JPH0279789A - Pwm-signal generating method of inverter - Google Patents

Abstract

PURPOSE:To simplify a circuit by operating a time for controlling the gates of six power transistors in a three-phase inverter in an MPU based on a frequency command, a voltage command and functions incorporated in a ROM, and outputting the result. CONSTITUTION:Pulse generators 2ai (i=a-f) latch 3 a timer time instructed from an MPU 7 and latch 4 the sign of a voltage. The time of a free running timer 8 is compared with the time of the latch circuit 3 all the time. When both times agree, a latching signal is inputted into an FF 6, and output pulses are changed. The value of the time of the pulse change is operated in the MPU 7 based on a first frequency command W1*, a motor output voltage command V1* and functions incorporated in a ROM 10 for every output of an interruption timer 9. The pulse change times and the signs are sequentially computed, and gate signals + or -U to + or -W of six power transistors are outputted. In this way, a simple control circuit is formed at a low cost, and one chip is implemented.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a pulse width modulation (PWM) signal generation method for generating a gate signal of a three-phase inverter for variable speed driving of an AC motor.

[Conventional technology]

Conventionally, there has been a PWM signal generation method in which control is performed so that the locus of the line-to-line voltage synthesis backwheel draws a circle.

For example, I Peck Tokyo (1983) No. 4
07 to 417 (IPEC-Tokyo (1983
) discussed in PP407-417).

This involves storing a switching pattern in ROM that corresponds to the magnetic flux rotation phase and voltage command, and then extracting the switching pattern from the ROM based on these commands and switching the three-phase P.
Outputs WM signal.

In addition, there are two control voltage vectors that apply voltage to the motor windings and advance the magnetic flux, and whose switching states differ by only one.

If Vn is Vn and VZ is a zero voltage vector that does not apply voltage to the motor windings, Vi VA for one phase command.
I VB, VBe VZlVZ as one set, VZ+V
he VBe VB, VAI VBt VZ (1)
The voltage vectors are output in sequence. In addition, VA, VB
, VZ is set as one set, and the voltage vector I~ is changed every 30'.

[Problem to be solved by the invention]

Although the above-mentioned conventional technology can perform high-frequency switching by outputting a PWM signal using hardware, it is necessary to store all switching patterns with phase and voltage commands as parameters in the ROM, which makes it difficult to control the pulse width with precision. The ROM capacity becomes relatively large. Further, counters, latch circuits, etc. are also required, making the control circuit complicated and increasing the cost of the control circuit. Also, the zero voltage vector v
x has a Vo vector of (U, V, W) = (0, 0, 0) and a v7 vector of (1, 1, 1).

If Vz-+V^→Ve-*V^→VZ is output as one set, three phases change simultaneously when Vo and v7 are switched, resulting in an increase in switching loss. Furthermore, if all three phases are turned off at the same time, the current from the DC power supply is cut off in the voltage source inverter, so the inductance voltage due to the wiring from the power supply to the switching element will jump, potentially damaging the switching element.

Furthermore, if a switching change occurs in two of the three phases at the same time, a reverse polarity pulse is generated in the line voltage in the 180° section, causing motor current distortion. Therefore, it is ideal to have a switching pattern in which only one of the three phases changes over a 360° interval.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a PWM signal generation method for an inverter that can flow a highly accurate sinusoidal motor current using an inexpensive control circuit and can respond to instantaneous torque that can be applied to vector control.

[Means to solve the problem]

First, as a means to achieve the above purpose, ROM is used.

RAM, )ISO(lligh 5peed
17) PWM signals (U+, U-, V+, V-.

The system outputs W+, W-). Further, the switching pattern of the voltage vectors VA and VBVZ is changed every 60°. Note that the normal 60゛ interval is an arbitrary carrier period T depending on the sample value of the phase command O*.
For example, VZ→■＾→V Fl −) V n→■
^→Output VZ as a set of switching patterns,
Only immediately after switching the 60° section, between Tc/2 v8 → V^ → VZ
output as a set of switching patterns. Furthermore, as the zero voltage vector VZ, V
0 and V7 voltage vectors are applied alternately. Also, 1! TA is the flow time of pressure vectors VA and VBVZ.

When TB and 'rz are given, the phase command 0* in the 60° interval is given by equations (1) to (3).

Tz=Tc/2-TA-Ta-(3)
Here, Tc is a constant or arbitrary sampling period of the phase command, V1** is the inverter output voltage command, and Vmax is the maximum output voltage of the inverter.

