JPH0568950B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a power conversion device that converts direct current power to low frequency alternating current power approximately equal to a commercial frequency.

[Technical background of the invention and its problems] Figure 6 shows an example of the main circuit configuration of a power conversion device to which the present invention can be applied, which converts DC power into AC power at a commercial frequency and connects to an AC system. .

This device converts a DC power source 1 into high frequency AC (for example, 20 KHz) via a high frequency inverter 2. After that, the power is transmitted via a high frequency transformer 3 that insulates the DC power supply side and the AC power supply side, and once again converted to DC by a rectifier 4, and then passed through an energy storage element (a reactor in this figure) 5. Then, the low frequency inverter 6 for connecting to the AC system converts the power into AC power at a commercial frequency (for example, 60 KHz), and connects it to the AC system 7.

Generally, between the DC power supply 1 and the AC system 7,
Due to problems such as ground fault protection and voltage resistance of the switching elements 21, 22, 23, and 24 in the inverter 2,
Requires transformer 3 for insulation. At that time, if the inverter 2 converts the DC power directly into commercial frequency AC power instead of high frequency AC power,
The rectifier 4 and commercial power supply interconnection inverter 6 shown in FIG. 6 become unnecessary. (However, reactor 5 cannot be omitted even in this case.) In this way,
In the case of direct conversion to commercial frequency alternating current, the isolation transformer 3 must transmit low frequency power of about the commercial frequency, so compared to the case of transmitting high frequency power as in the power converter shown in Fig. 6. , the size of the transformer increases significantly. Therefore, the combination of the high-frequency transformer 3, the rectifier 4, and the low-frequency inverter 6 for connecting to the AC grid can be made more compact as a whole than the commercial frequency transformer.
Also, in terms of efficiency, a high-frequency transformer can have lower loss than a low-frequency transformer, so even if you include the loss of the rectifier and interconnection inverter, it is possible to improve the overall efficiency of the entire device. be.

However, the problem here is saturation of the high frequency transformer 3. In other words, the more the device is made smaller and more efficient, the more the saturation problem becomes more prominent. Therefore, unless some measure is taken, there is a possibility that an accident such as destruction of the switching elements 21 to 24 or burnout of the transformer 3 may occur.

[Object of the Invention] The present invention has been made in consideration of the above-mentioned circumstances, and it is possible to control the switching element 2 in the high frequency inverter 2 while controlling the desired current to flow in the AC system.
1, 22, 23, and 24, the magnetic flux generated by the transformer 3 is controlled, and on the other hand, due to the accumulation of control errors, the transformer 3
The purpose of the present invention is to provide a power conversion device that prevents the transformer from becoming completely saturated by quickly resetting the control circuit and returning to normal control when the transformer begins to become saturated. It is something to do.

[Summary of the Invention] To achieve this object, the present invention converts DC power into high-frequency AC through a high-frequency inverter, and after transmitting the power through a high-frequency transformer,
Once converted to DC power through a rectifier, it is converted back to commercial frequency AC power using an energy storage element and an inverter for connecting to a commercial power supply.The power converter connected to the commercial power supply converts the sinusoidal power supply current to the commercial power supply. The switching element of the high-frequency inverter is turned on and off by a signal from a voltage pattern generation circuit that causes the high-frequency transformer to flow, and the on-time of the switching element is controlled according to the polarity of the voltage applied to the high-frequency transformer. , a magnetic flux control circuit is provided that indirectly controls the magnetic flux generated within the high-frequency transformer, and on the other hand, when the high-frequency transformer begins to saturate due to accumulation of control errors in the control circuit, a circuit that immediately detects saturation is provided. The saturation of the high frequency transformer is prevented by resetting the magnetic flux control circuit using the signal from the saturation detection circuit, quickly operating in the direction of desaturation, and returning to normal control.

[Embodiments of the Invention] First, the principle of the present invention will be explained with reference to FIG.

