JPH0937592A - Pwm controller and control method for three level inverter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress fluctuation of neutral potential, at the time of controlling a three level inverter by space vector system, by generating an output voltage vector using an instantaneous load current. SOLUTION: A first means 1 divides the 360 deg. entire space of a vector into twelve sections and outputs a vector command V. A second means 2 determines the number of a section where a command vector V is present from a rotational angle and a third means 3 calculates a modulation factor α from the magnitude of vector. A fifth means 5 calculates a current ratio corresponding to each section number from the section number of command vector V and an instantaneous level of load current detected by a current detection means 4. A sixth means 6 determines transmission system and sequence for suppressing the neutral potential a three level inverter based on the modulation factor α and the current ratio. A seventh means 7 the output time of each vector based on the transmission system and sequence and generates a PWM control signal for controlling a voltage type inverter 8.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a PWM control method and control device for a voltage type three-level inverter used as a control power source for driving a motor in general industry and railway fields.

[0002]

2. Description of the Related Art Three-level inverters are becoming popular mainly in electric railways as large-capacity, high-voltage inverters. FIG. 7 is a diagram showing a circuit configuration of a three-level inverter. In FIG. 7, 10 is a DC voltage power source, 11 is a DC reactor, 12C1 and 12C2 are capacitors of the same capacity, and both ends P and N are connected to the positive terminal and the negative terminal of the DC voltage power source 10, respectively. Furthermore, the capacitor 12
The neutral point terminal O is led out from the connection point of C1 and 12C2. Reference numerals 13, 14, and 15 are three-phase arms. 13u1 to 13u4, 14v1 of each arm 13, 14, and 15
14v4 and 15w1 to 15w4 are a positive terminal P and a negative terminal N
Is a main switching element group of U, V, and W phases connected in series between and, respectively.
3D4, 14D1 to 14D4 and 15D1 to 15D4.

13D5 to 13D6, 14D5 to 14D6
15D5 to 15D6 are neutral-point clamp diodes connected to the neutral-point terminal O and the connection point between the main switching element groups 1, 2, 3 and 4 with polarities as shown in FIG. In FIG. 7, for example, in the arm 13, when the switching elements 13u1 and 13u2 are turned on, the U-phase terminal becomes + E, and the switching element 13u
When 2 and 13u3 are turned on, the U-phase terminal is 0, and when the switching elements 13u2 and 13u3 are turned on, the U-phase terminal is -E. Similarly, two switching elements for each phase are turned on in pairs, and U,
+ E, 0, -E outputs are generated in the V and W phases. When the output of each phase is 0, current flows in or out of the neutral point terminal O. Also, the sum i of the currents (iu, iv, iw) of each phase
u + iv + iw is 0.

As described above, the potentials of the U-, V-, and W-phase terminals output three types of levels + E, 0, and -E, and the spatial voltage vector thereof is as shown in FIG. In the figure, for example, (01-1) indicates that the voltages of the U, V, and W phases are (0, + E, -E), respectively. Therefore,
It is possible to output 19 types of discrete voltage vectors in 3 ³ = 27 switching states. Therefore, it is possible to select how many vectors are adjacent to the command voltage vector V and control so that the combined vector matches the command vector. In the space vector method, the vector adjacent to the command voltage vector is selected as described above to determine the switching mode, and the switching element is controlled so that the vector is output for a period corresponding to the size of the command voltage vector. , PWM control the inverter.

By the way, in FIG. 7, the neutral point potential Vn of the DC side smoothing capacitor fluctuates depending on the current flowing into or out of the neutral point. Since the fluctuation of the neutral point potential causes the overvoltage of the switching element, it is necessary to suppress the fluctuation of the neutral point potential.