[Effect]

Therefore, as will be explained in detail in the embodiment, in the case of a three-phase inverter, there are zero voltage vectors of V0 and V7 and control voltage vectors of Vl to VB. °≦f) ago<
60° section, VA = Vl (U, V, W = 0, 0, 1),
VB=VB (0,1,1) VZ=Vo (0,0,
O) vector, 60”≦(laeo<120’
The interval is V^=VB (1,0,1), Vo=Vx (
Op O+ 1), select VZ=V7 (11111). In addition, the normal O°≦Oago, <60° section is Vo (0,0,O) →V1* (0,0,1) →V
B (0,1,1)→VB (0,1,1)→V5(0
, 0, 1)-+vo(o, o, O) are arranged with the 'rc section as one set. As a result, the U phase does not switch in a 60" interval, and only the two phases, V and W, are switched. In this case, the switching transition occurs when only one phase changes at a time. Explain.O
The interval between °≦θδ6° and 60° is always Vo (0, 0, O
) vector interval, Vr (0,0,1) →Vr+(L
, 0.1) → V7 (1,1,1) arrangement, 60
Even when switching between ° sections, switching changes only occur in one phase. Note that 60°≦θseo<120”
The normal interval is Tc interval V7 (1, 1, 1) -+V
s(1, 0.

1) +Vt (0,0,1)-+Vt (0,0,1)
→VB(1,0,1)→V? Switching transitions are made with (1, 1, 1) as one set. Due to this operation, only one phase changes at the same time over the entire 360゛ range, so 1
No reverse polarity pulses occur at 80° line voltage.

Further, the flow times of V^, VB, VZ 1B, and pressure vectors are given by equations (1) to (3), and the locus of the line voltage composite vector operates in a circular manner.

〔Example〕

An embodiment of the present invention will be described below. The hardware configuration is shown in Figure 1. FIG. 1 shows a functional configuration diagram of Intel Corporation's 1-chip microcomputer 0196 as an example of the 1-chip microcomputer 1. As shown in FIG. The C196 microcontroller has a total of 9 timers (Cl
96 microcontroller has a function called H8D),
These timers can be used as pulse generators (2a-2f). The pulse generator 2a is a timer time latch circuit 3
, a code latch circuit 4, a comparator 5, and a flip-flop circuit 6, and receives pulse (PWM signal) change time tn- from an MPU (microprocessor unit) 7.
and the sign (rising edge = 1. falling edge = 0) in the latch circuit 3,
Set to 4. As a result, the time of the free-run timer 8 and the time setting value of the latch circuit 3 (n$) are constantly compared in the comparator 5, and when tn=tnI, a latch signal is input to the flip-flop circuit and the output pulse is output. Change. By sequentially calculating and setting the pulse change time un- and the sign in this way, six PWM gate signals U+, U-, and V+ are generated.

Outputs V-, W+, and W-. Next, the value of the pulse change time is determined by the phase command 0, which is obtained by integrating the first layer wave number command ω1 for each interrupt signal of a constant or arbitrary period given by the interrupt timer 9, or the motor output voltage command V1. *Season,
Function table (Sino
The MPU 7 calculates based on the umbrella and 5in (c/3-0 fist) values.

Next, the PWM signal generation method of the present invention will be explained in detail. Figure 2 shows the DC power supply 11 and power transistor tJ+
, U-, V+, V-, W+, and w-.

Therefore, in the case of a three-phase inverter, transistors U+, V
The on/off states of + and W+ are 2δ= as shown in Figure 3.
There are 8 ways. For example, V4(U+.

The voltage vector of V+, W+) = (1, 0, O) is.

U10 on (U-off), ■+ and W+ off (V-9W
- indicates the case of ON), the terminal of the U-phase winding 150 is connected to the positive side of the DC power supply 11, and the ■-phase and W-phase winding terminals are connected to the negative side. As a result, the line voltage composite vector is v4
The direction will be Note that the direction of the arrow indicates the positive side of the DC power supply. In this way, Vt~6 is the voltage vector applied to the motor winding, and the magnetic flux is the integral value of this voltage vector. On the other hand, Vo (0,0,O) and V? (1, 1, 1)
In this case, the motor winding is short-circuited and does not contribute to the generation of magnetic flux, resulting in a zero voltage vector VZ. Next, FIG. 4 shows the basic software processing of the microcomputer that is activated at regular or arbitrary intervals. First, the one-number frequency command ω111 and the inverter output voltage command vi* are input.

After this, an integral action is performed by integrating the ω1 fist, and the fifth
The phase command θ range is output for each 60' section shown in the figure.

