JPH01303093A - Controller for pulse width modulation of inverter - Google Patents

Abstract

PURPOSE:To reduce a current ripple, a voltage ripple and a torque ripple effectively by changing the number of division of ON time of a transistor with the value of the torque ripple. CONSTITUTION:The torque ripple of an induction motor 1 is detected by a torque ripple detector 13. A division-number setting circuit 14 is built into a micro-computer 8, and has a division-number table in which the value of the number of division N of ON time of a dividing circuit 11 is stored according to the value of the torque ripple previously generated in the induction motor 1. The number of division N of the division-number setting circuit 14 is set and stored so as to increase by stages with the elevation of the torque ripple. The value of the number of division N corresponding to the magnitude of the torque ripple detected by the ripple detector 13 is read and set by the division- number setting circuit 14, output to the dividing circuit 11, and used for dividing the ON time.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention relates to a pulse width modulation control device for an inverter, and particularly relates to an improvement in a device that increases the carrier frequency and performs precise waveform control.

(Prior art) In recent years, MOSFETs have been used as high-speed switching devices.
Elements such as (metal oxide gate field-effect transistors) have appeared, and if they are used in inverter pulse width modulation, precise waveform control becomes possible, reducing electromagnetic noise and increasing motor efficiency. It has become possible to obtain such effects.

Therefore, in the past, high carrier frequencies (for example, 20Klz
) is known to enable pulse width modulation control using the method (2), thereby ensuring the above-mentioned effect of reducing electromagnetic noise, etc. (For example, see ``Application of high-frequency switching to general-purpose inverters'' in the proceedings of the 1986 National Conference of the Industrial Application Division of the Institute of Electrical Engineers of Japan, presented by Chihiro Okadochi, etc.).

(Problems to be Solved by the Invention) However, the conventional device described in item 1 had drawbacks such as a complicated circuit, complicated various adjustments, and high cost.

Therefore, is it possible to consider using a one-chip microcomputer (hereinafter referred to as a microcomputer), which is inexpensive and has a simple circuit configuration?In this idea, a microcomputer is used for a series of processes necessary to generate a PWM control pattern. It takes a long calculation time, for example, about 200 μs, and the carrier frequency remains at about 5 Kllz at maximum. For this reason, it is generally difficult to control the pulse width change by 1 SO using a carrier frequency of high frequency (20 KII z or more).

The present invention has been made in view of the above, and its purpose is to provide a low-cost and simple device while employing a chip microcomputer, by creating a situation equivalent to increasing the carrier frequency. The purpose of this invention is to enable pulse width modulation control at an equivalently high carrier frequency with a simple circuit configuration, thereby achieving effects such as reduction of electromagnetic noise and increase in motor efficiency by one point due to precise waveform control.

(Means for Solving the Problems) In order to achieve the above object, the present invention changes the generation algorithm of the PWM control pattern (that is, the ON time of a plurality of switching elements provided in the inverter), and
Even if the calculation time (calculation cycle) of the M control pattern is long, the calculated ON time of each switching element is divided into a plurality of pulses, and the carrier frequency is equivalently raised by 1.

In that case, when controlling the drive of a motor using such pulse width modulation control, there is generally torque ripple in the motor, so in order to effectively reduce this, it is necessary to increase the carrier frequency at the point when the torque ripple increases. In order to increase this, it is preferable to increase the number of divisions of the ON time of the switching element. Furthermore, when the ON time of each switching element is equally divided into equal width pulses, as shown in FIG.
If there is a large change in the ON time, it will not be able to respond well to this, so it is preferable to divide the pulse into unequal width pulses as shown in FIG. 15. In this case, if the number of divisions is increased, as shown in FIG. As shown in Figure 2, the reproducibility of the signal wave is further improved.

Therefore, in the present invention, by fixing the number of divisions of the ON time of the switching element and changing it appropriately, the reproducibility of the signal wave is improved. The purpose is to perform modulation control.

