JP2011239564A - Neutral point clamp type power conversion equipment - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a mechanism for accurately constraining a ripple of neutral point current and neutral point potential of a three-phase NPC converter for even high frequency component with an easy control algorithm.SOLUTION: A neutral point clamp type power conversion equipment according to an embodiment of the present invention comprises: a neutral point clamp (NPC) type three-phase inverter that is connected to a DC power supply having three potentials to drive three-phase load and can output three levels, means 12 for calculating a three-phase voltage command; neutral point potential fluctuation constraining means 23 for calculating the compensation amount of zero-phase bias voltage in order to constrain the fluctuation of the neutral point potential according to the three-phase voltage command and a current feeding into the load; and means 14 for correcting the three-phase voltage command by adding the compensation amount of zero-phase bias voltage to the three-phase voltage command. The neutral point potential fluctuation constraining means 23 comprises means for calculating the compensation amount of the zero-phase bias voltage according to at least the absolute value and a sign of the three-phase voltage command, and the load current.

Description

  The present invention relates to a neutral point clamp type power converter used for a railway vehicle or the like.

  In a railway vehicle, a three-level output NPC converter may be used for a main converter (single-phase converter + three-phase inverter). This system is characterized by less harmonics than a normal two-level output converter. Since the high-frequency band of the current flowing through the rail is used as a train signal, the system with small harmonics is suitable.

  However, a high frequency three times the inverter output frequency, which is a phenomenon unique to the system, is generated. The frequency spectrum component of this high frequency moves with the fluctuation of the inverter output frequency, and crosses the signal band (asynchronous component in the vicinity of 500 Hz to 2 kHz) used in ATC or the like, which may cause a problem (such a signal band of 500 Hz to 2 kHz). Depending on whether it is present and its tolerance level). This is because the current flowing from the PC inverter to the neutral point causes the potential at the neutral point to pulsate at three times the inverter output frequency, and this component flows from the single-phase NPC converter to the return line (rail).

  In particular, in the case of a railway vehicle, the return current allowable value of the signal frequency band such as the ATC is on the order of 1 / 10,000 of the fundamental current (50 Hz / 60 Hz), which is a weak level, and strict suppression is required. Is done.

  With respect to this problem, registration 3585733 separates the neutral point of the inverter and the converter so that the influence of the inverter does not appear in the converter output. However, since capacitors are individually divided into converters and inverters, there are problems such as an increase in total capacity for stabilizing the DC.

  On the other hand, as a control method for suppressing the fluctuation of the neutral point potential, registration 3182322 has a method of superimposing a constant bias voltage on the three-phase output voltage of the inverter. However, there is a problem that it cannot be used in a region where the output voltage is high due to the limitation of the magnitude of the bias voltage.

  Although there is a method of controlling the bias voltage amount according to the unbalance amount of the neutral point potential, since feedback control is performed, the inverter output frequency becomes high and the neutral point potential pulsation frequency becomes high. The effect is not obtained.

  Further, as a feedforward method, Japanese Patent Laid-Open No. 8-317663 proposes a calculation method in which a three-phase NPC inverter theoretically does not cause a neutral point potential fluctuation, but the calculation formula is complicated and practical use is not possible. was difficult.

  An object of the present invention is to realize a method for accurately suppressing pulsation of a neutral point current and a neutral point potential of a three-phase NPC converter to a high frequency component with a simple control algorithm.

  A neutral point clamp type power converter according to an embodiment of the present invention is a neutral point clamp type (NPC) capable of outputting three levels, which is connected to a DC power source having three potentials and drives a three-phase load. Three-phase inverter, means 12 for calculating a three-phase voltage command, and zero-phase bias voltage compensation so as to suppress fluctuations in the neutral potential according to the three-phase voltage command and the current flowing through the load current A neutral point potential fluctuation suppressing means 23 for calculating the amount, and a means 14 for correcting the zero phase bias voltage compensation amount by adding it to the three-phase voltage command. And means for calculating the zero-phase bias voltage compensation amount according to at least the absolute value of the three-phase voltage command, its sign and the load current.

