JPH0561871B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention relates to a DC-AC conversion method that converts DC to AC using pulse width modulation (PWM) to reduce harmonic components. The switching transistor _Q1 shown in FIG.
The PWM inverter control method is used to approximate the line-to-line output voltage waveform of the output line UVW of a three-phase bridge inverter consisting of Q ₆ to a sine wave and remove harmonic components. in this kind of way
To obtain the PWM output voltage, the transistor Q ₁ ~
_Q6 must also be controlled by PWM waves. Traditional
As shown in Figure 2 as an example of the PWM control method,
A triangular wave is generated at a constant phase position, and a control pulse is determined by the intersection of the triangular wave and the voltage control voltage V _R of a selected value corresponding to the inverter output voltage _. For control, there is a method of obtaining the control pulses shown in FIG. 2B. However, both analog and digital techniques make the control circuitry complex and expensive. SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a DC/AC conversion method that can relatively easily obtain an output voltage approximating a sine wave. To achieve the above object, the present invention provides that the waveform in the range of 0 degrees to 180 degrees of the single-phase output voltage or multi-phase line-to-line output voltage of a single-phase or multi-phase DC/AC converter is
N rectangular wave pulses having the same pulse width not larger than the predetermined angular interval θs of 120 degrees/n (where n is an integer) are applied in the range of 30 degrees to 150 degrees with a repetition period of the predetermined angular interval θs. , any one or more of the pulses located in the range of 30 degrees to 60 degrees in the original pulse train arranged symmetrically about 90 degrees from each position 30 degrees without changing the amplitude and pulse width.
move each pulse to a symmetrical position with respect to the angle, and reverse the polarity of any one or more pulses within the range of 60 degrees to 90 degrees without changing the amplitude and pulse width. 60 degrees towards 0 degrees from the position
of the pulses located in the range of 90 degrees to 120 degrees, and reverse the polarity without changing the amplitude and pulse width from each position.
Move any one or more pulses within the range of 120 degrees to 150 degrees by moving them 60 degrees in the direction of 180 degrees and moving them 150 degrees from each position without changing the amplitude and pulse width. If the pulses are moved to symmetrical positions and n pulses are arranged symmetrically around 90 degrees, an equivalent pulse train will be obtained, and the above single-phase output voltage or line-to-line output voltage will be in the range of 180 degrees to 360 degrees. A direct current characterized in that the direct current is controlled to convert the direct current into an alternating current that approximates a sine wave by controlling the direct current to alternating current converter so that the waveform becomes a pulse train equivalent to the polarity inverted pulse train in the range of 0 degrees to 180 degrees. It is related to AC conversion method. According to the present invention, an inverter output is obtained based on a pulse train in which n rectangular wave pulses of the same width are arranged at an interval of an integer multiple of θs = 120 degrees/n, and pulses in the range of 60 degrees to 90 degrees are obtained. An arbitrary number of pulses in the range of 90 degrees to 120 degrees are moved 60 degrees in the 0 degree direction and the polarity is reversed, and an arbitrary number of pulses in the range of 90 degrees to 120 degrees are moved 60 degrees in the 180 degree direction. Since the pulse is inverted in polarity, it is possible to easily obtain an approximate sine wave output with suppressed harmonic components. Next, a DC/AC conversion system according to an embodiment of the present invention will be described with reference to FIGS. 3 to 7. 3, 4, and 7 are examples in which n=20 pulses are arranged in a half cycle to facilitate understanding. Of these, Figure 3A shows the original pulse train, Figure 3B shows the output pulse train from 0° to 180°,
FIG. 4 shows the output pulse train from 180° to 360°. Third
The original pulse train in Figure A is 120° from 30° to 150°.
= 20 rectangular wave pulses P ₁ to P 20 of the same amplitude (height) h and the same pulse width α ₀ at a repetition period of a predetermined angular interval θs of 120°/n divided into ₂₀ equal parts, 30° to They are arranged in a range of 150°. The output pulse train in FIG. 3B is equivalent to changing the pulse arrangement of the original pulse train in FIG. Two pulses P ₂ and P ₄ of P ₁ to _{P 5} are moved from their respective positions to positions symmetrical with respect to 30° without changing the amplitude and width, and within the range of 60° to 90°. The pulse located at P ₆ ~
Pulse P ₈ of P ₁₀ is inverted in polarity without changing its amplitude and width and moved by 60 degrees in the 0 degree direction from its respective position, and pulse P ₁₁ located in the range of 90 degrees to 120 degrees
~Pulse P ₁₃ of P ₁₅ is reversed in polarity without changing its amplitude and width and moved in the direction of 180° from each position.
60 degrees, and pulses P ₁₇ and P ₁₉ of pulses P ₁₆ to _{P 20} located in the range of 120 degrees to 150 degrees are moved symmetrically to 150 degrees from their respective positions without changing the amplitude and width. Move each position, and center around 90°
