JP3302565B2 - PWM controller for voltage source inverter - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a PWM controller for a voltage-type inverter used as a control power supply for driving a motor in the field of electric railways and the like. PWM of voltage source inverter that can be limited to a range
It relates to a control device.

[0002]

2. Description of the Related Art FIG. 2 is a diagram showing a circuit configuration of a voltage source inverter. In FIG. 2, 8 is a DC voltage power supply, 9 is a DC reactor, 10 is a capacitor, and both ends P and N are connected to the positive terminal and the negative terminal of the DC voltage power supply 8, respectively. 11su1, 11su2, 12sv1, 12sv
2 and 13sw1 and 13sw2 are U, V, and W phase main switching element groups connected in series between the positive terminal P and the negative terminal N, respectively.
1Du2, 12Dv1, 12Dv2 and 13Dw1, 1
An anti-parallel diode indicated by 3Dw2 is provided.

The potentials at the U, V and W phase terminals output two types of voltage levels + E and 0. A complex coordinate system is defined on the stator of the induction machine as shown in FIG. 3A, and a voltage vector is defined according to the following equation (1). In the following equation (1), Vu, Vv, and Vw are voltages at the U, V, and W phase terminals.

[0004]

(Equation 1)

Therefore, in the voltage source inverter shown in FIG. 2, eight kinds of discrete voltage vectors can be output in 2 ³ = 8 switching states. FIG.
(A) is a diagram showing the eight types of voltage vectors, and vectors V0 (000) and V7 (111) are zero vectors. 1 is the voltage level + E, 0 is the voltage level 0
For example, in the vector V1 (110), the U phase and the V phase indicate the voltage level + E, and the W phase indicates the voltage level 0.

In order to control the voltage source inverter shown in FIG. 2 with the above-mentioned voltage vector, several vectors are selected from the vectors adjacent to the command voltage vector V so that the combined vector coincides with the command vector V. To control. Here, if the resistance on the primary side of the motor is ignored, the magnetic flux vector λ is defined by the following equation (2).

[0007]

(Equation 2)

FIG. 3B shows an example of a magnetic flux vector. The circumference is a magnetic flux vector locus λ1 corresponding to the case where the input voltage of the motor is a sine wave, and is called an ideal magnetic flux vector. λ1 has a radius of | V | as shown in FIG.
/ Ω1. Here, | V | is the magnitude of the command vector V, and ω1 is the output angular frequency of the inverter.

FIG. 4 shows a part of FIG. 3B. The broken line λ in FIG. 4 is output by a software PWM control method having a voltage vector output arrangement equivalent to a conventional triangular wave carrier PWM control method. Represents the magnetic flux vector,
The voltage vector is output so that the magnetic flux vector locus is an approximate circular locus having a radius | V | / ω1. Zero vectors (V0, V7) are arranged on the circular locus, and the number of arranged zero vectors is determined by the relationship between the output frequency of the inverter and the carrier frequency.

In the conventional PWM control method, the magnetic flux vector section corresponding to the half cycle Tc of the triangular wave carrier is shown in FIG.
Can be represented by the arc AB as shown in FIG. The figure shows a case where a region of 60 ° (∠AOG) sandwiched between two adjacent voltage vectors is divided by an odd number N = 3, and the relationship between the output frequency f1 of the inverter and the carrier frequency fs is represented by m = f
When s / f1 (hereinafter, m is referred to as a pulse mode), a case where m = 9 is shown.

In the one divided section shown in FIG. 4, the following switching pattern is output in a section from the zero vector V0 to the next zero vector V7.

For one divided section (arc AB), the length of each vector, that is, the time T1 and T6 for outputting each vector V1 and V6 is given by the following equation based on the relationship between the sides and angles of the triangle in FIG. It is obtained by the equations (3) and (4). Equation (3),
In (4), | V | represents the magnitude of the command voltage vector V, and | V1 | and | V6 | represent the magnitude of the output voltage vector of the inverter, respectively. Times T1 and T6 at which the vectors are output vary depending on the magnitude of the voltage command vector V. In the case of FIG. 4, θ is related to the angle at which the magnetic flux vector rotates in the circumferential direction, and can be expressed by the following equation (5).