Conventionally, the fluctuation of the neutral point potential in the three-level inverter has been suppressed as follows. In the carrier system, when the carrier frequency is very high and the harmonics of the currents (iu, iv, iw) of each phase are very small, each phase current (i
u, iv, and iw) do not change much and can be regarded as constant. FIG. 9A shows a method for suppressing the neutral point potential fluctuation of the three-level invertor shown in Japanese Patent Application No. 7-170750. The figure shows the command voltage Vu for each phase.
^The figure shows the case where ^* , Vv ^* , Vw ^* are modulated by the carrier A and the carrier B and a PWM signal is output. When the command voltages Vu ^* , Vv ^* , Vw ^{* of} each phase are larger than the carrier A, the voltage + E is When it is output and is smaller than the carrier B, the voltage -E is output, and when it is between the carriers A and B, 0 (neutral point voltage) is output.

FIG. 9B shows a carrier waveform that is modulated as shown in FIG. 9A and the half cycle of the carrier is enlarged. As shown in the figure, the minimum value of the carrier A in the upper row matches the minimum voltage command value of the three phases, and the maximum value of the carrier B in the lower row matches the maximum voltage command value of the three phases. Since carrier A and carrier B are modulated in this way, the period (neutral point voltage output time) in which the command value of each phase is sandwiched between two carriers is the two sides of the parallelogram that face each other and have the same length. It Therefore, the neutral point voltage output time (tu, tv, tw) of each phase per carrier half cycle
Will be the same. That is, tu = tv = tw = Δt.
The amount of charges flowing into and out of the neutral point during this period is given by the following equation (1). (Iu + iv + iw) × Δt = 0 (1) That is, in the carrier system, each phase current (iu, iv, iw) during the period in which the carrier frequency is very high and the neutral point voltage is output is When it can be regarded as constant without changing so much, the fluctuation of the neutral point potential becomes zero.

On the other hand, in order to suppress the fluctuation of the neutral point potential by the space vector method, the neutral point potential detection circuit shown in FIG. 10 is required. That is, in the case of the space vector method, as shown in FIG. 8, there may be a plurality of switching modes that output the selected voltage vector. Therefore, the neutral point potential detection circuit shown in FIG. By utilizing this, the neutral point potential Vn is controlled. In FIG. 10, R is a resistor, 12C1 and 12C2 are capacitors shown in FIG. 7, and AMP is an amplifier. The voltage divided by the resistor R and the voltage at the connection point of the capacitors 12C1 and 12C2 are compared by the amplifier AMP. , Detect the fluctuation of neutral point voltage. The allowable range ΔVn for the fluctuation of the neutral point potential
Is set, and the switching mode is determined so that the neutral point potential Vn is always within this allowable range. When the neutral point potential rises beyond the allowable range, the switching mode (for example, mode (010)) in which the current at the neutral point flows out according to the direction of the already selected voltage vector and load current. select. On the contrary, when it falls, the switching mode (for example, the mode (-10-
1)) is selected.

[0009]

As a method for suppressing the fluctuation of the neutral point potential described above, in the carrier method, it is assumed that the switching frequency is sufficiently high and that the current does not change within the control cycle Tc. is necessary.
Therefore, in this method, a suppression error occurs when the switching frequency is low, and the effect of suppressing the fluctuation of the neutral point potential becomes low. In particular, this method has the drawback that it cannot be applied to the transient process of current.

On the other hand, the known space vector method described above has a drawback in that the neutral point potential detecting device shown in FIG. 10 is required. Further, in order to suppress the fluctuation of the neutral point potential Vn, according to the fluctuation of the neutral point potential Vn, for example, between the two switching modes within the control cycle Tc,
It is necessary to switch the mode (010) and the mode (-10-1) several times. However, this method is not desirable because the number of times of switching (frequency) increases and three phases must switch at the same time.

The present invention has been made to solve the above problems, and the present invention generates an output voltage command vector V of an inverter in accordance with an instantaneous load current, and at the same time, detects the neutral point potential. An object of the present invention is to provide a PWM control method and control device for a three-level inverter that can suppress fluctuations.