  A system for accurately suppressing pulsation of the neutral point current and neutral point potential of the three-phase NPC converter up to a high frequency component with a simple control algorithm is realized.

The figure which shows the structure of the neutral point clamp type power converter device which concerns on one Example of this invention. The figure which shows the detailed structure of the zero phase bias compensation amount calculating part which concerns on one Example of this invention. The figure for demonstrating Formula (3). The figure which shows the detailed structure of the zero phase bias compensation amount calculating part which concerns on 2nd Example of this invention. The figure which shows the detailed structure of the zero phase bias compensation amount calculating part which concerns on 3rd Example of this invention. Motor current without this compensation, three-phase modulation rate command before compensation, three-phase modulation rate command after compensation, zero-phase bias compensation amount, neutral point current Inp flowing from the inverter to the neutral point of DC link FIG. Motor current when this compensation is applied, three-phase modulation factor command before compensation, three-phase modulation factor command after compensation, zero-phase bias compensation amount, neutral point current Inp flowing from the inverter to the neutral point of the DC link FIG. Motor current when this compensation is applied, three-phase modulation factor command before compensation, three-phase modulation factor command after compensation, zero-phase bias compensation amount, neutral point current Inp flowing from the inverter to the neutral point of the DC link FIG.

  Embodiments of a neutral point clamp type power converter according to the present invention will be described below with reference to the drawings.

(First embodiment)
A first embodiment is shown in FIG. This embodiment is described for a railway AC electric vehicle drive system. Single-phase AC overhead power is received through the pantograph 1 and the main transformer 2 and converted from AC to DC by a single-phase NPC (neutral point clamp type) PWM converter 3. The wiring connecting the converter 3 and the inverter 6 and the series circuit of the filter capacitors 4-1 and 4-2 constitute a DC link unit. This DC link section has three levels due to the filter capacitors 4-1 and 4-2. A three-phase NPC-PWM inverter 6 is connected to the DC link unit, and the three-phase NPC-PWM inverter 6 drives a main motor (induction motor) 7.

  The control unit for the main circuit controls the main circuit based on vector control. Here, a DQ axis rotation coordinate system is introduced in which the magnetic flux axis of the rotor of the main motor 7 is the D axis and the axis orthogonal to the D axis is the Q axis. Although not shown, the D-axis current command Id * and the Q-axis current command Iq * are determined according to the torque command and speed corresponding to the notch command. In accordance with the DQ axis current commands Id * and Iq *, the slip frequency calculator 17 calculates the slip frequency ws as shown in the following equation.

Ws = R2 / L2 * Iq * / Id * (1)
Here, R2 is the secondary resistance value of the main motor 7, and L2 is the self-inductance of the main motor 7. The main motor 7 has a rotation speed detector 8, and the inverter output frequency W 1 is determined by adding the detected rotation speed Wr and the slip frequency Ws by the adder 18. W1 is integrated by the integrator 19 and becomes a phase difference θdq from the reference axis (for example, U phase) of the stationary coordinate system to the D axis.

  The current of the main motor 7 is detected by the current detector 9 and three-phase DQ converted by the coordinate converter 20 to become the D-axis current Id and the Q-axis current Iq. The D-axis current deviation (= Id * −Id) and the Q-axis current deviation (= Iq * −Iq) are input to the current control unit 10, and each current is supplied to each current command by a PI controller or the like. The DQ axis voltage command (Vd *, Vq *) of the inverter output is determined so as to follow. The coordinate converter 11 calculates three-phase voltage commands Vu *, Vv *, and Vw * by DQ three-phase conversion.

