JPS60190168A - Pulse number switching system of pwm inverter - Google Patents

Abstract

PURPOSE:To eliminate the variation in the torque at pulse number switching time by storing the phase difference of pulses between carrier signals for switching the number of pulses and calculating to correct the phase difference at the switching time. CONSTITUTION:A noninterference calculator 6a calculates the primary voltage of alpha-beta shaft, converts by coordinates converter 7 the primary voltage into polar coordinates data, and then inputs to a PWM waveform calculator 8. PWM pattern necessary to calculate in the calculator 8 is given by a pulse pattern generator 10 and a pattern data converter 11. The generator 10 is, for example, formed of an ROM to give a pattern data in response to the number of pulses. The converter 11 converts the pattern data in response to the angular frequency and outputs it. A gate circuit 9 obtains a PWM wave of actual each phase inverter in accordance with the data from the calculator 8, and controls ON or OFF of the switching element of the inverter 1 in accordance with the voltage.

Description

DETAILED DESCRIPTION OF THE INVENTION (Technical Field) The present invention relates to a pulse number switching system for PWM (Barne Radiation Modulation) impark.

(Prior Art and Problems) In a PWM inverter, a synchronous type is used in which the positive and negative PWM waveforms are completely equalized in order to prevent the occurrence of DC'L pressure component. In this synchronous PWM inverter, simple control is to make the carrier wave a constant number of pulses with respect to the sine wave period, but in this case, at low frequencies, the low-order harmonics are high, and at high frequencies, the switching frequency is There are some drawbacks. Therefore, conventionally, a pulse number switching method has been adopted in which the carrier wave frequency is switched according to an increase or decrease in the sine wave frequency (velocity). In this pulse number switching type PWM inverter, the inventors of the present invention have discovered through research that when the pulse number is switched, an angular correction error occurs in the AKPWM rectangular shape, causing torque fluctuations in the load motor. This torque fluctuation is
Vector control, especially precise control such as non-interfering vector control in which the secondary magnetic flux and secondary current of the 1a motor are brought into direct contact with each other without interfering with each other, poses a serious problem.

(Object of the Invention) An object of the present invention is to provide a switching method that eliminates torque fluctuations when switching the number of pulses.

(Summary of the Invention) The present invention is characterized in that in order to switch the number of pulses, the phase difference between each carrier signal at each pulse is stored, and at the time of switching, an operation is performed that corrects the phase difference. .

(Embodiment) FIG. 1 is a block diagram of a non-interference vector control device showing an embodiment of the present invention. In order to perform variable speed vector control on the inductor tjfilJm2 whose primary voltage is the output of the voltage source inverter 1, the angular frequency set value ω-, which becomes the speed setting, is controlled by the relaxation obtained from the pulse pickup 6 and the frequency-voltage converter 4. Motive speed detection signal ω. The output of the speed control amplifier 5 which performs a proportional integral calculation on this deviation is obtained as the β-axis component 11β of the primary current corresponding to the secondary current of the motor 2.

On the other hand, the α-axis component lIa* of the primary flow corresponding to the secondary magnetic flux of the electric motor 2 is set. In other words, to control the speed of the electric motor 2 by primary voltage control, a two-phase voltage eIa+el/ having α and β axes rotating in synchronization with the negative voltage is set, and the α axis is determined as the motor secondary magnetic flux. By aligning the secondary current with the β-axis and making the current-magnetic flux orthogonal, this two-phase voltage eI
From the equivalent circuit of the motor, α-axis and β-axis components of the primary current are set for a + ``Iβ''. Then, the α-axis current component 'l corresponding to the magnetic flux setting is set.
With a* fixed, the β-axis current i, which corresponds to the current, /
By adjusting , the speed can be controlled while making the secondary current and magnetic flux orthogonal.

The non-interference calculation unit 6 calculates the α-β-axis primary voltage eI according to the α-axis-order current setting value Lla* and the β-axis-order current command il/.
cf r eIβ is calculated, and a non-interference calculation is performed to eliminate the mutual interference component. This arithmetic expression is expressed as follows.

Here, r is the motor primary resistance, and ω0 is the angular frequency command (
(inverter operating frequency), L is the equivalent leakage inductance, and Ll is the primary inductance.

The above formula requires a current of 1. +L due to β
, ωoi + β, so this interference is subtracted from r1i□, and the secondary current setting is set by
``Since there is an interference amount of tila, this means performing a non-interference calculation of adding the interference amount to r, i, and β.