In the invention according to claim (1), a specific means for solving the problem is as shown in FIGS. 1 and 2, the switching elements (Tra) to (Trc '
), each of the switching elements (Tra) to (Trc゛) of the bridge circuit (4)
The DC pulse width is modulated by the 0N10FF operation of 1.
The present invention is based on a pulse width modulation control device for an inverter that applies a three-phase AC voltage to the three-phase winding (2). Then, as shown in Fig. 6, Fig. 7, Fig. 13, and Fig. 15, the ON time of each switching element (Tra) to (T+/c') is calculated in the calculation period according to the carrier frequency. and a calculation means (10) that calculates the ON time of each switching element (Tra) to (Trc') calculated by the calculation means (10) according to the ripple based on the current of the three-phase winding (2). a dividing means (11) for dividing into a plurality of pulses, and a plurality of pulses divided by the dividing means (11) to each of the switching elements (Tra) to (1゛rc'
) is provided (19.).
It was completed.

Furthermore, in claim (2) of the present application, the above dividing means (1)
Instead of 1), the ON time of each switching element (1'ra) - (Trc') calculated by the calculation means (10) is divided into a plurality of pulses, the number of which corresponds to the rate of change of the ON time. This is a dividing means (11°).

(Operation) With the above configuration, claim (1) and claim (
In the invention described in 2), each switching element (
The ON time (PWM control pattern) of Tra)-(Trc') is repeatedly calculated by the calculation means (10) at each calculation period according to the carrier frequency, and the ON time (PWM control pattern) of each switching element (Tra)-(Trc' ) is divided into multiple (for example, 4) pulses by the dividing means (1]-) and (1, 1°), so the carrier frequency is multiplied by the number of divisions, equivalently The same situation occurs when pulse width modulation control is performed at a high carrier frequency (for example, about 20 KIlz). As a result, each of the switching elements (Tra) to (Trc'
) or ON control by the control means (2), a precise Senwauchi waveform close to a sine wave is obtained, which effectively reduces electromagnetic noise and effectively increases motor efficiency by 1. Become.

Here, the carrier frequency of pulse width modulation control is the normal value (
(approximately 5KHz) and takes a long calculation time] Even a chip microcontroller can sufficiently calculate the PWM control pattern.
Pulse width modulation control using a high carrier frequency can be performed at low cost and with a simple circuit configuration.

Furthermore, the switching element (T
The ON time division of ra)-(Trc') is determined by the three-phase winding (
2) It is divided into multiple pulses according to the ripple based on the current, such as torque ripple, current ripple, etc. In a situation where this ripple is large, the number of divisions can be larger than normal, so this large ripple If the carrier frequency in the situation is high enough, this ripple is effectively suppressed.

Further, in the invention according to claim (2), the number of divisions is changed according to the rate of change in the ON time of each switching element (Tra) to (Trc'), and when the rate of change in the ON time is large, the number of divisions is changed. Since the number of signals can be increased, good reproducibility of the signal wave can be ensured. Therefore, it is possible to equivalently perform pulse width modulation control using a sufficiently high carrier frequency without attempting to improve the reproducibility of the signal waveform.

(Example) Embodiments of the present invention will be described below with reference to the drawings.

1 and 2 show a pulse width modulation (hereinafter abbreviated as PWM) control device for an inverter according to the present invention. In each figure, (1) has three windings (2a), (2b) 1-
(2c) is an induction motor having a three-phase winding (2) with Y-connection; (3) is a voltage type inverter connected to the induction motor (1); A transistor bridge circuit (4) connected to the three-phase winding (2) of the induction motor (1) is provided, and the bridge circuit (4) includes free-wheeling diodes (Da) to '(Dc), respectively.
') transistors (switching elements) such as MOSFETs (Tra), (Tra
'), (Trb), (Trb'), (Trc),
(Trc'). Therefore, the inverter (3)
is equipped with a rectifier (6) that rectifies the three-phase alternating current of the three-phase power supply (5).
) is applying DC voltage.