  The voltage of the capacitor 4-1 in the DC link unit is called a positive DC voltage VdcP, and the voltage of the capacitor 4-2 is called a negative DC voltage VdcN. Each is detected by the DC voltage detector 5 and added by the adder 22 to calculate the DC voltage Vdc of the entire DC link unit.

  The three-phase modulation rate command calculation unit 12 calculates the three-phase modulation rate commands Au *, Av *, Aw * from the three-phase voltage commands Vu *, Vv *, Vw * and the DC voltage Vdc as follows:

Au * = Vu * / (Vdc / 2)
Av * = Vv * / (Vdc / 2) (2)
Aw * = Vw * / (Vdc / 2)
Here, the NPC converter has three levels as described above, and its neutral point varies. It is the neutral point potential suppression control unit 23 that suppresses this. The neutral point potential suppression control unit 23 includes a coordinate conversion unit 16, a zero phase bias compensation amount calculation unit 13, and an adder 14. The coordinate converter 16 receives the DQ axis current commands Id * and Iq * and calculates the three-phase current commands Iu *, Iv * and Iw *. The zero-phase bias compensation amount calculation unit 13 calculates a zero-phase bias compensation amount NPcmpAL based on the three-phase modulation rate commands Au *, Av *, Aw * and the three-phase current commands Iu *, Iv *, Iw *. The zero-phase bias compensation amount NPcmpAL is added to the three-phase modulation rate commands Au *, Av *, Aw * by the adder 14, and the corrected three-phase modulation rate commands Au **, Av **, Aw ** are output. Is done.

  The corrected three-phase modulation rate commands Au **, Av **, and Aw ** are compared with the carriers CAR1 and CAR2 for the NPC converter generated by the carrier generating unit 21 in the gate command generating unit 15, and A gate command to each switching element 6 is generated.

  Here, a detailed configuration of the zero-phase bias compensation amount calculation unit 13 which is a characteristic part of the present embodiment is shown in FIG. The basic configuration is to calculate the zero-phase bias compensation amount NPcmpAL by dividing the neutral point current reference InpRef by the neutral point current change amount InpRef2.

  The neutral point current reference InpRef is a motor current Iu, Iv, Iw that matches the three-phase current commands Iu *, Iv *, Iw * by switching the inverter according to the three-phase modulation rate commands Au *, Av *, Aw *. When the current flows, the current flowing to the neutral point is predicted. The neutral point potential Vnp is discharged when the neutral point current InINV of the inverter flows from the neutral point (the connection point between the capacitors 4-1 and 4-2) as indicated by the arrow in the figure, and the capacitor 4-2 is discharged. Since 4-1 is charged, it goes down. Conversely, when the neutral point current InINV flows into the neutral point, the capacitor 4-2 is charged and the capacitor 4-1 is discharged, so that the neutral point potential Vnp rises. One embodiment of the present invention aims to minimize this neutral point current as much as possible.

  On the other hand, the neutral point current change amount InpRef2 changes how much the neutral point current InINV changes when the zero-phase bias compensation amount NPcmpAL is set to 1 (= 100%) and the adder 14 compensates the three-phase modulation rate. That is, the sensitivity to the zero-phase bias compensation amount of the neutral point current. Accordingly, the zero-phase bias compensation amount NPcmpAL can be obtained by dividing the neutral point current reference InpRef, which is a predicted value of the neutral point current, by the neutral point current change amount InpRef2 by the divider 27. More practically, the output of the divider 27 includes a limiter 28 for limiting the value of the zero-phase bias compensation amount, a filter 29 for removing unnecessary high-frequency components generated by the calculation, and neutral point potential suppression by the zero-phase bias compensation. A changeover switch 30 for controlling the validity / invalidity of the control is provided.

  Details of the calculation of the neutral point current reference InpRef and the neutral point current change amount InpRef2 will be described below.

  The neutral point current reference InpRef is calculated by adding the predicted neutral point current value of each phase by the adder 26. For example, the predicted neutral point current value of the U phase can be calculated by subtracting the absolute value of the U phase modulation rate command Au * from 1 and multiplying it by the U phase current command Iu *, as will be described later.