The coordinate conversion unit 7 converts the primary voltage ela*ej/ into polar coordinate data. This conversion is as shown in the following equation based on the relationship shown in FIG.

IEI = e(B" + 6.," +++m (3
The JPWM waveform calculation unit 8 calculates the polar coordinate signal I of the voltage vector t.
E1. From φ, the signals corresponding to the three-phase fixed axis voltages e, eb, and ec are treated as PWM waveform processed data. Incidentally, as is clear from FIG. 2, the relationship between the voltages e, eb, ec and the polar coordinate signal is as follows. Here, θ is the angle between the a axis and the voltage Bekulele 9, and the α-β axis is the angular frequency ω. Since it rotates at , in steady state θ, = ω. It becomes t. However, if the voltage eIcl+el/ changes at the torque fluctuation or the epicenter, the phase φ
Since , changes, this change is expressed as θ, resulting in the following equation.

θ,=ωot+ψ (6) ψ is the amount of phase change in one sample period in digital calculation, and is ψ−φ.

ld-φn. However, φ. 1° is the phase one sample before in equation (4) above, and φnew is the phase at the time of the current sample.

The gate circuit 9 obtains the PWM waveforms of the actual inverter compensating voltages e, eb, and ec of each phase according to the data from the calculation unit 8, and instantly controls the switching elements on and off according to this voltage. No (Rusno (turn generation part 10)
The pattern data switching unit 11 provides PWM backing data necessary for calculation in the calculation unit 8, as will be explained in detail later.

Each of these sections 6 to 11 performs digital processing, and each section except the gate circuit 9 can be implemented as a microcomputer, for example, in accordance with a necessary operation xe program. For this reason, %Kf, ls. *.

11β and angular frequency ω. The input is taken in as a digital quantity that has been sampled and analog-to-digital converted.

The slip frequency calculation circuit 12 calculates the slip frequency ω8 from the output current setting signal 11- and the secondary current command 11β. This calculation is performed according to the following equation (7). Here, τ is the ratio Lt/rz of the secondary inductance L2 and the secondary resistance r.

This performance'r4. The slip frequency ω8 output of the circuit 12 is sent to the adder 16 as the speed detection signal ω of the converter 4. The angular frequency ω is summed with the output. I get criticized. In addition, these circuits 12.13
When processing each part 6 to 11 digitally, the converter 4 is configured as a counter, the operation of the amplifier 5 is also a digital operation, and the 4,5° 12 side is used as a microcomputer on the manuta side for sequence control, failure diagnosis, etc. have the process done.

Next, the arithmetic processing of the three-phase voltage signals e, eb, ecg) will be explained in detail below, focusing on each section 8 to 11.

The pulse pattern generator 10 has a control rate μ==1 as a reference control rate, a reference angular frequency of 9, and P synchronized with a sine wave.
WM waveform pulse pattern data θn is the carrier wave pulse number P
(Generated in accordance with the sine wave half cycle period. This pattern generating section 10 is composed of, for example, an ROIVI, and is configured to read pattern data according to the number of pulses P as a data table.Pattern data switching section 11
is the angular frequency command ω. The pattern data θn? Switch and take out. This pattern data switching is for setting an appropriate number of pulses Pi depending on the level of the angular frequency command ω0. The PWM waveform encyclopedia section 8 converts the pattern data θn seen throughout the switching section 11 into control rates μ and ω. The adjusted data Ua(-μ·θn) is output as data for forming a PWM waveform. Here, the control rate μ is unitized using the data of the voltage IE+ and the DC voltage Edc of the inverter 1 using the following formula.

E.

had/. 4.2 Extraction of PWM waveform data based on the sine e pulse pattern and its performance side in the pulse pattern generating section 10 will be explained with reference to FIG. Figure 3 shows the case where the number of carrier wave pulses P = 9. is as shown in the same figure (bJ).

Triangular wave C is the pulse pattern data of this PWM waveform.
l7) Store the angle on (n=x~2P) from zero point P, ~PI8 to the father point of the sine wave and triangle vC as a numerical value. This angle θ. are divided into groups according to the number P of carrier wave pulses and made into a large table. Here, the number of pulses P is set as p=5m+3 (m=0.1°2...,k) in order to use a fully synchronous equal-pulse sine wave PWM method.
Let it be IM.

The above pattern data θ. is different from the actually required 9 turns, and the control rate μ and the angle T starting from the apex of the triangular wave C
Although it is different from θ, this can be determined by the following calculation.