The remaining time is the output time T0, T7 of the zero vectors V0 and V7, and if these times are equally divided, it is calculated by the following equation (6).

[0014]

(Equation 3)

By outputting each voltage vector at the time calculated as described above, it is possible to realize a PWM in which the magnetic flux vector is approximated as a circular locus.

As a conventional technique in the application of electric railways, in PWM control of an inverter in a high-speed region, the inverter is operated in a one-pulse mode in order to increase the voltage utilization rate of the inverter. In the inverter PWM control in the medium speed range, GT
In order to lower the switching frequency of O, a synchronous modulation method is used. Assuming that the carrier frequency is fs and the pulse mode is m, fs is expressed by the following equation (7).

[0017]

(Equation 4)

In equation (7), f1 is the inverter output fundamental frequency. Since there is a limit to increasing fs due to an increase in switching performance and circuit loss of a semiconductor element, it is necessary to reduce m as f1 increases.
Further, m cannot take an arbitrary value, and it is necessary to switch m at an odd multiple of 3 in order to balance three-phase voltages and not generate even harmonic components.
That is, there are the following restrictions on the pulse mode m. Sine wave modulation: m = 1, 3, 9, 15, 21, ...

[0019]

In the conventional triangular wave carrier PWM control method, when the pulse mode m is switched, the torque ripple may become extremely large depending on the pulse mode m as shown in FIG. FIG. 5 (a)
(B) and (c) show the current waveform and the torque waveform in the 27-pulse mode, the 5-pulse mode, and the 1-pulse mode, respectively. The horizontal axis represents time, and the vertical axis represents current and torque. FIG. 3A shows a voltage of 215 V, a current of 230 A,
Torque: 212 kg · m, frequency: 8 Hz, slip: 28.5
%, Pulse mode 27, (b) voltage 855 V, current 1
70A, torque 183kgm, frequency 34Hz, slip 4.6%, pulse mode 5, (c) voltage 1100V,
Current 148A, torque 161kgm, frequency 48H
z, slip 3.1%, pulse mode 1 are shown.

According to FIG. 5, the torque ripple corresponding to the one-pulse mode in FIG. 5C is 19%, whereas the torque ripple in the five-pulse mode in FIG. The torque ripple corresponding to the pulse mode is 50% and 29%, respectively. That is, the torque ripple may be large depending on the pulse mode m, which causes a problem such as noise and a decrease in adhesion performance.

The present invention has been made to solve the above-mentioned problems of the prior art, and determines the pattern of the pulse mode m according to the modulation rate or frequency of the inverter, so that the torque ripple can be reduced to a certain value or less. An object of the present invention is to provide a voltage-type inverter control device which can be limited and can improve noise and adhesion performance.

[0022]

First, the principle of the present invention will be described. It should be noted that the present invention uses the IGB as shown in FIG.
It is intended for an inverter constituted by a high-frequency switching element such as T. That is, it is assumed that the carrier frequency can be made sufficiently high. The position of the magnetic flux vector λ, that is, the magnetic flux vector section number n where the magnetic flux vector λ exists (FIG. 3)
(B)) shows the definition of the magnetic flux vector λ and the command vector V
Can be determined by the following equation (8) based on the rotation angle ΔV. The modulation factor a of the inverter can be calculated by the following equation (9). In the inverter of V / f1 constant control, there is a relationship as shown in Expression (10) between the voltage command V and the frequency f1 of the inverter.
The frequency f1 of the inverter is obtained from Expression (10).
However, in the equation (10), C is a constant.

[0023]

(Equation 5)

As shown in FIG. 6, the actual magnetic flux vector trajectory λ rotates at a different rotation speed from the magnetic flux vector λ1 in the ideal state, so that a vector deviation occurs. This deviation is divided into a deviation Δθ in the rotation direction and a deviation ΔR in the radial direction as shown in FIG. 6, and the deviation Δθ in the rotation direction is calculated from the deviation ΔR in the radial direction.
Is considered to have a greater effect on torque ripple. If the radial deviation ΔR of the magnetic flux vector is ignored, the torque ripple can be expressed by the following equation (11).