[0012]

FIG. 1 is a block diagram showing the principle of the present invention. In the figure, 1 is a first means for sending a vector command, and the first means 1 divides the entire 360 ° space of a vector into 12 sections from section A1 to section A12 shown in FIG. The vector command V is output. The magnitude | V | of the vector command V is the third means 3
The rotation angle [V] of the vector command V is given to the second means 2. The second means 2 uses the rotation angle [V] to determine the section number in which the command vector V exists (8
A1 to A12) are determined. Further, the third means 3 calculates the modulation factor α that determines the output time of the vector from the magnitude | V |. On the other hand, when the load current is detected by the current detection means 4, the fifth means 5 uses the section number of the command vector V and the load current instantaneous value detected by the current detection means 4 to determine each section number. Corresponding current ratio,
For example, iu / iv etc. are calculated. As will be described later, the sixth means 6 determines a preset vector transmission method and transmission sequence based on the modulation rate α calculated by the third means 3 and the current ratio calculated by the fifth means. The seventh means 7 determines the output time of each vector based on the vector transmission method and transmission sequence determined by the sixth means 6, generates a PWM signal, and controls the switching element of the voltage type inverter 8.

Next, assuming that the command vector V is in the section A1 as shown in FIG. 8, the present invention will be described with reference to the vector composition diagram of FIGS. 2 (a) and 2 (b). 2 shows the section A1 of FIG. 8 and its adjacent section. V is a command vector, and Vs1, ..., Vs4, Vs, VL, Vm are shown in FIG.
Corresponding to each vector shown in Fig. 1, T1 is the output time of the vector in the direction 1 in Fig. 7A, T2 is the output time of the vector in the same direction 2, and Tmm is the output time in the direction 1 in Fig. 2B. Of the vector, Tsn is the output time of the vector in the same direction 2, and Tc is the control cycle.

The sixth means 6 is a vector transmission system preset as follows based on the modulation rate α calculated by the third means 3 and the current ratio calculated by the fifth means as described above. And select the calling order. In order to reduce the load current ripple, when the modulation factor α calculated by the third means 3 is 0.5 or less, only the short vector Vs sandwiching the command vector V is inserted, as shown in FIG. use. Further, when the modulation rate α is 0.5 or more, as shown in FIG.
As shown in (b), a short vector Vs and long vectors Vm and VL sandwiching the command vector V are used. Then, in order to suppress the fluctuation of the neutral point potential, the following three transmission methods are defined by focusing on the vector clamped to the neutral point in the transmission sequence during the control cycle Tc.

(1) Modulation rate: When α <0.5. In the range of the modulation rate α, using only the short vector Vs,
As shown in FIG. 3, the command vector V is generated in the transmission order S11 of the transmission method 1. In this case, since the command vector V is short, the transmission sequence S11 generates the command voltage vector only by the two short vectors including the adjacent regions and the zero vector. Then, in the control cycle Tc, the positive side switching mode (0
10) and the negative side switching mode (-10-1) are used alternately. However, the vector change is ordered so that it is performed by one switching.

(2) Modulation rate: When α ≧ 0.5. (a) Transmission method 2 Transmission order 21 This transmission method is used when the command vector V is large, and the command vector V is generated in the transmission order S21 of the transmission method 2 as shown in FIG. That is, the transmission order is S11.
When only a short vector such as
21 is used, and the short vector Vs belonging to the section A1 and the intermediate vector Vm are used. In FIG. 3, T0
, Ts1, Ts2, ..., Tl, Tm, ..., T0 are the vectors (000), (010) ,.
1-1) (the vector shown in FIG. 2) is the output time in one control cycle.

Transmission order S22: Used when the command vector V is larger, the command vector V is generated in the transmission order S22 of the transmission method 2 as shown in FIG. In addition, as will be described later, when the output time of a short vector is adjusted to suppress the fluctuation of the neutral point potential, the transmission order S
22 may be used.