Further, (8) is a one-chip microcomputer that controls the ON time of the six transistors (Tra) to (Trc') of the bridge circuit (4), that is, pwM control. 8) includes each of the above transistors (T
It is equipped with a pace driver (8a) that operates 0N10FF of ra) to (Trc'), and the microcomputer (8a)
)Nor transistor (Tra) - (Trc') 0
DC is pulse width modulated by N10FF control.

1. Inside the microcomputer (8), the ON time (total P) of each transistor (Tra) to (Trc') is stored.
The ON time calculation circuit (10) as a calculation means for calculating the WM control pattern) and the ON time calculation circuit (1o) that calculates the ON time of each transistor (Tra) to (Trc') calculated by the ON time calculation circuit (1o) a dividing circuit (11) serving as a dividing means for dividing the circuit into two; and a control means for controlling each transistor (Tra) to (Trc') to turn on in the pace driver (8a) using the pulses divided by the dividing circuit ('']). It also has a built-in pulse width control circuit (12).

In addition, in Fig. 1, (13> is "1 induction motor (
”) ripple detection circuit that detects torque ripple of (
14) is built in the microcomputer (8), and the value of the division number N of the ON time of the division circuit (11) is determined in advance according to the value of the torque ripple generated in the induction motor (]) = ]
- A division number setting circuit having a division number table stored in the division number setting circuit (4), wherein the division number N of the division number setting circuit (4) is set and stored so as to increase stepwise as the torque ripple becomes larger. ing. And the above ripple detection circuit (]
The division number setting circuit (14) reads and sets the value of the division number N corresponding to the magnitude of the torque ripple detected in step 3), and outputs it to the division circuit (11) for use in dividing the ON time. It is composed of

Next, the formation of the PWM control pattern by the microcomputer (8) in item 1 will be explained.

In general, this PWM control pattern is formed by determining a PWM control pattern so that the locus of time integration of the output voltage approaches a circular locus.

To explain this in detail, first, let the potentials of the output terminals of the inverter (3) be va, vb, vc, the potential of the neutral point of the three-phase winding (2) be VQ, and the output defined by the following formula. Consider a voltage vector vP and a time integral P of the voltage vector vP.

Vp = J'n d2. V
+ is the effective value of the fundamental wave voltage) When the voltage vector Yp and its time integral 2P are applied, the voltage vector Yp and its time integral 2P draw a circular locus on the complex plane.

On the other hand, in the voltage source inverter (3), one of the transistors in each phase arm is always in the ON state, so for convenience, the ON state on the -1+ side is indicated by "]", and the ON state on the - side is indicated by "]".
0'', and rlolJ, r in the order of C phase, b phase, and C phase.
When written as ``O]J'', the state of the inverter (3) is 8.
Exists as expected. Voltage vector Vp (P=0
~7) has a magnitude of J sword Vd (Vd is the DC voltage of the rectifier (6)), and its direction is the direction shown in FIG. Here, V O% V 7 is a zero vector since Vo=Vy=0.

The time integral 4p of the above voltage vector is c31p/dt-v
Since P, the time integral λ when driving the inverter <3)
P is 1Vpl-JT-v in the direction of voltage vector vP
It moves at a speed of d (however, it stops if the vector is zero).

From the above, the PWM control pattern of the voltage source inverter (3) appropriately selects the voltage vector vP so that it moves at an angular velocity ω along the vector locus on the complex plane of the time integral 2P of the voltage vector or along the circumference of the specified radius R. and decide. (The designated radius R is R=V, /ω, where the effective value of the line voltage of the fundamental wave voltage is 1 and the angular frequency is ω).

That is, for example, as shown in FIG. 4, the angle φ is 0≦φ≦π
/3, the voltage vector v4. , vO and a zero vector (for example, Vo), stay at point po for time τ○ (this state is indicated by symbol 0), then take v4 for time τ4 to reach point q1, and then set vO for time τ6 Consider the case of reaching point P1. In this case, △P
OQIP+-]3-, pop,=V,-T.

