the voltage into a DC voltage and a negative DC voltage and provides an ignition signal to each of the switching elements, wherein a voltage command vector generating means for outputting a voltage command vector in a stationary coordinate system; A two-phase to three-phase conversion means for converting the stationary coordinate system voltage command vector into a three-phase coordinate system voltage command signal; and a carrier generation means for generating one of a large vibration carrier and a small amplitude carrier for each phase. The static coordinate system voltage command signal, the output current of each phase of the NPC converter, and the smoothing capacitor voltage can be input and output in the NPC converter space vector diagram. Neutral point potential control means for outputting a firing option selection signal for designating which firing option of a vector having two firing options in the vector, the three-phase coordinate system voltage command Comparing means for comparing a signal with the carrier signal and outputting a temporary firing signal of the NPC converter; and determining a firing signal to be applied to the NPC converter based on the temporary firing signal and the firing option selection signal. , An ignition signal determination means, and a control device for the NPC converter. 7. The neutral point potential control means sets the neutral point potential and the neutral point current in the control section at the current sampling time for all possible ignition option combinations in the current control cycle. The neutral point potential at the next sampling time is predicted from the value integrated in step (1) and the capacitance of the capacitor.
7. The control device for an NPC converter according to claim 6, wherein a combination of the ignition options close to is selected. 8. A plurality of switching elements are connected in a bridge to convert AC power to DC power or DC power to AC power, and direct current applied between DC terminals to the positive side by two sets of smoothing capacitors. A pulse width modulation type NPC converter control device that divides the DC voltage into a DC voltage and a negative DC voltage and supplies an ignition signal to each of the switching elements; a means for generating a static coordinate system voltage command vector; Two-phase to three-phase conversion means for converting a voltage command vector into a voltage command signal in a three-phase coordinate system; a voltage command vector in a space vector diagram having a predetermined area divided by inputting the static coordinate system voltage command vector. A carrier selection signal and a space that specify whether each phase modulation signal should cross a large amplitude carrier or a small amplitude carrier according to the region where the vector exists. A space vector region determining means for outputting a vector region number; a modulation signal for processing the three-phase coordinate system voltage command signal and outputting a modulation signal with the three-phase coordinate system voltage command signal and the region number as inputs; Processing means; carrier generation means for outputting either a large amplitude carrier or a small amplitude carrier for each phase in accordance with the carrier selection signal; the stationary coordinate system voltage command signal, the NPC converter phase output current, and With the smoothing capacitor voltage as an input, a firing option selection signal for specifying which one of the vectors having two firing options is to be selected from among the output possible vectors in the NPC converter space vector diagram. Neutral point potential control means for outputting, comparing the processed modulated signal and the carrier signal,
Comparing means for outputting a temporary firing signal of the NPC converter; and ignition signal determining means for determining a firing signal to be applied to the NPC converter based on the temporary firing signal and the firing option selection signal. A control device for an NPC converter. 9. The neutral point potential control means determines the neutral point potential and the neutral point current at the current sampling time for all possible ignition option combinations in the current control cycle. The neutral point potential at the next sampling time is predicted from the value integrated in step (1) and the capacitance of the capacitor.