[0025]

(Equation 6)

In equation (11), ΔT is the amount of change in torque, T is the average torque, and Kθ is a parameter determined by the constant of the motor. From Expression (11), the maximum value of torque ripple ΔTmax / T can be expressed by the maximum value Δθmax of the deviation of the magnetic flux vector in the rotation direction as in Expression (12).

In FIG. 4, when the region of 60 ° (∠AOG) sandwiched between two adjacent voltage vectors is divided by an odd number N as described above, the relationship as expressed by the equation (13) is obtained with the pulse mode m. is there. FIG. 4 shows that N =
3, m = 9, that is, 9 pulse mode.

[0028]

(Equation 7)

Since it is considered that Δθmax occurs when a zero vector is output, it can be expressed by the following equation (14). In Equation (14), a is the modulation factor of the inverter, and m is the pulse mode. FIG. 7 shows a value of Δθmax when the inverter frequency f1 is changed from 0 to 50 Hz in the inverter of the V / f1 constant control, which is calculated by using the pulse mode m as a parameter using Expression (14).
Here, the horizontal axis in FIG. 7 is the voltage modulation rate a of the inverter and the frequency f1 of the inverter.

[0030]

(Equation 8)

In FIG. 7, a horizontal line L represents the value of Δθmax corresponding to the torque ripple in the one-pulse mode, and represents a target value for limiting the value of the torque ripple in all the pulse modes m to within this range. . The abscissa of the intersection of the horizontal line L and the branch of each pulse mode m indicates the frequency f1 of the inverter to switch the pulse mode m and the modulation factor a at that time. If the pulse mode m is successively changed at the intersection of the branch of the pulse mode m and the straight line L in accordance with the rise or fall of the frequency f1 of the inverter, a PWM waveform is generated so as not to exceed the torque ripple of the one-pulse mode. can do. The thick line in FIG. 8A illustrates the above procedure.

In the figure, the dotted line indicates the torque ripple (に よ る Tmax) according to the conventional method of determining the pulse mode m.
/ T), the pulse mode m, and the frequency f1 of the inverter or the modulation factor a. FIG. 8B shows FIG.
FIG. 6 is a diagram illustrating characteristics of a carrier frequency fs corresponding to FIG.

When the pulse mode m is switched, it is necessary to minimize fluctuations in the inverter output voltage. Since the method of switching the pulse mode m is well known, the description is omitted here. The length of each voltage vector output at this time, that is, the output time of each vector, is calculated by Expressions (3), (4), (5), and (6) as in the conventional case. In this way, the torque ripple of the motor is limited to the range of the torque ripple corresponding to the one-pulse mode.
A WM waveform can be generated.

FIG. 1 is a functional block diagram of the present invention. The present invention solves the above problem as follows. In FIG. 1, a voltage command vector V is input from torque vector control. That is, as shown in FIG. 3 (b), for each section divided into A1 to A6, the torque vector control means outputs the output current of the voltage source inverter 6 detected by the current detection and the load detected by the speed detection. The command vector V in each of the above sections is obtained based on the rotation speed of the motor 7. Since the method of calculating the command vector V is well known, the description is omitted here.

The means 1 determines the position of the command vector V according to the rotation angle ΔV of the given voltage command vector V, ie, FIG.
It is determined which of the sections A1 to A6 in (b) is in the above equation (8), and the output order of the inverter voltage vector for generating the voltage command vector V is determined. The means 2 calculates the modulation factor a of the inverter and the frequency f1 of the inverter by the equations (9) and (10) using the given voltage command vector V.

Means 3 is a storage means, and has a pulse mode m
The modulation factor a or the frequency f1 of the inverter for switching is stored. That is, as described with reference to FIGS. 4 to 8, the torque ripple in a plurality of pulse modes m assumed based on the deviation between the rotation direction component of the ideal magnetic flux vector trajectory λ1 (circumference) and the actual magnetic flux vector λ (circular line). Is calculated, and from the desired torque ripple limit value and the calculation result of the torque ripple of each pulse mode m, the modulation factor a or frequency f1 of the inverter for switching the pulse mode m offline is obtained in advance, and the result is stored in the storage means 3.