Next, the conduction times TA, Ta, and Tz of the three voltage vectors ■^, VB, and V'l are calculated from equations (1), (2), and (3). Here, 0≦Os<so' in sin O.

5in (c/3-0 fist) is R as a function table
It is stored in OM9. Also, the carrier period Tc is 1
It is the period for outputting the set switching mode, and Vma
x is the maximum output voltage of the inverter, and V1* is the inverter output voltage command.

Next, check whether the phase command 0 is immediately after switching the 60° section.
- is judged based on whether or not it overflows, and the normal 60
In the case of the ° interval, (if not immediately after switching), the pulse generator 2 is used to generate the three voltage vectors VA shown in Table 1,
VBVZ to Tc section VZ-VA-VB VB-VA-
With VZ as one set of switching patterns, the respective voltage vectors are Tz, Tae TB, Ta, Ta,
Output Tg time. In addition, the 601 section number Ne in Table 1
o is + if the θ umbrella overflows as shown in Figure 5.
The three voltage vectors and conduction times selected for each 60° section shown in Table 1 will be explained with reference to FIG. 6. The direction of the v3 vector in Figure 3 is taken as the origin (θ5eo=o°).The two voltage vectors VA move in a circular locus in the direction of the 6 electric signs.
, VB can be controlled to cause a sinusoidal current to flow through the motor.

Therefore, if the axis of θ = 06 is bound to the B axis and the axis of 60° is AM, then for example, the section Neo:O becomes VBfd''Va, and VA'':V1. Also, the A-axis direction component of the VII command vector is vllll
・5ino city, and the B-axis direction component is V1*sin
(π/3-θ palm). As a result, by calculating the flow time of the VA and VB vectors from equations (1) and (2), the voltage command vector ■1* draws a circle, so that a sine wave current flows to the motor.

In addition, as can be seen from FIG. 3, the voltage vectors selected in Table 1 are V s sVB,
Let the Vo vector be the 60° section where N5o=1 is Vl,
By selecting Vll and V7, the voltage vectors are sequentially selected so as to rotate. Also, the section where Nso=0 is V
A: V 1 teV a = V a and na'J, N
[1O=1 (71 section is V^” V B and Va :'
V l. That is, the V^ vector is a voltage vector in the direction of 60'' section phase command θ umbrella, and the VB vector is a voltage vector in the direction of θ, = Q a.

Table 1 Next, the process immediately after the θ fist changes to the 60' section will be explained using flowchart 1- in FIG. First, the 60' section number Neo is +1 and the voltage vectors V^ and VoVZ corresponding to Neo are selected from Table 1. After this Tc/2
The switching patterns of the section VB→■^→VZ are set as one set, and each is T o.

Output T^, Tz time. A switching time chart at this time is shown in FIG. Note that FIG. 7 is a switching time chart when the phase command is 60° (Nso=O-) Neo=1). When 60° section number Neo=O,
Tc section V Z (V 13 ) → V A (V I
) 4 V u (V s )-)VB (VB)nV
^(Vt) → VZ (VB) is given as one set of switching patterns, switched to Neo=1, given a switching pattern of VZ (V7), and then the Tc interval, V Z
(V 7 ) −* VA (V 1%) →V
u(Vl) →vB (Vt)→V^(VB) -+v
g (V?) is given as a set of switching patterns.

By adopting such a switching pattern, only two of the three phases are switched as shown in Figure 7, and the number of pulse generators (timers) (number of pulse change time setting values) shown in Figure 1 is reduced. For example, by using Intel Corporation's 0196 microcontroller, which can set nine timer times, a three-phase ( U+, U+
.

v+, v-, w+, w-) gate signals can be output.

As a result, the control circuit has the advantage of being extremely inexpensive and simple in configuration. Note that 1 in FIG. 7 is the fall time of W-, and 11 is the rise time of W+ delayed by the dead time for preventing short circuit between the upper and lower arms of the inverter.

Furthermore, as shown in Fig. 7, even when switching between 60° intervals, only one of the three phases always changes;
There is also the effect that a sinusoidal motor current flows because no reverse positive or negative pulses occur in the line voltage in the ° section, and as a result, no current distortion occurs.