Angle θ due to control rate μ. The change in is regarded as the angle θ which is approximately proportional to the control rate μ.

θ = μ·θ (9) If Tax is the period θa of the triangular wave C, it can be determined from the formula in the following table.

Also, from Figure 3, when the slope of the triangular wave is positive, the angle Tax
Actual PWM waveform pattern data can be obtained by specifying that the section is at a high level, and when the slope is negative, the Tθ section is at a low level.

Therefore, pattern data θ with control rate μ=1. A data table is created for each type of pulse P, and the data θn(
i- is selected, and the arithmetic unit 8 calculates the above (
9) Calculation of equation and conversion to angle T11x according to the above equation 'J4. (including the distinction between high level and low level of TIlx) and ω. By converting the data into real-time data, pattern data for forming an actual PWM waveform can be obtained. ω. The conversion of tx to real-time data by is 1 for the angle anti-data Tθ, cage = 'l'a! -Side/Page: Determined by the calculation in (10).

In addition, although the pattern data for only one phase has been shown above, if the above data is for C phase θX&, data for b phase and C phase is θxb e θworking. Since they are in phase delayed by 120°, it can be determined by picking up data at a point delayed by 120° with respect to θ0, and C
The phase is θXe” (θX,
1) It can also be intertwined.

Next, the gate circuit 9 is configured as shown in FIG. The figure shows a bus-coupled configuration with a microcomputer 8A as the calculation section 8. The programmable timer 21 is composed of a counter timer T□ and a monostable multivibrator Tlt, and a value ν2 corresponding to 172 of the period OT of the triangular wave C is preset to the timer TII via H/X and 8B, and this value is clocked. By counting down with the period of CLK, one pulse output is obtained every half synchronization τ72 of the triangular wave C. 1 This pulse is input to the multivibrator 'I' + t which is synchronized with the clock CLK. A timing signal tII is obtained every half cycle. This timing signal tll is matched with the timing of the positive and negative vertices of the triangular wave C.

The triangular wave slope state latch circuit 22 consists of two D-type frisodes 7.
The flip-flop FF receives triangular wave slope state data D from the computer 8A. (“1” when the slope is positive.

When negative, “0”) is given by write command WR,
The Q output of the flip-flop FF is tied 12.
1 timing signal tl! is taken in. Therefore, the output D1. of the latch circuit 22. Each issue corresponds to the upward and negative periods of the triangular wave C to the high and low levels.

The programmable timer 26 has counter timers TI3 + T14 + T15 (
data t corresponding to the angle TOx from the apex of the triangular wave for each phase from the computer 8A.
x (the above-mentioned (formula) is preset. This presetting is performed by transferring the data given from the computer to the data latch to the counter using the timing signal tll, and the preset value t is counted using the c-1 clock CLK. The programmable timer 26 outputs logic "1" for only the period. Therefore, the programmable timer 26 outputs for each phase to No. 18 T, , Tb, and Tc' of the time width from the apex of the triangular wave to the intersection with the sine wave.

The conventional data exchange between the control means and the computer 8A is executed by applying the timing signal tll to the computer 8A as the interrupt signal 1NTR0.

The logic section 24 receives the output D1. of the latch circuit 22. and the output T of the timer 26, i'6. From Tc to each phase a, b, c
PWM waveforms e, , eb, and ec are formed. For example, for the a phase, by taking the AND of the output T8 of the timer 26 and the output 2 of the latch circuit 22 at the gate G, a signal E having a width from the top of the triangular wave to the intersection with the sine wave in the upward slope period is obtained. , + is obtained, and by taking the AND of the inverted signal of the output Ta obtained from the inverter G2 and the output D□ at the gate G, a signal E, which has a width from the peak of the triangular wave to the intersection with the sine wave in the slope negative period, is obtained. - is obtained, and the a-phase PWM waveform eat- is obtained by taking the logical sum of both signals Ea+ and Ea- at gate G4.

Furthermore, an inverted signal ζ of ea is obtained by gate G.

As described above, by giving each sum real time data TXk to the gate circuit 9 every half period ν2 of the triangular wave, each phase a, b, c
The voltage signals ea, eb, and ec of the PWM waveform can be obtained, and the envelope frequency (inverter operating frequency) f of the signal e@+eb+ec. It becomes.