Since Poq+=J7g¥−Vd ・τ4 q 1 p+ = JTIT vd ・τ6 and τ0+τ4+τG−To, solve Equation 1 to obtain the voltage vector vI within the period To.

vO, time to take vO τ4. τ6. τ0 is obtained.

r4/ T□ =kS -8in(π/3-φ○)T6
/ To = kS dispute Sjn φ○ro/ To =
1-ks-8in(φ○+π/3) (3) However, ks is the voltage control rate, and kS=JT'V1/Vd.

Equation (3) in ``1'' is a relational expression when the angle φ is in the range O≦φ≦π/3, but in other sections, since the inverter (3) performs symmetrical three-phase operation, the following By replacing each symbol as shown in Table 1, a relational expression in the range 0≦φ≦2π can be obtained.

Next, based on the time τ of the voltage vector in the above equation (3), 0N10F of each transistor (Tra) - (Trc')
Find the F pattern (PWM control pattern). In this case, the relationship between the voltage vector time τ and the PWM control pattern changes depending on the order in which the voltage vectors are taken, so for the sake of simplicity, the same pattern is repeated in each period TO, and If we add the constraint that the 0N10FF switching of the transistor is only once, the PWM control pattern is represented by the four patterns shown in Figure 5 (a) to (d) (Figure 18 τ1 is on the →- side Transistor O
N time, and τ- indicates the ON time of one side transistor, respectively).

In this embodiment, the PWM control pattern shown in FIG. In the voltage source inverter (3), the PWM control pattern is uniquely determined by knowing the name of the transistor that turns on at the beginning of period To and the time it takes to turn on SF, so (3) Referring to the equation and FIG. 5(a), the PWM control pattern is determined by the following equation within the range of angle φ or 0≦φ≦π/3.

= 16-ra/To =1. -ff # (V+ /Vd
)・5in(φ○+π/3) τb/T○−1. -JT ・(V+ /Vd)・
Sin φ0τc /T○−1 (always ON) ・・・(4) In the relational expression (4) of the PWM control pattern in the range of 0≦φ≦π/3, each symbol is replaced in the same way as above. Then 0
The relational expression is within the range of ≦φ≦2π.

Next, the operation of the 1-chip microcontroller (8) is shown in Figures 6 and 7.
This will be explained based on the control flow shown in the figure with reference to FIG.

The control flow in FIG. 6 is for each transistor (Tra) to (
This is a calculation flow of the ON time (PWM control pattern) of the transistor (Tra) (Trc"), and the control flow of FIG. To explain from the control flow, this control flow is repeatedly performed every calculation period To (for example, 200 μs) according to the carrier frequency (for example, 5 KHz),
After inputting the phase ωt (-φ○) of the output voltage and the amplitude ■1 of the output voltage in step SAI, in step SA2 each transistor ('l''ra ) - (Trc') ON time τ(
0 month) is calculated.

After that, in step SA3, the value of the division number N corresponding to the magnitude of the torque ripple signal received from the torque ripple detection circuit (13) (see Fig. 1) is set in the division number setting circuit (14) in its built-in table. Transistor (Tra) − (T
rc') ONN time (n+1.) is divided into N
(for example, 4), each ON time τ(n+1) is divided by a plurality of N (4) pulses τ' (n+1) (τ゛(
n+1)-τ(n+1.)/4). and,
This divided pulse τ' (n+1) in step SA4
is stored in one switching time register for each phase (in a voltage-type inverter, one transistor in each phase arm is always in the ON state, so only one for each phase) as shown in FIG. Return.

In addition, the control flow in FIG. 7 is faster than the repetition period To° or the calculation period To in FIG. − −To/N (The number of divisions N is N = 2” (m−1,,2
) is preferable). On the other hand, after the divided pulse τ' (n+1) is stored in the switching time register of each phase in the control flow shown in FIG. 6, as shown in FIG. Enter the contents of the switching time register in SBI, and proceed to step S8.
2, the corresponding transistors (Tra) to (Trc') with divided pulses τ' (n+1) every period T o '
is repeated by repeating ON control l~.

Therefore, in the calculation flow of the PWM control pattern shown in FIG.
Turn on each transistor (switching element) (Tra) to (Trc') based on the relational expression (/l) of the control pattern
Calculating means (10
).