In order to make the power supply current I _U in phase with the power supply voltage V _AC (that is, power factor = 1), switching elements 61, 62, 63,
64 is controlled at a phase angle α=0°. In other words, the power supply voltage
Switching elements (thyristor in the figure) 61, 64 or 62, 63 at the timing when V _AC becomes 0
Turn it on with a combination of Therefore, the voltage V _INV on the input side of the inverter 6 is
Depending on the conduction state of 2, 63, and 64, the commercial power supply has a full-wave rectified voltage waveform. On the other hand, in the high frequency inverter 2, the high speed switching elements 21,
Of 22, 23, 24, 21, 24 or 22,
When the combination 23 is turned on, a DC voltage V _DC is applied to the primary side of the high frequency transformer 3 (the high frequency inverter side is the primary side and the rectifier side is the secondary side).
A voltage V ₂ therefore appears on the secondary side. In other words, assuming the turns ratio of the transformer is n ₁ :n ₂ , when switching elements 21 and 24 are on, V ₂ = n ₂ / n ₁ V _DC . When switching elements 22 and 23 are on, V ₂ = n ₂ / n ₁ V _DC , etc. In the case of V ₂ ≒ 0 or no excitation. Therefore, the output voltage V _REC of the rectifier 4 is V _REC = n ₂ /n ₁ V _DC when the switching elements 21 and 24 are on or 22 and 23 are on, and V _REC ≈0 in other cases. Therefore, the voltage applied to reactor 5 is
Since V _REC −V _INV , switching elements 21 and 2
4 or a combination of 22 and 23, _n2 / _n1 V _DC -V _INV ...(A) In other cases, -V _INV ...(B). Therefore, if the turns ratio n ₁ :n ₂ of the transformer is determined so that n ₂ /n ₁ V _DC > (maximum value of V _INV ),
The current I _L flowing through the reactor 5 increases in case (A) _{and decreases} in case (B).

What should be noted here is that in either combination of high-speed switching elements 21, 24 or 22, 23, as a result, the current flowing through the reactor
It is possible to increase _IL . Therefore, when considering the saturation of the transformer, in the period when I _L must be increased, 21, 24 or 2
All that is needed is to control the amount of time that 2 and 23 are on in a well-balanced manner.

An embodiment of the present invention will be described below with reference to FIG. This figure has the same main circuit configuration as in Figure 6, and controls the power supply current I _U (actor current I _L ) in the figure so that it becomes a predetermined sinusoidal current, and the high-frequency transformer 3. High-speed switching element 2 in high frequency inverter 2 to avoid saturation.
1, 22, 23, and 24. FIG.

The switching pattern generation circuit 100 outputs switching signals SW14 and SW23 of the switching elements in the high frequency inverter 2 in order to make the power supply current _IU a sinusoidal current. To do this, first, the voltage pattern VPAT to be generated as the output V _REC of the rectifier 4 is output from the voltage pattern generation circuit. This voltage pattern generation circuit is a circuit that generates an on/off pattern, that is, a voltage pattern VPAT, by comparing it with a comparator so that the controlled current exists near a certain current reference, as is usually done. direct current comparison type pulse width modulation control),
A circuit that generates the voltage pattern VPAT to be generated as V _REC by comparing a triangular wave with a voltage reference that causes a predetermined current to flow (triangular wave comparison type pulse width modulation control),
Alternatively, a circuit or the like may be used in which a voltage pattern is prepared in advance by calculation or the like. The point to note here is that the generated voltage pattern VPAT is
This is a good point because it can be used in any pattern to control the power supply current _IU , that is, the reactor current _IL , without being particularly concerned with transformer saturation. However, this is not amplitude modulation control but pulse width modulation control. Thus, the generated voltage VPAT may include wide pulses.

The voltage pattern signal VPAT generated by the voltage pattern generation circuit is input to the switching element selection circuit. This circuit selects the combination of switching elements 21, 24 or 22, 23 to be turned on next. For example, when 21 and 24 are to be turned on, the 21 and 24 selection signal SEL14 is output to the delay circuit for preventing vertical short circuit. Therefore, for example, after a delay of several μSEC, the 21, 24 ON signals SW14 are output to the switching element drive circuit, and the switching elements 21, 24 receive the drive signals D1, D4, respectively.
is output. Then, the elements 21 and 24 become conductive, and the DC voltage V _DC is applied to the primary side of the high frequency transformer 3. As a result, a voltage appears at V _REC and the reactor current I _L increases. On the other hand, switching element 2
The 1, 24 ON signal SW14 and the 22, 23 ON signal SW23 are also output to the magnetic flux control circuit 100. Therefore, first, SW14 and SW23 are input to a polarized total conduction time calculation circuit. There, SW
14 or SW23 signal is output with a sign. In other words, the time during which SW14 is output is positive, and the time during which SW23 is output is negative, and the sum is taken. The result is the polarized total conduction time TN. This TN is a high frequency transformer 3
It can be seen as a simulated signal of the magnetic flux generated within the magnetic flux, that is, the generated magnetic flux level. Therefore, in the present invention, saturation of the magnetic flux of the high frequency transformer 3 is prevented or controlled by controlling the polarized overall conduction time TN not to exceed a certain level. This time data TN is then output to the area discrimination circuit. Upper limit value ULMT with TN
If it exceeds the upper limit, an upper limit over signal ULV is output. Also, if it falls below a certain lower limit DLMT,
A lower limit over signal DLV is output. Moreover, when the position is within these limit values, the switching signal ALT is outputted alternately. Those signals (UL
oV, DLV, ALT) are output to the switching element selection circuit, and the switching element sets 21, 24, 22, 23 are selected based on the signals.