20 pulses P ₁ to P ₂₀ are arranged symmetrically. The output pulse train in FIG. 4 is a pulse train in the range of 180° to 360° arranged according to the same rule as in FIG. 3B, and corresponds to the pulse train in FIG. 3B with the polarity reversed. Since the above-mentioned output pulse train is an arrangement of rectangular wave pulses with the same pulse width and same amplitude with a regularity that is an integral multiple of θs, the inverter must be controlled to obtain this pulse train at the output stage of the inverter. This can be achieved extremely easily. Furthermore, since the output voltage is adjusted by changing the pulse width of each rectangular wave pulse, the output voltage can be easily adjusted while suppressing fluctuations in harmonic components. Next, the reason why a sine wave approximation output can be obtained by the output pulse trains as shown in FIGS. 3B and 4 will be described. In Fig. 5, which shows a sine wave (sin) in the range of 0° to 90°, the vertical axis is the amplitude of the sine wave, the lateral movement is the phase, and points A to T are provided in the drawing for explanation purposes. _Z = _area DJRSKED=area EKSTLFE _W = _area AHBA When there is a relationship of ₃ = 60° + θ ₀ , the following relationship holds true regardless of the value of θ ₀ : W=x−y=Y−X (1). The following equations (2) to (11) prove that equation (1) holds true. X=∫ ⁶⁰゜₆₀゜_- 〓 ₀ sinθ・dθ = [−cosθ] ⁶⁰゜₆₀ ° _- 〓 ₀ = −cos60゜＋cos (60゜−θ ₀ ) ...(2) Y＝∫ ⁶⁰゜⁺ 〓 ₀ 〓 ₀₌₆₀゜sinθ・dθ = [−cosθ] ⁶⁰゜⁺ 〓 ₀₆₀ ° =−cos (60゜＋θ ₀ )＋cos60゜ ......(3) W=∫〓 ₀₀ sinθ・dθ＝[−cosθ]〓 ₀₀ =-cosθ ₀ +cos0゜=-cosθ ₀ +1 Z=θ ₀ ×1=θ ₀ ...(4) x=Z-X=θ ₀ +cos60゜-cos (60゜-θ ₀ )...(5
) y=Z−Y=θ ₀ +cos(60°+θ ₀ )−cos60°……(6
) From equations (5)(6) and (2)(3), x-y=Y-X=-cos(60°+θ ₀ ) +cos60°+cos60°-cos(60°-θ ₀ ) =2cos60°- cos(60°+θ ₀ ) −cos(60°−θ ₀ )=1−cos(60°+θ ₀ ) −cos(60°−θ ₀ )……(7) According to the cos addition theorem, cos(60°+θ ₀ )＝cos60゜cosθ ₀ −sin60゜sinθ ₀ ……(8) cos(60゜−θ ₀ )＝cos60゜cosθ ₀ +sin60゜sinθ ₀ ……(9) (8)(9) can be converted into (7) Substituting into the equation (7) = 1−cos60゜cosθ ₀ +sin60゜sinθ ₀ −cos60゜cosθ ₀ −sin60゜sniθ ₀ =1−cos60゜cosθ ₀ −cos60゜cosθ ₀ ...(10) cos60゜= Substituting 0.5 into equation (10), equation (10) = 1-cosθ ₀ = equation (4) = W (11). Incidentally, the area of each rectangular wave pulse in the systems shown in FIGS. 3 and 4 is the same. Therefore, it becomes possible to easily make the area of the constant angle range of the sin waveform in FIG. 5 correspond to the total area of the plurality of rectangular wave pulses in FIG. 3. Therefore, if an original pulse train consisting of n positive pulses is arranged between 30° and 150°, and n ₀ pulses exist between D and E and between E and F in Fig. 5, then The relationships n ₀ −n _x /n ₀ =k(z−x/z) and n ₀ −n _y /n ₀ =k(z−x/z) hold. However, n _x is the number of pulses moved from between D and E to obtain an approximate sine wave, n _y is the number of pulses moved from between E and F, and k is 0<k
It is a constant that satisfies ≦1. Therefore, the following relationships (12) and (13) hold true. n _x = n ₀ kx/Z+n ₀ (1-k) ...(12) n _y = n ₀ ky/Z+n ₀ (1-k) ...(13) Therefore, the n ₀ pieces between D and E Out of pulse(12)
Move the n _x pulses obtained from the formula to positions symmetrical to 30 degrees for each pulse, that is, between A and B.
Among the n ₀ pulses between ~F and n _y pulses obtained from equation (13), the polarity of the pulses is reversed and shifted by 60 degrees in the 0 degree direction, that is, between A and B. and,
If the effective number of positive pulses between A and B is _nw , this _nw is expressed by the following equation. n _w = n _x −n _y = n ₀ k (x-y) / Z = n ₀ k (Y-X) / Z = n ₀ k / Zw ... (14) After transferring n _x , D ~ If the number of pulses remaining between E is _nX , this _nX is expressed by the following equation. _{n _ _ _ _ _ _ _ _} ) Furthermore, if n _Y is the number of pulses remaining between E and F after shifting n _y , this n _Y is expressed by the following equation. n _Y = n ₀ k/ZY ...(16) As is clear from equations (14), (15), and (16), n _w , n _x ,
n _Y is proportional to W, X, and Y, respectively. Therefore, the following equation holds. n _W /W=n _X /X=n _Y /Y=n ₀ k/Z=K (constant)...(
17) If the width x height (amplitude) of each pulse, that is, the area of the pulse, is a, then the following equation is established. an _W /W=an _X /X=an _Y /Y=an ₀ k/Z=aK……(18
) Equation (18) shows that if the values _{of n} This means that it is proportional to the area of each sine wave in the same section. Furthermore, equation (18) shows that when n ₀ and the value of a are constant, if the value of the proportionality number K is changed by choosing a different value of k, A to B can be obtained while maintaining the proportional relationship of equation (18). The effective total area of the pulses in each section between D and E, and between E and F changes. This means that the output voltage can be set arbitrarily by selecting the value of k. The above equations (17) and (18) hold true no matter what the value of θ ₀ is. As an example, when θ ₀ =30°=π/6, n=264, and n ₀ =n×30°/120°=66, the values are as follows. Z=π/6×1.0=0.5236 W=0.1340 X=0.366 Y=0.500 x=0.1576 y=0.0236 And when k=1, n _x has the following value. n _x = n ₀ kx/Z+n ₀ (1-k) = 66×1×0.1576/0.
5236 +66(1-1)=19.8655 Similarly, when k=1, _ny , _nW , _nX , and _nY have the following values. _{n y =} 2.9748 _{n W} = n _x −n _y = _19.8655 − _{2.9748 =} 16.8907 n Equation (17) in the case of 1 has the following value.