Further, in step SA3, the division number setting circuit (1
From the built-in table of 4), read out the value of the division number N corresponding to the 1× ripple of the induction motor (1), that is, the ripple based on the current of the three-phase winding (2), and use this division number N to calculate the calculation means ( 10) Each transistor (Tr
a) It constitutes a dividing means (11) which divides the ON time τ of -(Trc') into the plurality of N pulses, and according to the control flow shown in FIG. A control means (12) 114 is configured to turn on each transistor (Tra) to (1re') using a plurality of N equal-width pulses divided by (]]).

Therefore, in the first embodiment, P W M j
In the calculation flow of the control pattern (Fig. 6), each transistor (T
The ON time τ of ra) to (Trc') is calculated by the calculation means (10)
After calculating each ON time τ, dividing means (]
]) is divided into a plurality of N (for example, N-4) pulses τ", and these divided pulses τ" are stored in registers between the input signals 2 and 144 of each phase of a, b, and C.

Then, in the subsequent period T○, as shown in FIG.
) ~ (Trc') ON time τ is calculated and divided, and in this current period TO, its TO/N (
−T○゛), a calculated in the previous period T o
, l), each transistor (Tra) corresponding to the divided pulse τ' in the switching time register of each phase C
-(Trc') is controlled to be turned on by the control means (12), so the frequency of the high frequency component is higher than that of the conventional one (which does not divide the ON time into multiple pulses) as shown in Fig. 21. The carrier frequency can be equivalently multiplied by the number of divisions of the ON time N (N=4), and the original carrier frequency (5KIlz) can be multiplied by a higher carrier frequency (20
KII z) can be increased.

Note that FIGS. 8 and 21 show the case where the ON time of the + side transistor of each phase is calculated.

Here, the original carrier frequency (5 KHz), that is, the operation period T○ (200 μs) of OIV 1 time, is a sufficient period for calculating the PWM control pattern with one chip microcontroller (8), so High carrier frequency (2OKI) without using chip microcontroller (8)
By making PWM control possible (around Iz), it is possible to precisely control the three-phase AC waveform to the induction motor (1) by making use of the capabilities of high-speed switching elements such as MOSFETs, while keeping the circuit configuration simple and at low cost. This makes it possible to reduce electromagnetic noise and increase motor efficiency.

In addition, the gear rear frequency (5KIIz) is the same as before.
If this is sufficient, the calculation time of the one-chip microcomputer (8) can be shortened, and the processing capacity for processes other than PWM control can be increased.

Moreover, as shown in FIG. 10, the torque generated by the induction motor (1) has a ripple that increases every π/3 in electrical angle, or this ripple is detected by the torque ripple detection circuit (13). Then, the division number N corresponding to this ripple value is set by the division number setting circuit (14). As a result, in a situation where l·le ripple is large, the division number N is set to a large value, and the ON time is divided by the dividing means (11) more often than usual in a situation where l·le ripple is large. As a result, the torque ripple of the induction motor (1) is effectively suppressed, and the induction motor (1) is rotationally driven with a nearly uniform torque. That will happen. Note that the number of divisions N is 10 in the configuration shown in Figure 2.
The settings may be made in advance based on the diagram; in this case,
Torque ripple detection circuit can be omitted.

In addition, in the above embodiment, the value of the division number N of the ON time of each transistor (Tra) to (Trc') was changed depending on the torque ripple of the induction motor (1), or the value of the division number N of the ON time of each transistor (Tra) to (Trc') was changed, or the three-phase winding ( 2) When suppressing current ripples and voltage ripples leading to 2), each transistor (Tra) to (Trc')
By understanding the rate of change of the ON time and the phase of the signal wave,
As shown in the figure, the number of divisions is N in the range where the rate of change in ON time is large.
If the number of divided pulses is set to a large value and the number of divided pulses is increased, the current ripple and voltage ripple can be effectively suppressed as in item 1.