(B) Transmission method 3: Transmission sequence S31: When it is necessary to reduce the fluctuation of the neutral point potential, the transmission sequence S31 is used as shown in FIG. That is, the transmission order S31 occurs when it is necessary to reduce the fluctuation of the neutral-point potential in the transmission order S21 or S22. Specifically, the sixth means 6 described above is based on the current calculation result given from the fifth means 5 and, as will be described later, according to the condition of FIG.
Transmission method 2 or transmission method 3 is selected.

Transmission order S32: This is a transitional transmission order. In the order shown in FIG. 3, the transmission order S21 or S22 shifts to the transmission order S31.

By the way, if the total amount of charges flowing into and out of the neutral point during the control period Tc becomes zero, the fluctuation of the neutral point potential also becomes zero. This is the basic principle of the present invention, and according to this, the output time of the vector in each transmission order of each transmission method must satisfy the following equation. In the following equation, Ts1, Ts2, ..., Tl, Tm are vector output times, and iu, iv, iw are phase currents. <Transmission order S11> 2Ts1 × (iu + iw) + Ts2 × iw + 2Ts3 × (iu
+ Iv) + Ts4 × iv = 0 (2) <Calling sequence S21> Ts × (iu + iw) + Tm × iu = 0 (3) <Calling sequence S22> Ts × (iu + iw) + Tm × iu = 0 (4) <Calling Order S31> Ts xiv + Tm xiu = 0 (5)

Here, the condition for suppressing the fluctuation of the neutral point potential in each vector section is determined as follows. The transmission method 1 is selected when the modulation rate α <0.5, and if the output time of each vector is determined so as to satisfy the expression (2), the fluctuation of the neutral point potential can be suppressed at all.

The transmission method 2 is selected when the modulation rate α> 0.5. When the command vector V is in the section A1, similarly, the output time of each vector is expressed by equation (3) or equation (4).
If it is determined to satisfy the above condition, the fluctuation of the neutral point potential can be completely suppressed. The condition that can be suppressed is the following formula (6) as described later. iu / iv ≦ Tsn / Tmn (6)

The transmission method 3 is also selected when the modulation rate α> 0.5. When the command vector V is in the section A1, similarly, if the output time of each vector is determined so as to satisfy the equation (5), the fluctuation of the neutral point potential can be completely suppressed. The condition that can be suppressed is the following formula (7) as described later. iu / iv <-Tsn / Tmn (7) The conditions that can be suppressed in the entire 360 ° space are obtained as shown in FIG.

The sixth means 6 selects the transmission method and transmission order based on FIG. That is, the load current detected by the fourth means is given to the fifth means 5. On the other hand, in order to grasp the position (section number n) of the command vector V used by the fifth means 5, the second means calculates the section number n from the rotation angle ψ of the command vector V as follows. n = (ψ / 30 °) −2 (8) The fifth means 5 uses the load current instantaneous value detected by the fourth means 4 and corresponds by the section number calculated by the above second means. Calculate the current ratio. The sixth means selects the transmission method and transmission sequence as will be described later on the basis of this current ratio and the modulation rate α calculated by the third means.

When the transmission method and transmission sequence are selected as described above, the seventh means 7 calculates the output time of each vector in each transmission sequence. The method of calculating the output time of each vector will be described below. When the transmission method 1 is used, first, in order to generate the voltage command vector V, the output time T of the vector in the direction 1 in FIG.
Calculate the output time T2 of the 1 and direction 2 vectors. This is calculated by the following equations (9), (10) and (11) from the relationship between the sides and the corners of the triangle in the figure. (9)
In equations (10) and (11), Tc is the control period, Vs is the magnitude of the vectors Vs1, Vs2, Vs3, Vs4, and Vc is the magnitude of the command vector.

[0026]

[Equation 1]

On the other hand, in order to suppress the fluctuation of the neutral point potential, the output time of each vector in the transmission sequence S11 is expressed by the equation (2).
Must be satisfied. If the output time of each vector is determined as shown below from the equations (9), (10) and (2), the voltage command vector V can be generated and the fluctuation of the neutral point potential can be suppressed. Ts1 = T1 / 4 (12) Ts4 = T1 / 2 (13) Ts3 = T1 / 4 (14) Ts2 = T1 / 2 (15) That is, the above (12), (13), (14) and (15) are Substituting into equation 2), the left side of (2) is iu + iv +
It becomes iw, and as described above, since iu + iv + iw = 0, it can be seen that the equation (2) is satisfied.