  The neutral point current change amount InpRef2 is also calculated by adding the neutral point current change amount of each phase by the adder 31. For example, in the case of the U-phase neutral point current change amount, the sign of the U-phase modulation rate command Au * is basically calculated by the encoder 32 (1 if the input is positive, -1 if the input is negative), By multiplying the calculated value and the U-phase current command Iu *, the neutral phase current change amount of the U-phase can be obtained. In addition, a high-precision compensation unit 40 (described later) is provided as compensation for more strictly suppressing the neutral point current.

  Thus, the neutral point current can be suppressed. The configuration of this embodiment is based on the following idea.

  When zero-phase bias voltage compensation is not performed, the neutral point current of the U phase can be calculated by the following equation.

U-phase neutral point current = (1- | Au * |) × Iu (3)
FIG. 3 is a diagram for explaining the expression (3), and shows a waveform processed by the gate generation unit 15. Although only the U phase will be described here, the same applies to the V phase and the W phase. Although the modulation factor Au * is a sine wave, it shows a waveform expanded in the time direction, and is almost a straight line. When the modulation factor Au * is a positive value, the modulation factor Au * crosses the triangular wave carrier CAR1 as shown in FIG.

  The neutral point current flows only when the gates gu1, gu2, gu3, and gu4 of the inverter 6 in FIG. 1 are off, on, on, and off. That is, when the modulation factor Au * is located inside the carrier CAR1 (downward in FIG. 3), a neutral point current flows. As indicated by triangles a, b, and g, a neutral point current flows when Au * is between periods d and f. Similarly, when the modulation factor Au * takes a negative value and crosses the carrier CAR2, a neutral point current flows when the modulation factor Au * is located inside the carrier CAR2 (upper in FIG. 3).

  Consider right triangles a, b, and c including hypotenuses a and b of carrier CAR1. In this case, the neutral point current flows between d and e. The modulation factor Au * is a value from 0 to 1, and passes the point a when it is “1”. Since the triangles a, b, c and the triangles a, d, e are congruent, the modulation rate Au * is equal to the line segment length ratio ae / ac. Therefore, when the length of the base bc is 1, the length of the line segment de is 1-Au *. That is, the time ratio at which the neutral point current flows is 1-Au *. Considering the case where the modulation factor Au * is negative, the neutral point current flows at a time ratio of (1− | Au * |) as shown in Expression (3).

  The total neutral point current flowing through the neutral point current, that is, the neutral point current reference InpRef can be calculated by calculating and adding Equation (3) for each phase of the UVW phase. On the other hand, when the zero-phase bias voltage compensation amount NPcmpAL is added, the U-phase neutral point current can be calculated by the following equation.

“U-phase neutral point current” = (1− | Au * + NPcmpAL |) × Iu
(4)
When the expressions (3) and (4) are compared and the signs of Au * and Au * + NPcmpAL are the same, the amount of change in the neutral point current due to the zero-phase bias voltage compensation amount NPcmpAL is as follows.

“U-phase neutral point current change due to zero-phase bias voltage compensation” =
-Sign (Au *) × NPcmpAL × Iu (5)
This equation (5) is for the case where the zero-phase bias voltage compensation amount is NPcmpAL, and if it is normalized to be 100%, equation (6) is obtained.

“Neutral point current change amount” = − Sign (Au *) × Iu (6)
If this is calculated for the UVW phase and added, the neutral point current change amount InpRef2 can be obtained.

  Therefore, the zero-phase bias voltage compensation amount NPcmpAL may be considered as follows to make the neutral point current zero.

“Zero-phase bias voltage compensation amount NPcmpAL” =
“Sum of formula (3) (UVW phase)” / “Sum of formula (6) (UVW phase)” (7)
On the other hand, it is assumed that “Au * and Au * + NPcmpAL” have the same sign as described above. This assumption does not necessarily hold. For more precise compensation, the following is considered.