12 to 17 show a second embodiment, in which, instead of dividing the pulse into a plurality of equal width pulses in the first embodiment, each transistor (Tra) - (Trc') has a plurality of ON times. This is applied to pulses divided into unequal width pulses.

In other words, this division into unequal width pulses is based on the ON time τ(n) of the transistor in the period FO and the subsequent period To
This is performed by linear interpolation between the ON time τ (n→1) and the ON time τ (n→1). facing, as shown in Figure 12.
The ON time of each transistor (Tra) - (Trc') calculated by the ON time calculation circuit (10') is calculated by the dividing circuit (]
]') when dividing into multiple N unequal width pulses,
The value of this division number N is determined by the division number setting circuit (14°) to determine the phase ωt of the output voltage (that is, the rate of change in ON time) and the amplitude V.
The number of divisions is set to N according to 1. Here, P
As can be seen from the relational expression (4) of the WM control pattern, for example, when the angle φO is in the range of 0≦φO≦π/3, the ON time τa of the transistor is < 5 in (φ
0 + π/3), and the rate of change in the ON time is small, so good waveform reproducibility can be ensured even when the number of divided pulses is a normal value, so the value of the number of divisions N is set to the normal value ( For example, set it to N-4). In addition, the 0IV1 time τb of the transistor is in the range of Sinφ○, and the rate of change of its ON time is large, so the value of the division number N is increased (for example, N-8) to ensure good waveform reproducibility. Set to . Considering the above, the angle φ○ is O≦φ○
In order to expand the range to ≦2π, the entire interval is divided into 6 intervals at intervals of π/3, and the result is shown in Table 2 below. The division number table created in consideration of the above is stored in advance in the division number setting circuit (14°).

Next, the division into unequal width pulses will be explained in detail. In the control flow shown in FIG. 13, calculation processing is performed in a period of 1゛○ cycle, and in step SCI, the phase ω and amplitude V1 of the 11'' force voltage are input, and in step SC2, each of the previous transistors (T
The calculation result of the ON pulse r'(1-1) (divided pulse divided by the number of divisions N) of ra)-(Trc') is input.

After that, in step sC3, each transistor (Tra) to (
While calculating the ON time τ(n) of Trc "), "1
The phase ω of the read output voltage and the value of the division number N corresponding to the amplitude V1 are successively set in the division number setting circuit (14°), and these multiple values are set from the ON time τ(n) calculated above. A divided pulse τ°(n) divided into N is calculated. after that,
At step SC4, each transistor (Tra) to (Trc
The interpolated value Δτ of the previous divided pulse τ' (n-1) is determined according to the difference between the previous divided pulse τ' and the current divided pulse τ'). -1 is calculated based on the following formula.

△τn-1-'τ゛(n)-τ'(n-])]HaI
N: Number of divisions. Step S
At C5, each divided pulse τ' (n-1) is sequentially given Δτ, 2・Δτ, 3・Δτn−1, 1l−J14・Δτn−1・
In step Sc6, each divided pulse of the lever is stored in a plurality of N (N-4) switching time registers for each phase, and in step SC7, each divided pulse τ゛(
Memorize Kano 1) and return.

In addition, the control flow in FIG. 14 is as shown in FIG.
○1", each transistor (Tr
a) to (Trc') are ON-controlled, and the control period T○゛ is the calculation period To of the control flow in Fig. 13.
I/IJ (N is the number of divisions).

In this control flow, as shown in FIG. 16, the divided pulses τ' for each phase stored in the first switching time register are read in step SDI, the switching time register is shifted in step SD2, and the switching time register is shifted in step SD3.
Then, each transistor (Tra) to (Trc') is turned ON using the read divided pulse r' (1st issue) and the process returns.Then, in the same way, the second and third transistors are sequentially controlled every control period T○'. Reading the divided pulses for each phase stored in the third and fourth switching time registers and controlling each transistor (Tra) to (Trc') to turn on is repeated.