[Table] The above is theoretically possible, but n _x and n _y must be integers because they represent the number of pulses. Therefore, practically, the values of n _x and n _y may be approximated to the above values. Therefore, if n _x =19.8655≒20 n _y =2.9748≒3, the following function holds true. n _W = 20-3 = 17 (positive 20, negative 3) n _X = 66-20 = 46 n _Y = 66-3 = 63

[Table] Although the proportional coefficients of equation (20) are not completely the same as those of equation (19), they show extremely good approximation in practical terms. Next, when θ ₀ = 30°, which phase within 30° should be selected for n _x = 20 and n _y = 3 can be determined by changing the value of θ ₀ and performing the same calculation as above. It is possible to find n _x and n _y within the value of ₀ . The following Table 1 shows θ ₀ = 30° θ ₀ =
The values of n _x and n _y for 20° θ ₀ = 10° are shown. Table 2 shows the values of n _x and n _y in each 10° interval obtained from Table 1.

【table】

[Table] When k=0.5, the following relationship holds true. n _x =42.9328 n _y 34.4874

[Table] Then, to make n _x and n _y integers, n _x = 43
If n _y =34, the following relationship holds true.

[Table] Comparing equations (20) and (22), the mutual approximation of each proportional coefficient is somewhat worse in the case of equation (22). The value of K is (2
In the case of equation 2), it is half of that in equation (20). This is k
=0.5 means that the output voltage is half of that when k=1. As mentioned above, by transferring n _x and n _y obtained from equations (12) and (13), the effective number of pulses in each section,
Alternatively, the total effective area of the pulse is approximately proportional to the area of the sine wave in that section. For k=1
Table 3 shows how closely the transitional pulse train approximates a sine wave when n _x and n _y are selected based on Table 2. Table 3 shows the voltage of each harmonic up to the 29th when the voltage of the fundamental sine wave is taken as 100%. As is clear from this table, the transition pulse train does not contain even-order harmonics because the pulse arrangement is symmetrical about 90 degrees. In addition, the transition pulse train does not include third harmonics and harmonics multiples of three in the line voltage, whether it is single-phase or multi-phase.

[Table] For comparison, Table 4 shows the harmonic components of the output voltage waveform in the case of the method shown in FIG. In Table 4, since the case is for a three-phase line voltage, third harmonics and harmonics multiples of three are not included, but in the case of a single phase, these are included.