Means 4 is based on the modulation factor a or the frequency f1 of the inverter calculated by the second means.
The pulse mode m is determined with reference to the data stored in. The means 5 calculates the output time of the voltage vector in each pulse mode m determined by the fourth means by using equations (3), (4),
It is calculated by (5) and (6). In accordance with the output time of the voltage vector obtained by the above-described means, the voltage is supplied to the voltage-type inverter 6 in a vector transmission order determined by the position of the command vector V, the voltage-type inverter 6 is controlled, and the load motor 7 is driven.

According to the present invention, as described above, the modulation factor a or the frequency f1 of the inverter in which the torque ripple is within the desired limit value is obtained and stored, and the pulse mode m is determined by referring to the stored data. Since the output time of each voltage vector is calculated, the motor can be operated with the torque ripple of the motor smaller than the torque ripple corresponding to the one-pulse mode in the motor operation region, so that the electromagnetic noise can be reduced and the adhesion characteristics can be improved.

[0039]

FIG. 9 is a diagram showing a configuration of a control device according to an embodiment of the present invention. 2, 8 and 9 are the DC power supply and the DC reactor shown in FIG. 2, 10 is an input capacitor, 11, 12, and 13 are the three-phase bridge inverters shown in FIG. 2, and 7 is a load motor. . 14 is a load current detector and 15 is a motor speed detector.

Reference numeral 16 denotes a digital signal processor (hereinafter abbreviated as DSP), and reference numeral 17 denotes a microcomputer (hereinafter abbreviated as MC) that performs processing according to the present invention.

In FIG. 9, a DSP 16 is a voltage command vector for torque vector control based on the operation command of the motor 7, the load current detected by the load current detector 14, and the rotation speed of the motor detected by the motor speed detector 15. Generate V. Since the process of generating the voltage command vector V is well known, a description thereof will be omitted (for example, if necessary, refer to Japanese Patent Application Laid-Open No.
No. 5588, Japanese Patent Application No. 7-147461, etc.).

The MC 17 is provided in advance in FIG.
The pattern of the pulse mode m shown in (a) is obtained, and the result is stored in a memory (not shown).

The DSP 16 has a DP-RA (not shown).
The voltage command vector V is transmitted to the MC 17 via M. MC 17 receives the voltage command vector V,
As described above, the position of the command vector V is determined, and the modulation factor a and the frequency f1 of the inverter are obtained. The pulse mode m is determined from the pattern of the pulse mode m stored in the memory according to the frequency f1 of the inverter, and finally, the output time of each vector is calculated, and the voltage source inverters 11, 12,. 13 switching elements are driven.

FIG. 10 is a flowchart showing the processing in the MC 17 shown in FIG. 9, and the processing in the MC 17 will be described with reference to FIG. MC17, as described above,
The torque ripple in a plurality of pulse modes m is obtained by the above equation (14), and from the desired torque ripple limit value and the calculation result of the torque ripple of each pulse mode m, the modulation factor a or frequency f1 of the inverter that switches the pulse mode m offline is determined in advance. And the result is stored in a memory (not shown) (steps S1 and S2).

In step S3, when receiving the voltage command vector from the DSP 16, the MC 17 determines the section of the magnetic flux vector in which the command vector V exists based on the rotation angle ΔV of the command vector V according to the equation (8). In step S4), the modulation factor a and frequency f1 of the inverter are calculated based on the length | V | of the command vector according to the equations (9) and (10).

In step S6, the above step S5
In accordance with the frequency f1 of the inverter obtained in
In step S7, the pulse mode m for keeping the torque ripple within a predetermined range is determined with reference to the data stored in the memory. In step S7, the output time of each vector is calculated by the above equations (3) to (6). According to the output time of each vector obtained as described above, the voltage type inverters 11, 1
Two or thirteen switching elements are driven.

[0047]

As described above, according to the present invention, when the induction motor is driven by the magnetic flux vector PWM control, the modulation factor a of the inverter in which the torque ripple falls within a desired limit value.
Alternatively, since the frequency f1 is obtained and stored, and the pulse mode m is determined with reference to the stored data to calculate the output time of each voltage vector, the operation can be performed with the torque ripple limited to a certain value or less, It can reduce noise and improve adhesive performance.

[Brief description of the drawings]

FIG. 1 is a functional block diagram of the present invention.