When the transmission method 2 is used, first, in order to generate the voltage command vector V, the output time Tsn of the vector in the direction 1 and the output time Tmn of the vector in the direction 2 are obtained. This is calculated by the following equations (16), (17) and (18) from the relationship between the sides and the corners of the triangle in FIG.

[0029]

[Equation 2]

On the other hand, in the transmission method 2, the output time of the vector Vs (010) for suppressing the neutral point potential is Ts.
When the output time of z and the vector Vm (01-1) is expressed as Tmz, the output times Tsz and Tmz must satisfy the above expression (3), and the relationship between the expression (3) and iu + iv + iw = 0. From the following equation (19) is obtained. Tsz = (iu / iv) × Tmz (19) Here, since the vector Vm (01-1) has no choice, the output time Tm is set as in the following equation. Tm = Tmn = Tmz (20) The actual output time Ts of the vector Vs (001) and the output time Tl of the vector VL (-11-1) are determined by comparing Tsn and Tsz as follows.

In the case of Tsn> Tsz In this case, the fluctuation of the neutral point potential can be completely suppressed by outputting the long vector VL (-11-1).
The output time of each vector in the transmission order S22 is given by the following equation. Tm = Tmn = Tmz (21) Ts = Tsz = (iu / iv) * Tm (22) Tl = (Tsn-Ts) / 2 (23) T0 = Tc-Ts-Tm-Tl (24) That is, the transmission method In the case of 2, it is necessary to satisfy the relation of the above formula (19) in order to suppress the fluctuation of the neutral point potential,
Expression (23) is obtained from Expression (19) and Expression (20).
Further, since Tsn> Tsz = Ts, the output time of the vector VL irrelevant to the fluctuation of the neutral point potential can be calculated from Tl by the equation (24).

In the case of Tsn = Tsz The fluctuation of the neutral point potential can be completely suppressed. Moreover, the output of the long vector VL (-11-1) is not necessary. The output time of each vector in the transmission order S21 is given by the following equation. Tm = Tmn = Tmz (25) Ts = Tsz = Tsn (26) T0 = Tc-Ts-Tm (27)

In the case of Tsn <Tsz In this case, there is a portion in which the fluctuation of the neutral point potential cannot be suppressed. That is, even if the output time Tm of the vector Vm and the output time Ts of the vector Vs satisfying the above (22) and (23) are selected, Tsn <Tsz, and therefore the vector VL satisfying the expression (24) cannot be output. It is impossible to completely suppress the fluctuation of the neutral point potential.

The output time of each vector of the transmission method 3 can be calculated basically in the same manner as the calculation method of the transmission method 2. As described above, in the transmission method 2, the fluctuation of the neutral point potential can be suppressed by Tsn ≧ T in the section A1.
This is the case of sz. Since Tsz = (iu / iv) × Tmz as shown in the equation (19), the equation (19) and T
From sn ≧ Tsz, Tsn ≧ (iu / iv) × Tmz = (iu / i
v) x Tmn is obtained. That is, the variation of the neutral point potential can be completely suppressed by the transmission method 2 when the condition of Tsn / Tmn ≧ (iu / iv) is satisfied in the section A1, and the above equation (6) is obtained. be able to.

Further, with respect to the transmission method 3, if the fluctuation of the neutral point potential can be completely suppressed in the section A1 and the conditions are similarly obtained, the above equation (7) can be obtained. As described above, when the conditions for completely suppressing the fluctuation of the neutral point potential by the transmission method 2 and the transmission method 3 for each of the sections A1 to A12 are obtained, the above-mentioned FIG. 4 is obtained. As described above, the sixth means 6 refers to FIG. 4 and determines the transmission method 2 or the transmission method 3 based on the current ratio and the section number, and determines the transmission order.