[Table] As is clear from the above, if the pulses are arranged based on the idea of the present invention, the effective total area of the pulses in any phase interval is approximately proportional to the area of the sine wave in that interval. As a result, the waveform of this pulse train approximates a sine wave. In this method, 30°~ before changing the pulse.
The number of pulses in the original pulse train arranged up to 90 degrees affects the approximation to a sine wave of the transitional pulse train after the pulses are shifted. When the number of pulses is small, the approximation is poor, and as the number of pulses increases, the approximation becomes better. Theoretically, if the number of pulses is infinite, it will approximate to a sine wave infinitely if harmonics of infinite order are ignored. The number of pulses to be transferred from among the pulses between 60° and 90° of the original pulse train is smaller than the number of pulses to be transferred among the pulses between 30° and 60°. Therefore, if the number of pulses in the original pulse train is relatively small,
There is no need to transfer pulses between 60° and 90°. In addition, pulses transferred from 30° to 60° and pulses transferred from 60° to
It is possible that pulses shifted from 90° are shifted to the same phase. In this case, since the polarities are opposite, both pulses essentially disappear after the transition. However, in the claims, the number of pulses is expressed from the standpoint that both pulses exist with their height directions opposite. Furthermore, if the width of a pulse is equal to θs, it will come into contact with an adjacent pulse, and multiple pulses will apparently coalesce and be considered as one pulse, but in this case, there are multiple pulses with a pulse width of θs. regarded as. The above matters are the same even between 90° and 180°, which are completely symmetrical with respect to 90°. Figure 6 is a circuit diagram of this method, and Figure 7 is a circuit diagram of this method.
It is a diagram showing the state of each part in the diagram. In FIG. 6, 1 is an integral voltage input terminal, which in this embodiment is a terminal that supplies a DC voltage corresponding to the input voltage of the inverter. A voltage supplied from an input terminal 1 is sent to a sawtooth wave generation circuit 2, and then supplied to an integrating capacitor 4 via an operational amplifier 3, where it is integrated. Since a discharging transistor 5 is connected in parallel to the capacitor 4, the sawtooth wave V _A shown in FIG. 7 is generated in synchronization with the on/off period of the transistor 5. 6 is an output frequency instruction voltage input terminal, and the instruction voltage supplied from here is V-
The signal is converted into a frequency signal by the F converter 7, and pulses are generated at a period corresponding to the designated frequency. V-
The output pulse of the F converter 7 is input to the counter 8 as a clock pulse and is also input to the base of the transistor 5. In the example of FIG. 7, 30 pulses are generated at a period of θs during the period from 0° to 180°, and 30 sawtooth waves V _A are generated. A voltage comparator 9 generates a comparison output between the control voltage V _R applied from the control voltage input terminal 10 and the sawtooth wave V _A. As is clear from V _P in FIG. 7, this embodiment generates a comparison output V _P that is at a low level during a period in which the sawtooth wave V _A is lower than the control voltage _VR . That is, the circumferential base θs is from 0° to 180° (t ₀ to t ₁ ) in Fig. 7.
30 square wave pulse outputs V _P are generated. 11 is a memory, based on the idea of the present invention, each line voltage of the output lines U, V, W of the inverter is stored.
Data for controlling the transistors Q ₁ to _{Q 6} of the inverter is stored so that V _UV , V _VW , and V _WU become the pulse train shown in FIG. In this example, as is clear from FIG. 7, the memory 11 has a total of 60 addresses from address 0 to address 59, and a counter reset address 60 provided next to address 59, where the reset signal is stored. The moment the address is switched to 60, a reset signal for the counter 8 is generated from line _M3 , and the address is switched to 0. Note that each address 0 to 60 stores 9 bits of data, and this data is read out through 9 output lines _M1 to _M9 corresponding to the 9 bits. Note that data is read from the memory 11 using the same clock as the period θs of the sawtooth wave V _A. That is, it is performed sequentially based on the address designation of the counter 8. 12 is a distribution logic circuit for forming a control signal for a three-phase inverter, and includes eight AND gates.
It consists of A ₁ to _{A 8} , six OR gates B ₁ to B ₆ , and one inverter C ₁ . Note that the lower input terminal of the AND gate _A8 is an inverting input terminal. Also, memory output lines M ₁ to _{M 6} are connected to AND gate A _1
-A6 , and the output terminal of an inverter _C1 for inverting the output of the comparator 9 is coupled to the other input terminal of the AND gates _A1 - _A6 . In addition, the outputs of AND gates A ₁ to A ₆ are connected to transistors Q 1 to Q ₁ through OR gates B ₁ to _{B 6} .
Combined with the base of Q ₆ . One input terminal of AND gate A ₇ is coupled to the output terminal of comparator 9, and the other input terminal is connected to memory output line M ₇