FIG. 2 is a diagram showing a circuit configuration of an IGBT voltage inverter to which the present invention is applied.

FIG. 3 is a diagram showing a space voltage vector and a magnetic flux vector.

FIG. 4 is a diagram illustrating magnetic flux vector PWM control (when N = 3 and m = 9);

FIG. 5 is a diagram showing an example of pulse mode m and torque ripple in conventional PWM control.

FIG. 6 is a diagram showing a deviation ΔR in the radial direction and a deviation Δθ in the rotation direction in the magnetic flux vector PWM method.

FIG. 7 shows the maximum value (Δθ) of the deviation of the magnetic flux vector in the rotational direction.
max) and the maximum value of the torque ripple (ΔTmax /
T) is a diagram illustrating a relationship between a pulse mode m, a modulation factor a, and a frequency f1.

FIG. 8 is a diagram showing a pulse mode pattern in which torque ripple is limited and carrier frequency characteristics corresponding to the pulse mode pattern.

FIG. 9 is a diagram illustrating a configuration of a control device according to an embodiment of the present invention.

FIG. 10 is a diagram showing a processing flow of an embodiment of the present invention.

[Explanation of symbols]

1 means for judging the section where command V exists 2 means for calculating output frequency f1 or modulation rate 3 means for storing modulation rate a or frequency f1 4 means for determining pulse mode 5 means for calculating output time of each vector Reference Signs List 6 voltage type inverter 7 load motor 8 DC voltage power supply 9 DC reactor 10 capacitor 11, 12, 13 main switching element group 14 load current detector 15 motor speed detector 16 digital signal processor (DSP) 17 microcomputer (MC) V command Vector Vx Inverter output voltage vector (x = 0 to 7) λ Actual magnetic flux vector λ1 Magnetic flux vector in ideal state | V | Vector length Tc Control cycle, time corresponding to one division of magnetic flux vector Tx vector Vx Output time n Magnetic flux vector section number divided Number (n = 1 ~
6) ω1 Inverter output angular frequency f1 Inverter output frequency fs Carrier frequency a Inverter modulation rate m Pulse mode N 60 ° Number of divided magnetic flux sections T Average torque ΔT Torque variation Δθ Flux vector rotation direction deviation ΔR Radial deviation of magnetic flux vector Δθmax Radial deviation of magnetic flux vector corresponding to maximum value of torque ripple Kθ Parameter related to motor constant

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-4-79770 (JP, A) JP-A-63-302766 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H02P 5/408-5/412 H02P 7/628-7/632 H02P 21/00 H02M 7/42-7/98 B60L 1/00-3/12 B60L 7/00-13/00 B60L 15/00-15 / 42

Claims (1)

(57) [Claims] In an induction motor drive system using a voltage source inverter, a voltage vector (V0 to V7) defined on an induction machine stator coordinate and an air gap magnetic flux vector of the induction machine obtained by integrating the voltage vector. λ and the voltage vector (V1 to V6) sandwiched by a straight line that intersects at the origin of the orthogonal coordinate system as one section, and the entire 360 ° space of the magnetic flux vector is divided into six sections from section A1 to section A6. First means for determining the position of the command vector V, that is, where in the divided section numbers (A1 to A6), based on the rotation angle ΔV of the given voltage command vector V, Second means for calculating the modulation factor a or the frequency f1 of the inverter based on the magnitude | V | of the voltage command vector V, the ideal state magnetic flux vector locus λ1 and the actual magnetic flux vector
The torque ripple obtained from the deviation of the rotational component of λ is
Switch the pulse mode m below the desired limit
The modulation factor a or frequency f1 of the barter is obtained, and the pulse
Modulation rate a or frequency of inverter for switching mode m
According to the third means storing the wave number f1 and the modulation factor a or the frequency f1 of the inverter calculated by the second means, the pulse mode m is determined by referring to the data stored in the third means. And at least fifth means for calculating the output time of the voltage vector in each pulse mode m determined by the fourth means, wherein the first, second, third, fourth and fifth means are provided. A PWM control apparatus for a voltage-type inverter, wherein an output voltage pulse mode m of the inverter is determined based on the output voltage and an output time of each voltage vector is calculated.