9. The control device for an NPC converter according to claim 8, wherein a combination of the ignition options close to is selected. 10. The NPC converter according to claim 6, wherein either the actual current value of each phase or the current reference value of each phase is used for the neutral point potential prediction control. Control device. 11. The action of the neutral point potential control is as follows:
Always perform, always determine the firing state based on the temporary firing signal and the firing option selection signal, or perform when the value of the neutral point potential at the time of this sampling exceeds a certain allowable width, and exceed the allowable width If not, the temporary ignition signal is set to the ignition state as it is, or the neutral point potential predicted value at the next sampling is likely to exceed a certain allowable cap. If the arc signal is set to the firing state as it is,
The method according to any one of claims 8 to 10, wherein
The control device for an NPC converter according to any one of the above. 12. In the carrier generation means, the principle is to always use small vibration carriers for all three phases, and the value of the neutral point potential at the time of the current sampling exceeds a certain allowable radiation, or 11. The NPC converter according to claim 8, wherein the large-amplitude carrier and the small-amplitude carrier are used together when the value of the neutral point potential at the next sampling time is likely to exceed a certain allowable width. Control device. 13. An NPC converter comprising a plurality of switching elements connected in a bridge to convert AC power into DC power, and an NPC inverter comprising a plurality of switching elements connected in a bridge to convert DC power into AC power. Connecting the NPC converter and the positive and negative DC terminals of the NPC inverter to each other, and connecting two sets of smoothing capacitors for dividing a DC voltage applied between the DC terminals of the NPC inverter and the NPC converter; A control device for an NPC converter that performs pulse width modulation on a switching element, wherein the DC power on the positive side of the NPC inverter is supplied to a positive input terminal of a first subtractor, and the DC power on the negative side of the NPC inverter is supplied to the first subtractor. To the negative side input terminal of the subtractor to determine the inverter side DC power deviation, and the positive side DC power of the NPC converter The negative-side DC power of the NPC converter is input to the positive-side input terminal of a second subtractor and the negative-side input terminal of the second subtractor, respectively, to obtain a converter-side DC power deviation. The power deviation is input to a positive input terminal of a third subtractor, and the converter-side DC power deviation is input to a negative input terminal of the third subtractor to obtain a power deviation. A controller for an NPC converter, wherein a compensation output obtained via a gain is input to a neutral point compensation circuit of the NPC inverter. 14. An NPC converter comprising a plurality of switching elements connected in a bridge to convert AC power into DC power, and an NPC inverter comprising a plurality of switching elements connected in a bridge to convert DC power into AC power. Connecting the NPC converter and the positive and negative DC terminals of the NPC inverter to each other, and connecting two sets of smoothing capacitors for dividing a DC voltage applied between the DC terminals of the NPC inverter and the NPC converter; A control device for an NPC converter that performs pulse width modulation on a switching element, wherein the positive DC power of the NPC inverter is supplied to a positive input terminal of a first subtractor, and the positive DC power of the NPC converter is supplied to a first input terminal of a first subtractor. Input to each of the negative input terminals of the first subtractor via a gain to determine a positive DC power deviation; To the positive input terminal of the second subtractor and the negative DC power of the NPC converter to the negative input terminal of the second subtractor via a first gain. The positive DC power deviation is input to the positive input terminal of a third subtractor, and the negative DC power deviation is input to the negative input terminal of the third subtractor. A control device for an NPC converter, wherein a deviation is obtained, and a compensation output obtained from the power deviation via a second gain is input to a neutral point compensation circuit of the NPC inverter. 15. An NPC converter comprising a plurality of switching elements connected in a bridge to convert AC power to DC power, and an NPC inverter comprising a plurality of switching elements connected in a bridge to convert DC power to AC power. Connecting the NPC converter and the positive and negative DC terminals of the NPC inverter to each other, and connecting two sets of smoothing capacitors for dividing a DC voltage applied between the DC terminals of the NPC inverter and the NPC converter; A control device for an NPC converter that performs pulse width modulation on a switching element, wherein the DC power on the positive side of the NPC inverter is supplied to a positive input terminal of a first subtractor, and the DC power on the negative side of the NPC inverter is supplied to the first subtractor. To the negative side input terminal of the subtractor to determine the inverter side DC power deviation, and the positive side DC power of the NPC converter The negative-side DC power of the NPC converter is input to the positive-side input terminal of the second subtractor and the negative-side input terminal of the second subtractor, respectively, to determine a converter-side DC power deviation. When the power deviation, which is the output of the subtractor, exceeds a predetermined range, it is input to a function generator that generates an output. The inverter-side DC power deviation is input to the positive input terminal of a third subtractor. The output of the third subtractor is input to the negative side input terminal of the third subtractor via a first gain to obtain a power deviation, and the power deviation as the output of the third subtractor is calculated by a second gain. A compensation output obtained through the NPC inverter is input to a neutral point compensation circuit of the NPC inverter. 