And pattern data θ. For P=9, the call of θ1
．． θ7. U,...θ1. If we assume that the order of 018. is the positive phase rotation direction of electric warfare 2, then the call is reversed to 018. θ, 7
...θ7. By setting θl + "1g, the electric motor 2 can be rotated in the opposite phase. Therefore, switching between forward and reverse rotation of the electric motor can be easily realized by switching the reading direction of the data θ□ in the arithmetic unit 8.

Next, the frequency ω in the calculation unit 8. The process of calling data On regarding switching of the number of carrier wave pulses P due to the change will be explained. When switching the number of pulses P, if the number after switching is the same number n or n±1 for the calling number n before switching, the number of pulses P before and after switching is different, so the voltage 1
A large phase change and pulse width change appear in No. 3 ea, eb, and e, which causes torque fluctuation.

Therefore, it is possible to reduce the phase change by selecting a call number n that has carrier waves close in phase before and after switching to the sine wave. This is determined by the following formula.

Here, p new is the number of pulses after switching, Po1d is the number of pulses before switching, nneW is the calling number at the number of pulses pnew, and "old" is the calling number at the number of pulses Po1d. For example, from P=15 to P= 9, each triangular wave calling number P+ -Ra, (P+)
The relationship between ~(Pso) and the sine wave is shown in Figure 5,
If n=5 before switching, then from the above equation (13), nnew
3, and the fifth phase (for the sine wave) at P=15 will be the closest number to the third phase at P=9.

In this example, nnffw is an integer, but a phase difference will appear when switching with a leaf number that is a fraction in equation (13) above, and if this phase difference is in the direction of delay, it will become a braking torque on the electric motor and the speed will increase. If it is in the descending or leading phase, it becomes an accelerating torque, resulting in speed control with overshoot.

However, if the motor has a large GD", a slight phase difference will not be a problem, but it will not be suitable for applications that require accurate torque control using vector control.

For this reason, the present invention proposes a pulse number switching method that makes the phase difference zero. This will be explained below.

The phase difference Δ determined by the setting of n according to the above equation (13)
In order to correct X, the set value of timer 21 (0T/
2) and the set value TX (T, RI) of the timer 26 is corrected and set by 611 minutes only once at the time of switching.
Calculation of Xp can be determined by calculation, but since it is uniquely determined once the call number n is determined, the pattern data θ. Data θ can be added in advance as attached data to . can be read simultaneously when calling.

Fig. 6 shows the switching state by correcting △Xp, and when switching the number of pulses between P=9 and P=15,
By correcting the timer set value already at the apex by the difference ΔXp up to the apex of =9, a PWM waveform with no phase difference can be obtained. The set time of the timers 21 and 23 at this time is as shown in the following equation.

Also, the control flow centered on the arithmetic unit 8 in the microcomputer configuration is as shown in FIG.

Although the embodiments will be described using a non-interfering vector control device, it goes without saying that the present invention can be applied to a normal PWM inverter to obtain the same effects.

(Effect of optical cutting) As described above, according to the present invention, since the phase difference Δxpk between both carrier waves is corrected when switching the number of carrier wave pulses, the phase change caused by switching the number of pulses is eliminated, and high precision imaging over a wide range without torque shock is possible. Shift PWM control becomes possible. In particular, switching can be performed immediately at the timing of phase mismatch without waiting for the phase-failure point of both carrier wave pulses, and rapid switching of the number of pulses without phase change is possible even during low-speed operation. In terms of configuration, it is only necessary to add the data of the phase difference ΔX to the pattern data and add processing for switching using the data ΔX to the arithmetic processing.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a non-interfering vector control device according to an embodiment of the present invention, FIG. 2 is a vector diagram for explaining polar coordinate conversion processing in FIG. 1, and FIG. 3 is a pulse pattern according to the present invention. 4 is a circuit diagram of the gate circuit 9 in FIG. 1, FIGS. 5 and 6 are waveform diagrams for explaining the operation of switching the number of pulses, and FIG. 7 is a diagram showing the operation section 8. FIG. 1... Voltage type inverter, 2... Induction motor, 6...
...Pulse pickup, 6...Non-interference calculation section, 7.
...Coordinate transformation section, 8...PWM waveform calculation section, 9...
Gate circuit, 10... Pulse pattern generation section, 11.
...Pattern data switching section.

Claims (1)

[Claims] In a PWM inverter control device that obtains an inverter voltage signal with a sinusoidal PWM waveform by switching the number of carrier wave pulses in accordance with an angular frequency command, both carrier waves are added to the first l) WM hexagonal signal when changing the number of carrier wave pulses. A PWM inverter pulse number switching method that waits for switching after correcting the phase difference.