  The above assumption may not be satisfied when | Au * | is smaller than | NPcmpAL |. In this case, the sign of Au * + NPcmpAL in equation (4) may be referred to. This method is shown in FIG. 4 as a second embodiment.

First, the absolute value calculators 35 and 36 calculate the absolute value of the U-phase modulation rate command Au ** and the absolute value of the compensated U-phase modulation rate command Au **. When the comparator 37 compares the two, and the U-phase modulation rate command Au * is larger than the compensated U-phase modulation rate command Au **, the switch 33 makes the U-phase modulation rate command as described above. Multiply the sign of Au * and the U-phase current command Iu * to obtain the U-phase neutral point current change amount. On the other hand, when the U-phase modulation rate command Au * is smaller than the compensated U-phase modulation rate command Au ** by the comparator 37, the sign of the zero-phase bias compensation amount NPcmpAL used for the previous calculation by the encoder 38 The U-phase current command Iu * is multiplied to obtain the U-phase neutral point current change amount. The Z inverse (Z ⁻¹ ) 41 is a storage unit that stores and outputs a value used for the previous calculation.

  Considering that | Au * | is small, it is also possible to refer only to the sign of NPcmpAL instead of Au * + NPcmpAL. In the first embodiment shown in FIG. 2, an example in which the reference numeral NPcmpAL is referred to is shown.

  Furthermore, as described above, it is possible to perform the calculation using only the sign of Au * + NPcmpAL, instead of dividing the cases by the sizes of | NPcmpAL | and | Au * |. This is shown in FIG. 5 as a third embodiment. Compared with the first embodiment and the second embodiment, it can be realized with the simplest configuration without being classified.

  FIG. 6 shows the motor current (Iu, Iv, Iw) when the inverter output frequency is 30 Hz, each phase current amplitude is 300 A, the modulation factor is 70%, and the power factor is 0.9, and the three-phase modulation factor command (Au *) before compensation. , Av *, Aw *), three-phase modulation rate command after compensation (Au **, Av *, Aw **), zero-phase bias compensation amount NPcmpAL, neutral point flowing from the inverter to the neutral point of the DC link The current Inp is shown.

  FIG. 6A shows the case without this compensation, and it can be seen that a neutral point current Inp that pulsates at three times the inverter output frequency flows. As a result, the neutral point voltage Vnp of the DC link portion pulsates.

  FIG. 6B shows the result of the present example shown in FIGS. 1 and 2. It can be seen that the neutral point current is not strictly zero, but is suppressed to almost zero globally. Thereby, although not shown in figure, the pulsation of the neutral point voltage Vnp of a DC link part can be suppressed. It turns out that a big effect is acquired irrespective of a simple control algorithm.

  FIG. 6C shows the result when the high-precision compensator 40 is not provided. Compared to FIG. 6B, it can be seen that a spike-like current flows. It can be seen that the effect of the high precision compensator 40 can be confirmed, and even if the high precision compensator 40 is not provided, a great effect can be obtained in suppressing the neutral point current and the neutral point voltage Vnp.

  Further, the limiter 28 in FIG. 2 is effective in limiting the zero-phase bias compensation amount. The zero-phase bias compensation amount is calculated by dividing the neutral point current reference InpRef by the neutral point current change amount InpRef2. The three-phase voltage command, current command, or current command, which is the calculated input value, is obtained. The current used instead of includes noise and ripple, and the zero-phase bias voltage compensation amount obtained by dividing the current may vary greatly. By limiting this to a significant range, overcurrent and control failure can be avoided.

  Also, the filter 29 in FIG. 2 can avoid overcurrent and control failure by limiting the zero-phase bias voltage compensation amount including ripples and pulsations similar to those described above to a necessary band.