Therefore, in the calculation flow of the PWM control pattern shown in Fig.
(Trc”) ON time τ of each transistor (Trc”)
The ON time τ of a) to (Trc') is divided by the division number N, which is set to a large value when the rate of change of the ON time τ is large, and the ON time τ is divided into a plurality of N unequal width pulses (τ□, ( τ°ten△τ),
(τ□+2 ・Δte)2(τ°+3 ・Δτ), (τ°
+4 ・Δτ)・) The dividing means (I
I').

Therefore, in this embodiment, the dividing means (11゛)
If the ON time is divided by unequal width pulses as shown in FIG. Since it is repeated every control period T゛○ that a divided pulse larger by the interpolated value Δτ than Good waveform reproducibility can be ensured with respect to the average value (■) of the output voltage in the control period T'o corresponding to the carrier frequency. In addition, when dividing into unequal width pulses by linear interpolation, the calculation and processing time required by the microcomputer (8) for the division is relatively short, and the equivalent carrier frequency can be sufficiently It can be improved.

Furthermore, FIG. 18 shows a third embodiment, in which it is possible to determine whether the ON time of each transistor (Tra) to (Trc') is divided by equal-width pulses or by unequal-width pulses. This was applied to a method selected appropriately according to the rate of change.

In other words, in the same figure, within the phase range of the signal wave (indicated by the broken line in the figure), in the range where the rate of change in ON time is large, the ON time is divided into unequal width pulses, and the rate of change in ON time is small. In the phase range of the signal wave, the ON time is divided into equal-width pulses, and the setting of the number of divisions N is as specified in Section 2 of "1.
This is similar to the example.

Therefore, in the third embodiment, similarly to the second embodiment,
It is possible to improve the reproducibility of the signal wave in a range where the rate of change of the ON time is large, and to ensure good reproducibility of the signal wave in a range where the rate of change of the ON time is small. Compared to the case of dividing by pulses, this method has the effect of eliminating the need to calculate interpolated values and saving calculation and processing time.

In addition, as described above, each transistor (Tra) −(Trc
When the number of divisions N of the ON time of the inverter (
When the short-circuit prevention time Td of the upper and lower arms of the transistor is equal to or shorter than 3), the number of divided pulses is reduced to prevent pulses from disappearing, and the reproducibility of the demodulated waveform is ensured.

The control flow is shown in FIG. In the same figure, starting from the start, each transistor (Tra) is
The ON time of ~(Trc') is calculated, and in step SE2, the value of the division number N of the ON time is read and set in the same way as above, and the value of the division number N of the ON time is calculated, and each of the above transistors (Tra,
) to (Trc') are divided into a plurality of N times.

Then, in step SE3, the width T of this divided pulse is compared with the short circuit prevention time Tcl of the upper and lower arms, and if T≦Td, it is determined that the output pulse disappears, and step SE4
Reduce the number of divided pulses, that is, the number of divisions N, and return to step SE2 above to divide each transistor (Tra).
The ON time of ~(Trc') is divided by this smaller division number N.

Then, in step SE3, when T>Td and the pulse does not disappear, in step SE5, each transistor (Tra) to (T+・c') is controlled to be turned on every cycle T○゛ using the divided pulses. and exit.

Note that the part for converting the contents of the switching registers of each phase into pulse widths may be processed by hardware such as an external pulse width modulation IC. Furthermore, Figures 9 and 16
If the configuration shown in the figure is used, even if the carrier frequency changes due to a change in the switching element, the dividing means (
It is sufficient to change only 11) and the control means (2).

Furthermore, if the switch time register is shared with the register of the pulse width control section (step 5B2), the process of step SBI in FIG. 7 can be omitted.

Furthermore, the calculation period TO in the calculation flow of the PWM control pattern is uniquely determined by the time required to actually calculate the PWM control pattern, or the ON of the transistor in the control flow of FIGS. The control period To' is
Basically, it is determined according to the desired carrier frequency, and for this purpose, the number of divisions of the ON time of each transistor N (To/
Basically, the value of T○゛) may be set to an appropriate value.

In addition, in the above embodiments, the PWM control pattern is determined based on the relational expression (4) when using voltage vector control, or based on the relational expression when using other PWM control methods such as the triangular wave comparison method. Of course, you can.