[0036]

FIG. 5 is a diagram showing the configuration of a control device according to an embodiment of the present invention. In the figure, 10 and 11 are a DC power supply and a DC reactor, 12 is an input voltage dividing capacitor,
Reference numerals 13, 14 and 15 are the three-phase three-level inverter bridge shown in FIG. 7, 9 is a load motor, 16 is a load current detector, 17 is a digital signal processor (hereinafter abbreviated as DSP), and 18 is It is a microcomputer (hereinafter abbreviated as MC).

The DSP 17 is the first to third units shown in FIG.
Means for processing the output of the load current detector 16 by the method described above, and outputs the vector output method (transmission method) and the output time of each vector. Calculate and output to MC18. The MC 18 generates a PWM waveform according to the result processed by the DSP 17, and drives the switching elements of the three-phase three-level inverter bridges 13, 14, 15 by a gate circuit (not shown).

FIG. 6 shows the DSP 17 and MC1 shown in FIG.
9 is a flowchart showing a process in 8. First, D
The motor control unit of SP17 performs a calculation for controlling the motor using a known method to generate the voltage command vector V. Next, the position of the voltage command vector V, that is, the section number n in FIG. 8 is calculated by the above equation (8). Further, the modulation rate α corresponding to the command vector V is obtained from the control unit of the motor.

Next, in step S1 of FIG. 6, it is determined whether the modulation rate α is larger than 0.5. When the modulation rate α is smaller than 0.5, the procedure goes to step S4 and the transmission method 1 is selected. Then, in step S7, the vector output time is calculated according to the condition for generating the command waveform. That is, a vector to be output is determined by the calculated section number n and the modulation rate α, and the equations (9) to (1
The vector output times T1 and T2 in the section number n calculated by the method described in 1) are obtained. Further, in step S8, the output time of the vector is adjusted according to the condition of suppressing the fluctuation of the neutral point potential. That is, the specific output time Ts according to the above equations (12) to (15)
Calculate 1, Ts2, Ts3, Ts4. The zero vector V0
The output time T0 of (000) is determined by the difference between the control cycle Tc and the total of the output times Ts1, Ts2, Ts3, and Ts4.

If the modulation factor is greater than 0.5, the process proceeds to step S2, where the output of the load current detector 16 fetched by the DSP 17 causes the currents iu and iv of the respective phases.
, Iw current ratio. On the other hand, the above equation (16)-
The output time Tsn of the vector in the direction 1 and the output time Tmn of the vector in the direction 2 in the transmission methods 2 and 3 are obtained by the method described in (18). Then, referring to the table storing the usage conditions of the transmission methods 2 and 3 shown in FIG. 4, based on the section number of the command voltage vector V, Tsn / Tmn, and the current ratio of the currents iu, iv, and iw of each phase. , Determine whether to adopt the transmission method 2 or the transmission method 3.

Step S5 according to the determined transmission method
Alternatively, in step S6, the vector output time is calculated according to the condition for generating the command waveform in step S7. That is, a vector to be output is determined by the calculated section number n and the modulation rate α, and the equations (16) to (1
The output times Tsn and Tmn of the vector in the section number n are obtained by the method described in 8). Further, step S7
In, the vector output time is adjusted according to the condition for suppressing the fluctuation of the neutral point potential. That is, the concrete output times Ts, Tm, Tl, and T0 are obtained by the method described in the equations (21) to (27).

When the output time of each vector is determined as described above, the output time is set to MC1 in step S8.
8 to generate a PWM waveform and drive the switching elements of the three-phase three-level inverter bridges 13, 14, and 15.