Next, in one embodiment, a unit such as a general-purpose inverter that integrates the primary station wave number command ω111 to generate the rotational phase command of the primary voltage.
The explanation has been given using an example of Vlf constant control in which the secondary voltage and primary frequency ratio are controlled to be constant. In such constant Vlf control, the magnitude of the primary voltage command vl- and the rotational phase 0 do not change stepwise. Therefore, another embodiment is shown in FIG. 8 which is applied to a general vector control device that changes Vl and 0* in steps according to the torque command. and the excitation current command knee are input, and the two-axis (d, q) axis components of the primary voltage Vla,
Vlq is calculated from equations (4) and (5).

vt4=ωt*(Qmu+fiz)4 is umbrella-rl・Ill
*...(4) Here, in the motor constant, rL is the water resistance Qt, Qz is the primary and secondary leakage inductance, and M is the mutual inductance. Next, the primary voltage vector calculation means 17 calculates the equation (6) to obtain the primary voltage magnitude command Vi- and the phase command (p*
Calculate.

V 1 ”=f−η]TJ...(6
)Vtd Therefore, the processing of the single-chip microcomputer 1 is the same as that of the previous embodiment, and the only difference is that the primary voltage phase command is calculated from equation (8).

0*=fω1◆・dt+ψshin・
(8) In other words, the process of adding . to the voltage phase command is added. In addition, in the case of vector control, V- and TI increase and decrease in steps. In particular, T increases and decreases in steps due to power and regeneration. Therefore, when θne increases, the voltage vector is arranged as shown in FIG. 7, and the O umbrella decreases. For example, when moving from the Neo=1 section to the Neo=O section, similarly, Between Tc VZ (Vt) -) VA (V
l5) +VB (Vt) -*VB(Vt)-+V^
(VB) →VZ (Vt> is placed, after section switching
Between T c / 2 VB (VB) → VA (Vt)
By arranging +VZ (Vo), only one phase changes at the same time.In this way, the present invention
Even in vector control where the fist changes in steps, it moves accurately. In addition, when changing the combination of three voltage speckles for each 30" section as in the conventional example, 12 section discrimination processes are required, but for each 60 degree section as in the present invention, only 6 section discrimination processes are required. The software processing time can also be made faster.As a result, the carrier period T', which is limited by the calculation time, can be reduced.
Another effect is that the carrier frequency can be increased because c can be shortened. Furthermore, the higher the frequency range is, the more the phase command θ umbrella between Tc changes. Further, there may be a case where the amount of change in OI exceeds the section phase, such as when the T umbrella increases or decreases significantly. In this case, the conventional method
Although it is limited by the change in I, in the present invention, the voltage back 1 - is switched every 60' interval, and the
Since it can be controlled as a simultaneous switching change of only the phases,
It is possible to provide an accurate primary voltage vector even in a high frequency region or vector control, and it is possible to flow an accurate motor current without current distortion.

Next, in one embodiment of the present invention, the normal 60' interval is set between Tc V z-+V A -* V a 4 V B-*
V A → V z 1? As explained in the example, the 9th
As shown in the figure, one normal 60° section is between Tc and VZ
(Vt) -+Vu (VB)-+VA(Vl) →V
＾(VZ) → VB (VB) → VZ (Vt) Tc, and only one interval VA (V 5) immediately after switching the 60° section
→V a (V l) →V z (V o ) Te device may also be used. In this case, Vo and vI in Table 1 may be exchanged. Conversely, as shown in Figure 10, 60°
-) VB (VB) is placed, and the normal 60° section is Tc
Between, VZ (Vt) + V^ (VB) nVR (Vl) →
VB(Vt)-+V^(VB)-+VZ(Vt) 1
'The arrangement is also good, and as shown in Figure 11, 60
° Immediately before switching the section, change the normal 60" section to Tcr!f
lVZ (Vo) −+VB(Vl)+V^(VB)+
Even if the arrangement is V^(VII)+VB(VZ)→VZ(Vo), a sine wave current without current distortion can similarly flow through the 360'' section chamber since only one phase changes at the same time.

〔Effect of the invention〕

According to the present invention, switching control is performed for only two of the three phases, and it is only necessary to set eight pulse change times between carrier periods, so that six gate signals can be output with only one chip microcomputer.

This has the effect of resulting in a low-cost control circuit. Furthermore, the switching pattern and its conduction time can be controlled so that the line voltage composite vector draws a circle, and the switching can be controlled so that only one of the three phases changes at the same time, resulting in highly accurate sinusoidal waveform It has the effect of becoming a motor current.

In addition, even if the phase of the primary voltage command changes stepwise, PWM control can be performed accurately and instantly so that only one phase changes at the same time, so it can also be applied to vector control that requires high torque response. It has the effect of saying.