  As described above, it is only necessary to compensate for a frequency component that is three times the inverter output frequency, and it is desirable to remove components higher than that. In an AC electric vehicle or the like, the inverter uses the output voltage high until the 1-pulse mode. In such a one-pulse mode, the neutral point current is not upstream in principle. Therefore, it is effective to set the cut-off frequency of the filter higher than three times the output frequency at which the one-pulse mode is set, not the highest frequency of the inverter output frequency.

  As described above, the fluctuation of the neutral point voltage can be suppressed by suppressing the pulsating component of the neutral point current.

  In an AC electric vehicle system, it is necessary to avoid that a current flowing in an overhead wire interferes with a signal system such as a track circuit (induction failure). For example, some signals used in the automatic train control device (ATC) use a 500 Hz to 2000 Hz band of return current (current flowing in the rail). This allowable signal level is very weak. In general, the PWM converter has an odd-order or even-order higher harmonic component synchronized with the power supply, and an asynchronous component therebetween generates only a low level. However, when the PWM converter is an NPC converter and the neutral point voltage pulsates, an asynchronous component is generated according to the pulsation frequency of the neutral point voltage and its magnitude. The pulsation frequency is three times the inverter output frequency, and “feedback compensation” in which the converter detects and compensates for it is not effective because the frequency is high and the allowable level is weak. On the other hand, with this configuration, the converter side current as described above is suppressed by suppressing the pulsation of the neutral point voltage itself by performing feedforward compensation on the inverter side which is the pulsation source of the neutral point voltage. Asynchronous components contained in can be reduced. In particular, the feedforward calculation in the known example is complicated, requires a processing load (calculation time, code amount) on the microcomputer, and lacks feasibility. While the calculation according to the present embodiment has a part that is not an exact solution, the algorithm is simple and there is no problem in implementation.

  In this embodiment, the current control 10 is provided so that each current command matches the actual current, and the actual currents Id, Iq match the DQ axis current commands Id *, Iq *. That is, it can be considered that the three-phase current commands Iu *, Iv *, Iw * and the actual three-phase currents Iu, Iv, Iw match. Therefore, the current commands Iu *, Iv *, and Iw * are used for the calculation of the neutral point current reference InpRef and the neutral point current change amount InpRef2. Can be obtained. In this embodiment, the neutral point potential compensation method for the three-phase modulation rate commands Au *, Av *, and Aw * is shown. However, as shown by the equation (2), the three-phase modulation rate command and the three-phase The voltage command has a direct current voltage double relationship and can be considered equivalent. Therefore, it is possible to construct a similar neutral point potential compensation method for the three-phase voltage command.

  DESCRIPTION OF SYMBOLS 1 ... Pantograph, 2 ... Main transformer, 3 ... Single phase NPC converter, 4 ... Filter capacitor, 5 ... DC voltage detector, 6 ... NPC three phase inverter, 7 ... Electric motor, 8 ... Rotation speed detector, 23 ... Medium Sex point potential suppression control unit.

Claims (10)