[0043]

As described above, the following effects can be obtained in the present invention. (1) According to the present invention, the command vector V can be generated and the fluctuation of the neutral point potential can be greatly suppressed. Moreover, the method of suppressing the fluctuation of the neutral point potential does not affect the generation of the command voltage waveform. (2) The fluctuation of the neutral point potential can be suppressed by detecting the load current. That is, it is possible to suppress the fluctuation of the neutral point potential by using the current information detected originally for controlling the motor, and no additional equipment such as voltage detection is required. (3) Since the output of the inverter is controlled so that the sum of the inflow and outflow charge amounts to the neutral point becomes zero at an arbitrary moment, it can be applied even in a transient process such as starting or command change. (4) The suppression effect depends on the modulation factor and the load power factor, but is not so greatly affected. That is, the suppression effect according to the present invention is robust to the modulation factor and the load power factor. (5) Since the calculation and decision logic are simple, the processing time is short,
It has no effect on the time response characteristics of the entire system.

[Brief description of drawings]

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

FIG. 2 is a vector composition diagram in transmission method 1, transmission method 2, and transmission method 3.

FIG. 3 is a diagram showing a calling sequence in a calling method 1, a calling method 2, and a calling method 3.

FIG. 4 is a diagram showing usage conditions of calling method 2 and calling method 3;

FIG. 5 is a diagram showing a configuration of a control device according to an embodiment of the present invention.

FIG. 6 is a diagram showing a processing flowchart of an embodiment of the present invention.

FIG. 7 is a circuit configuration diagram of a three-level inverter.

FIG. 8 is a space vector diagram in a 3-level inverter.

FIG. 9 is a diagram showing the principle of suppressing the fluctuation of the neutral point potential in the carrier system.

FIG. 10 is a diagram showing a neutral point potential variation detection circuit in the space vector method.

[Explanation of symbols]

9 Motor 10,11 DC power supply and DC reactor 12 Input voltage dividing capacitor 13,14,15 Three-phase three-level inverter bridge 16 Load current detector 17 Digital signal processor (D
SP) 18 Microcomputer (MC) V command vector Tc control cycle V0 zero vector Vs short vector VL long vector Vm other vector Txn vector V for generating command vector
Output time of x Txz Output time of vector Vx for suppressing fluctuation of neutral point potential Tx Output time of actual vector Vx T0 Output time of zero vector

Claims (2)

[Claims] 1. A region sandwiched by one long vector in an output space voltage vector of a three-level inverter and a vector of an intermediate length adjacent thereto is defined as one section, and the entire 360 ° vector space is divided into 12 sections. The section number of the above-mentioned 12 sections of the command vector V is discriminated by the rotation angle of the command vector V, the modulation factor α is calculated according to the size of the command vector V, and the section number of the command vector V and the load current are calculated. The current ratio corresponding to the section number is calculated by the value, and the transmission method and the transmission order for suppressing the fluctuation of the neutral point potential of the voltage dividing capacitor of the 3-level inverter are determined by the modulation rate α and the current ratio. And a three-level inverter characterized in that the output time of each vector in the transmission order is calculated and the three-level inverter is PWM-controlled. PWM control method of the inverter. 2. A region sandwiched by one long vector in the output space voltage vector of the three-level inverter and a vector of an intermediate length adjacent thereto is defined as one section, and the entire 360 ° vector space is divided into 12 sections. First means for dividing, second means for determining the section number of the above-mentioned 12 sections of the command vector V based on the rotation angle of the command vector V given by the vector control device, and the size of the command vector V The third means for calculating the modulation rate α by the fourth means, the fourth means for detecting the load current, the section number of the command vector V given by the second means, and the load current value detected by the fourth means According to the fifth means for calculating the current ratio corresponding to the section number, and the modulation ratio α calculated by the third means and the current ratio calculated by the fifth means, three levels are obtained. Sixth means for determining the transmission method and transmission order for suppressing the fluctuation of the neutral point potential of the voltage dividing capacitor of the inverter, and the seventh means for calculating the output time of each specific vector in the transmission order determined by the means. Means for calculating the transmission method and the vector output time according to the load current based on the means of the first, second, third, fourth, fifth, sixth and seventh, and suppressing the fluctuation of the neutral point potential. The PWM control device of the characteristic 3 level inverter.