and its output terminal is coupled to the input terminals of OR gates B ₁ -B ₃ . One input terminal of AND gate A ₈ is coupled to the output terminal of comparator 9, and the other inverting input terminal is connected to the memory output line
_M7 , and the output terminals are coupled to OR gates _B4 - _B6 . At each address in the memory 11, data is written in H and L as shown in the columns M ₁ to M ₉ in FIG. Read out sequentially. And the memory output line
Generate control signals for transistors Q ₁ to _{Q 6} based on the outputs of M ₁ to M ₆ and the output V _P of the comparator,
During the period when the control signal is at a high level, transistor Q ₁ ~
Q ₆ turns on. In addition, in FIG. 7, transistors Q ₁ ,
The control signals for Q ₃ and Q ₅ are shown, and the control signals for Q ₂ , Q ₄ , and Q ₆ are not shown independently, but Q ₂ has the opposite action to Q ₁ , Q ₄ has the opposite action to Q ₃ , Q ₆ has the opposite action to Q ₅ . When the transistors Q ₁ to Q ₆ are controlled by the control signals shown in FIG. 7, the output voltage waveforms between the output lines U, V, and W are V _UV , V _VW , V _WU in FIG.
becomes. That is, the voltage waveform between each line becomes the same as the waveform shown in FIG. Note that the output line _M8 of the memory 11 is the counter 8.
When the reset signal connected to the reset terminal of the counter 8 and written to the additional address 60 added after the final address 59 is read out
Reset. As a result, data is read from the first address 0 again. The output line M ₉ of the memory 11 is coupled to the base of a transistor 14 connected in parallel to the resistor 13 of the control voltage input terminal 10 line. Therefore, when the outputs shown in FIG. 7 are sequentially generated from the memory output line _M9 , the transistor 14 is turned on and off accordingly, and the value of the control voltage _VR changes. Therefore,
It becomes possible to change the width of the output pulse at a specific position. Note that when there is no need to change the control voltage _VR , the value of the memory output line _M8 is always kept constant. As is clear from the above, according to this example,
The following advantages are obtained: (a) Since the rectangular wave output pulses are arranged at intervals that are integral multiples of θs, switching control of the inverter to obtain the output pulses is facilitated. (b) Since a sine wave approximation output is obtained by arranging multiple pulses with substantially the same area, the effective total area of multiple pulses in an arbitrary phase interval can be calculated as the area of the sine wave in that interval. Approximation becomes easy, and a sine wave approximation output can be easily obtained. (c) The rising phase of each output pulse is always constant regardless of the output voltage and output frequency, and the rising interval of each output pulse is an integral multiple of θs, and the rising phase and pulse interval are fixed values determined in advance. It is. For this reason, control of the rising phase of the pulse can be realized extremely easily using extremely universal digital technology or analog technology, and furthermore, control of the pulse width after the rising edge is also easy, especially for multiphase output. In this case, the effect is large. (d) In this embodiment, since the input terminal 1 is configured to supply the DC voltage of the power source E of the inverter, it is possible to limit fluctuations in the inverter output voltage due to fluctuations in the inverter input voltage. That is, when the input voltage decreases, for example, the input voltage of the integrating capacitor 4 also decreases, and the slope of the sawtooth wave V _A shown in FIG. 7 becomes gentler. As a result, the pulse width of the inverter output voltage becomes wider. However, since the input voltage has decreased and the amplitude of the output pulse has become smaller, the area of the output pulse remains approximately constant, and there is substantially no variation in the output voltage or variation in the harmonic components. Such an operation also occurs when the inverter power supply E supplies a pulsating voltage. Therefore, there is a great advantage that voltage fluctuations including input voltage pulsations do not appear on the output side. (e) Since the width of a specific pulse can be controlled by switching the level of the control voltage V _R using the memory output line _M9 , control that improves the approximation to a sine wave can be easily achieved. Although the embodiments of the present invention have been described above, the present invention is not limited thereto and can be further modified. For example, in FIG. 6, the voltage _VR supplied from the control voltage input terminal 10 to the comparator 9 may be set as a constant reference voltage, and the voltage supplied from the input terminal 1 may be set as the output voltage command voltage. That is, the width of the inverter output pulse may be changed by changing the slope of the sawtooth wave V _A shown in FIG. 7 according to the output voltage. Alternatively, the voltage at the terminal 1 or 10 may be supplied based on the detection of the output voltage of the inverter, and closed loop control may be performed. Moreover, it is of course applicable to a single-phase inverter. It is also applicable to inverters that use switching elements such as thyristors.