[Brief explanation of the drawing]

Fig. 1 is a functional configuration diagram of a 1-chip microcomputer implementing an embodiment of the present invention, Fig. 2 is a 3-phase inverter configuration diagram, Fig. 3 is an explanatory diagram of voltage vectors, and Fig. 4 is a software processing diagram of the present invention. 5 is a time chart of phase command change, FIG. 6 is an explanatory diagram of the conduction ratio of two voltage vectors in the present invention, and FIG.
8 is a block diagram of another embodiment in which the present invention is applied to vector control, and FIGS. 9 to 11 are diagrams showing other embodiments of the present invention. 1 is a voltage vector arrangement diagram showing 1...1 chip microcomputer, 2... pulse generator, 3
... timer time latch circuit, 4 ... code latch circuit, 5 ... comparator, 6 ... flip-flop circuit,
7...MPU, 8...Free run timer, 9...
Interrupt timer, 10... Internal ROM, 11... DC power supply, 12... Power transistor, 13... Inverter, 14... Induction motor, 15... Motor winding, 16... Current/first factor/Figure 2 Figure 4 IF? , Q Figure 5 Figure 6 ↓ F3 detail Figure 7 Figure 8 W1゛

Claims (1)

[Claims] A PWM signal for an inverter is generated based on a rotational phase command θ* of the primary voltage, which is obtained by integrating the magnitude command V_1* of the primary voltage of the 1- and 3-phase inverter and the primary frequency command ω_1*. In the method of Select a zero voltage vector V_Z where the vector is zero, and apply these voltage vectors to V_Z, V_A, V_B, V_B in a regular 60° interval or an arbitrary carrier period Tc interval.
, V_A, and V_Z are arranged in a switching pattern as a set, and V_B, V_A, and V_Z are arranged in a switching pattern as a set in the Tc/2 section immediately after or immediately before the 60° section switching. PWM
Signal generation method. 2. In the method of generating the PWM signal of the inverter based on the magnitude command V_1* of the primary voltage of the three-phase inverter that changes instantaneously and the phase command θ*, such as vector control of an AC motor, a rotation phase of 60° is used. For each section, the inverter output voltage composite vector has an arbitrary size, and the two control voltage vectors V_A and V_B differ in only one switching state from each other, and the zero voltage at which the inverter output voltage composite vector becomes zero. Select the vector V_Z and apply these voltage vectors to V_Z, V_A, V_B, V_ in the normal 60° interval, which is constant, or in an arbitrary carrier period Tc interval.
An inverter characterized in that B, V_A, and V_Z are arranged in a switching pattern in which one set is arranged, and V_B, V_A, and V_Z are arranged in a switching pattern in one set in the Tc/2 section immediately after or immediately before the 60° section switching. PW of
M signal generation method. 3. Regarding the arrangement order of the voltage vectors V_A, V_B, and V_Z as described in claim 1 or 2, the normal 60° interval is V_Z, V_A, and V_Z in the carrier period Tc interval.
Arrange in the order of V_B, V_B, V_A, V_Z, 60°
Tc/2 section immediately after section switching, V_B, V_A, V_
Or, in the normal 60° section, arrange in the order of V_Z, V_B, V_A, V_A, V_B, V_Z between Tc, and immediately after switching the 60° section, arrange V_A between Tc/2,
V_B, V_Z are arranged in this order, or the normal 60° section is V_Z, V_A, V_B, V_B, V_A between Tc.
, V_Z, and Tc/ just before switching the 60° section.
2 sections, arranged in the order of V_Z, V_A, V_B, or a normal 60° section, V_Z, V_B, V_B between Tc.
Arrange A, V_A, V_B, and V_Z in the order of Tc/2 section immediately before the 60° section switching, V_Z, V_B, and V_A.
A PWM signal generation method for an inverter characterized by arranging the PWM signal in the following order. 4. Two types of zero voltage vectors V as ways of providing the zero voltage vector V_Z described in claim 1 or 2.
_0, V_7 are given alternately every 60° interval, and V_
PWM of an inverter characterized in that the transition of each switching pattern of Z, V_A, and V_B is given by a switching pattern in which only one phase out of three phases always changes.
Signal generation method. 5. Flow time T_A, T_B of voltage vectors V_A, V_B, V_Z according to claim 1 or 2
, T_Z is the phase command in the 60° interval θ*(0≦θ*<6
0°), the inverter maximum output voltage is Vmax, and the inverter output voltage command is V_1*, then T_A=(Tc
/2)×(V_1*/Vmax)・sinθ*T_B=
(Tc/2)×(V_1*/Vmax)・sin(π/
3-θ*) T_Z=Tc/2-T_A-T_B PWM of an inverter characterized by being given by the following relationship.
Signal generation method.