(Effects of the Invention) As explained below, according to the PWM modulation control device for an inverter of the invention described in claims (1) and (2) of the present application, the Since the ON time of the switching element to be calculated is divided into multiple pulses and each switching element is ON-controlled using the divided pulses, even if it takes a relatively long time to calculate the ON time of the switching element, carrier By increasing the frequency to an evenly polarized value, for example, even when using a low-cost, single-chip microcontroller with a simple circuit configuration, three-phase AC waveforms can be precisely controlled, reducing electromagnetic noise and increasing motor efficiency. I can try to figure it out. Moreover, since the number of divisions of the ON time is changed according to the ripple based on the current of the three-phase winding and the rate of change of the ON time of each transistor, more precise waveform control is possible, and the current ripple and Voltage ripple and torque ripple can be effectively reduced.

[Brief explanation of the drawing]

1 to 11 show a first embodiment of the present invention, in which FIG. 1 is a general schematic diagram, FIG. 2 is an electric circuit diagram, and FIG.
The figure is an explanatory diagram showing the various states of a voltage source inverter using eight types of voltage vectors. Figure 4 is an explanatory diagram of voltage vector control to bring the locus on the complex plane of time integration of the voltage vector closer to a circular locus. , FIGS. 5(a) to 5(d) are explanatory diagrams of types of PWM control patterns that can be taken within the range of 0≦φ≦π/3 of angle φ, and FIGS. 6 and 7 are each for one-chip microcontroller. Flowchart showing 0N10FF control of each transistor according to FIG. FIG. 11 is an explanatory diagram of the range in which the number of divided pulses is increased. 12 to 17 show the second embodiment, FIG. 12 is a block diagram of a microcomputer when the ON time of a transistor is divided into unequal widths, and FIGS. 13 and 14 are 0N1 of each transistor.
Flowchart 1 diagram showing 0FF control, FIG. 15 is an explanatory diagram of interpolation of divided pulses, FIG. 16 is an explanatory diagram of operation,
FIG. 17 is an explanatory diagram of the state of waveform reproducibility. Furthermore, FIG. 18 is an explanatory diagram showing the phase range of a signal wave for selecting division into equal-width pulses and division into unequal-width pulses, showing the third embodiment. In addition, FIG. 19 is a flowchart showing pulse width control for short-circuit prevention of the lower arm of the inverter, and FIG. 20 is an explanatory diagram when the ON time of the transistor is divided into a plurality of equal-width pulses. . Moreover, FIG. 21 is an explanatory diagram showing a conventional example. (2)...Three-phase winding, (3)...Voltage type inverter, (
4)3. Bridge circuit, (Tra) - (Trc'l-
L-transistor, (8)...1 chip microcontroller, (10
)・Calculating means, (11), (11')...Dividing means, (
12) Control means, (13) Ripple detection circuit.

Claims (2)

[Claims] (1) A bridge circuit (4) connected to the three-phase winding (2) and having a plurality of switching elements (Tra) to (Trc'), each switching element (Tra) of the bridge circuit (4) This is a pulse width modulation control device for an inverter, which applies a three-phase AC voltage to the three-phase winding (2) by pulse width modulating the DC by ON/OFF operation of ~(Trc'), O of each of the above switching elements (Tra) to (Trc') at a calculation period according to
a calculation means (10) for calculating N time;
Each switching element (Tra) to (T
rc') into a plurality of pulses according to the ripple based on the current of the three-phase winding (2); and a plurality of pulses divided by the dividing means (11). The above-mentioned switching elements (Tra) to (Trc
1. A pulse width modulation control device for an inverter, comprising: control means (12) for controlling ON of the inverter. (2) In the PWM modulation control device for an inverter according to claim (1), in place of the dividing means (11), the calculating means (
10) Each switching element (Tra) ~ (
A pulse width modulation control device for an inverter, comprising dividing means (11') for dividing the ON time of Trc') into a plurality of pulses, the number of which corresponds to the rate of change of the ON time.