16. An NPC converter comprising a plurality of switching elements connected in a bridge to convert AC power to DC power, and an NPC inverter comprising a plurality of switching elements connected in a bridge to convert DC power to AC power. Connecting the NPC converter and the positive and negative DC terminals of the NPC inverter to each other, and connecting two sets of smoothing capacitors for dividing a DC voltage applied between the DC terminals of the NPC inverter and the NPC converter; A control device for an NPC converter that performs pulse width modulation on a switching element, wherein the DC power on the positive side of the NPC inverter is supplied to a positive input terminal of a first subtractor, and the DC power on the negative side of the NPC inverter is supplied to the first subtractor. To the negative side input terminal of the subtractor to determine the inverter side DC power deviation, and the positive side DC power of the NPC converter The negative-side DC power of the NPC converter is input to the positive-side input terminal of a second subtractor and the negative-side input terminal of the second subtractor, respectively, to obtain a converter-side DC power deviation. The power deviation is input to the positive input terminal of the third subtractor, and the multiplier is configured to multiply the converter-side DC power deviation by a constant, and to the negative input terminal of the third subtractor via the first gain. A power deviation is obtained by inputting the power deviation, a compensation output obtained via the second gain is input to a neutral point compensation circuit of the NPC inverter, and an output terminal of the first subtractor and the third terminal Between the positive input terminal of the subtractor and the input terminal of a function generator that generates an output when the inverter side DC power deviation exceeds a predetermined range, and an output is generated from the output terminal of the function generator. The multiplier Together give calculation start command, the control device of the NPC converter, wherein the first gain is to be connected to the negative input terminal of the third subtractor. 17. An NPC converter having a plurality of switching elements connected in a bridge and converting AC power into DC power, and an NPC inverter having a plurality of switching elements connected in a bridge and converting DC power into AC power. Connecting the NPC converter and the positive and negative DC terminals of the NPC inverter to each other, and connecting two sets of smoothing capacitors for dividing a DC voltage applied between the DC terminals of the NPC inverter and the NPC converter; A control device for an NPC converter that performs pulse width modulation on a switching element, wherein the DC power on the positive side of the NPC inverter is supplied to a positive input terminal of a first subtractor, and the DC power on the negative side of the NPC inverter is supplied to the first subtractor. To the negative side input terminal of the subtractor to determine the inverter side DC power deviation, and the positive side DC power of the NPC converter The negative-side DC power of the NPC converter is input to the positive-side input terminal of a second subtractor and the negative-side input terminal of the second subtractor, respectively, to obtain a converter-side DC power deviation. The power deviation is input to a positive input terminal of a third subtractor, and the converter-side DC power deviation is calculated via a parallel circuit including two normally-closed and normally-open interlocking contacts and first and third gains. DC power deviation is obtained by inputting to each negative input terminal of the third subtractor, and a compensation output obtained through the second gain is input to a neutral point compensation circuit of the NPC inverter, An input terminal of a function generator that generates an output when the DC power deviation exceeds a predetermined range is connected between an output terminal of the third subtractor and an input terminal of the second gain, and Output from the output terminal of the Together provide a switching command to come the second interlocking contact, the controller of the NPC converter, characterized in that from the first gain to switch to the third gain. 18. One NPC for converting AC power into DC power by bridging a plurality of switching elements.
A converter, and n (n = integer of 2 or more) NPC inverters each of which is a bridge connection of a plurality of switching elements and converts DC power into AC power; The NP
A positive DC terminal of the C converter is commonly connected, and a negative DC terminal of the NPC converter is commonly connected to each negative DC terminal of each of the NPC inverters. An NPC converter control device for connecting two sets of smoothing capacitors for dividing a DC voltage applied between terminals and for pulse width modulating each of the switching elements, comprising: To the positive input terminal of the first subtractor and the positive DC power of the NPC converter to the negative input of the first subtractor via the product of the first and nth gains. Input to each of the NPC inverters to obtain a positive DC power deviation. One of the NPC inverters, the negative DC power is input to a positive input terminal of a second subtractor, and the NPC The negative DC power of the inverter is input to the negative input terminal of the second subtractor via the product of the first gain and the nth gain to obtain a negative DC power deviation, The positive DC power deviation is input to the positive input terminal of a third subtractor, and the negative DC power deviation is input to the negative input terminal of the third subtractor to determine a power deviation. A control device for an NPC converter, wherein a compensation output obtained via a second gain is input to a neutral point compensation circuit of the NPC inverter.