[Brief explanation of the drawing]

FIG. 1 is a circuit diagram showing an inverter, FIG. 2 is a waveform diagram showing a conventional PWM control method, and FIG. 3 is a waveform diagram for explaining the output voltage of an inverter according to an embodiment of the present invention. is the original pulse train, B
is a waveform diagram showing an output pulse train, FIG. 4 is a waveform diagram showing an output pulse train in an interval of 180° to 360° according to an embodiment of the present invention, and FIG. FIG. 6 is a circuit diagram showing an inverter according to an embodiment of the present invention, and FIG.
This figure is a waveform diagram showing the states of each part in FIG. 6. In the symbols used in the drawings, 1 is an integral voltage input terminal, 2 is a sawtooth wave generation circuit, 9 is a comparator, 10 is a control voltage input terminal, 11 is a memory, 12 is a logic circuit, Q ₁ to Q ₆ is a transistor.

Claims (1)

[Claims] 1 The waveform in the range of 0 degrees to 180 degrees of the single-phase output voltage or multi-phase line-to-line output voltage of a single-phase or multi-phase DC/AC converter is set at a predetermined angular interval of 120 degrees/n (where n is an integer). In the original pulse train in which n rectangular wave pulses having the same pulse width not larger than θs are arranged symmetrically about 90 degrees in the range of 30 degrees to 150 degrees with a repetition period of the predetermined angular interval θs, 30 degrees~
Any one or more of the pulses located within a range of 60 degrees are moved from each position to a position symmetrical with respect to 30 degrees without changing the amplitude and pulse width, and the Any one or more pulses within the range of pulses can be moved 60 degrees in the 0 degree direction from each position by reversing the polarity without changing the amplitude and pulse width, and moving the pulses within the range of 90 degrees to 120 degrees. An arbitrary number of pulses among the pulses located at
Move any one or more of the pulses located within a range of 150 degrees from each position to a position symmetrical with respect to 150 degrees without changing the amplitude and pulse width, and move 90 degrees. If n pulses are arranged symmetrically around the center, it becomes an equivalent pulse train, and the waveform of the single-phase output voltage or line output voltage in the range of 180 degrees to 360 degrees corresponds to the pulse train in the range of 0 degrees to 180 degrees. A DC/AC conversion method, characterized in that the DC/AC converter is controlled to convert a DC into an AC that approximates a sine wave by controlling the DC/AC converter to produce a pulse train equivalent to a pulse train with reversed polarity.