A neutral-point clamped (NPC) three-phase inverter that is connected to a DC power source having three potentials to drive a three-phase load and that can output three levels;
Means for calculating a three-phase voltage command;
Neutral point potential fluctuation suppression means for calculating a zero-phase bias voltage compensation amount so as to suppress fluctuations in the neutral point potential according to the three-phase voltage command and the current flowing through the load current;
Means for adding and correcting the zero-phase bias voltage compensation amount to the three-phase voltage command,
The neutral point potential fluctuation suppressing means includes a means for calculating the zero-phase bias voltage compensation amount according to at least the absolute value of the three-phase voltage command, its sign, and the load current. Point clamp type power converter. A neutral-point clamped (NPC) three-phase inverter that is connected to a DC power source having three potentials to drive a three-phase load and that can output three levels;
Means for calculating a three-phase voltage command;
Neutral point potential fluctuation suppression means for calculating a zero-phase bias voltage compensation amount so as to suppress fluctuations in the neutral point potential according to the three-phase voltage command and the current flowing through the load current;
Means for adding and correcting the zero-phase bias voltage compensation amount to the three-phase voltage command, and the neutral point potential fluctuation suppressing means,
Neutral point current estimating means for calculating a neutral point current flowing in the neutral point according to the absolute value and load current of the three-phase voltage command;
Neutral point current change amount estimating means for estimating a change amount of a neutral point current that changes by applying a zero-phase bias voltage compensation of a predetermined unit according to the sign of the three-phase voltage command and the load current;
A neutral point clamp type power conversion device comprising: a neutral point current estimated value and a means for calculating a zero phase bias voltage compensation amount according to a neutral point current change amount.   The neutral point current change amount estimation means sets the neutral point according to the sign of the zero-phase bias voltage compensation amount or the sign of the three-phase voltage command after addition correction in addition to the sign of the three-phase voltage command and the load current. The neutral point clamp type power converter according to claim 1 or 2, wherein the amount of change of the flowing neutral point current is estimated. A neutral-point clamped (NPC) three-phase inverter that is connected to a DC power source having three potentials to drive a three-phase load and that can output three levels;
Means for calculating a three-phase voltage command;
Neutral point potential fluctuation suppression means for calculating a zero-phase bias voltage compensation amount so as to suppress fluctuations in the neutral point potential according to the three-phase voltage command and the current flowing through the load current;
Means for adding and correcting the zero-phase bias voltage compensation amount to the three-phase voltage command,
The neutral point potential fluctuation suppressing means has means for calculating a zero-phase bias voltage compensation amount according to the absolute value of the three-phase voltage command, the sign of the addition-corrected three-phase voltage command, and the load current. A neutral point clamp type power converter. A neutral point clamp (NPC) three-phase inverter capable of driving a three-phase load connected to a DC power source having three potentials and outputting three levels, and a means for calculating a three-phase voltage command;
Neutral point potential fluctuation suppression means for calculating a zero-phase bias voltage compensation amount so as to suppress fluctuations in the neutral point potential according to the three-phase voltage command and the current flowing through the load current;
Means for adding and correcting the zero-phase bias voltage compensation amount to the three-phase voltage command,
The neutral point potential fluctuation suppressing means is:
Neutral point current estimating means for calculating a neutral point current flowing in the neutral point according to the absolute value and load current of the three-phase voltage command;
Neutral point current change amount estimation means for estimating a change amount of a neutral point current that changes by applying a zero-phase bias voltage compensation of a predetermined unit in accordance with the sign of the addition corrected three-phase voltage command and the load current; ,
A neutral point clamp type power converter comprising means for calculating a zero phase bias voltage compensation amount according to the neutral point current estimated value and the neutral point current change amount.   6. The neutral point potential fluctuation suppressing means for calculating the zero-phase bias voltage compensation amount includes limit means for limiting an upper limit and a lower limit of the zero-phase bias voltage compensation amount. The neutral point clamp type power converter device given in any 1 paragraph.   The neutral point potential fluctuation suppressing means for calculating the zero-phase bias voltage compensation amount includes a zero-phase bias voltage compensation amount filter. Point clamp type power converter.   8. The neutral point clamp type power converter according to claim 7, wherein the cutoff frequency of the filter is set to be higher than three times the inverter output frequency.   8. The neutral point clamp type power converter according to claim 7, wherein the cutoff frequency of the filter is set to be higher than a frequency that is three times the frequency at which the inverter has a maximum output voltage.   The neutral point clamp type power converter is applied to an inverter of a main converter of an AC electric vehicle, and the three phase load of the neutral point clamp type power converter is an electric motor, and the inverter A neutral point clamp type PWM converter is connected to the DC link section of the DC converter, and the AC side of the PWM converter is connected to a single-phase AC overhead line via a pantograph and a main transformer. The neutral point clamp type power converter device according to any one of claims 1 to 9.

Images