The above situation will be further explained based on the timing chart shown in FIG. Consider a case where a voltage pattern signal VPAT as shown in the figure is output from a voltage pattern generation circuit. As mentioned above, this VPAT is
The relationship is almost similar to the output voltage V _REC of the rectifier 4. The period before the time point indicated by A in the diagram is
Since the polarized total conduction time TN (that is, the generated magnetic flux simulation signal) is within the upper limit ULMT and lower limit DLMT, the respective limit over-signals ULV and DL
V is not output, and the alternate switching signal ALT is output. During the period when ALT is being outputted (that is, "H"), SW14 and SW23 are alternately outputted every time VPAT is outputted as shown in the figure. Here, which pair was in the conductive state last time is determined by the signal from the switching element memory circuit.
It is based on MEM14 and MEM23. As a result, positive and negative voltages are alternately applied to the high frequency transformer 3. However, SW14, SW23
The on time of depends on the on time of VPAT, so
Not necessarily the same. Therefore, the sum of T
The level of oN also gradually increases (or decreases),
As a result, the upper limit (or lower limit) is exceeded. FIG. 2 shows a case where the upper limit value is exceeded at point A. At this point, the upper limit over signal
ULV is output, and at the same time (actually, after a short period of time has elapsed to prevent a vertical short circuit, but it is omitted in the figure), the SW14 signal is reset and the SW23 signal is output, so that the high frequency inverter 2 The conduction state changes from the set 21 and 24 to the set 22 and 23, and the high frequency transformer 3 has
A voltage of positive to negative polarity is applied. On the other hand, the secondary voltage V ₂ is rectified by the rectifier 4, so
A voltage appears at V _REC so that the reactor current can continue to increase. Thereafter, TN decreases, and when the limit value over signal set within the upper and lower limits reaches the reset level RSTV (in the figure, it is set at the center, but it can be set arbitrarily) (in Figure 2, it is Point A') Reset the ULV signal. As shown in the figure, only SW23 is output during the period when ULV is output.
Thereafter, the ALT signal is restored, and after point A', alternate switching is repeated as before point A.

By controlling as above, the polarized total conduction time TN, that is, the generated magnetic flux simulation value falls within the upper limit (ULMT) and lower limit (DLMT) range,
Transformer saturation can therefore be prevented in the ideal case. However, in practice, the turn-on time of switching elements, variations in turn-on time, the effects of transient phenomena (vibration, etc.) during switching, temporal fluctuations in DC voltage, or accumulation of control errors (quantization errors in the case of digital control, analog In the case of control, errors due to drift, etc.) may cause the transformer to gradually move toward saturation, even though the magnetic flux control circuit is performing normal control. As a countermeasure, a saturation detection circuit is provided as shown in Figure 1, and the transformer saturation signal SAT output from the circuit is used to detect
The magnetic flux control circuit is reset, so to speak. Therefore, this saturation detection circuit will be explained.
Generally, when a transformer begins to saturate, the excitation current increases and the secondary voltage decreases. Figure 3 shows qualitative voltage and current waveforms. Therefore, in order to detect the saturation of the transformer, by monitoring the secondary voltage V ₂ or the primary current I ₁ , we detect a decrease in the voltage in the case of V ₂ and an increase in the excitation current in I ₁ in the case of I ₁ . Just detect it. However, in order to accurately detect an increase in excitation current, it is necessary to accurately and quickly detect I ₁ current, which is generally more difficult than detecting V ₂ . In addition, a magnetic flux sensor is embedded in the transformer, and B-
Another possible method is to directly monitor points on the H characteristic to detect the saturation point.

FIG. 4 shows an embodiment of the saturation detection circuit when the secondary voltage V ₂ is monitored. The signal V' ₂ for checking the level of whether the secondary voltage is saturated uses the result of dividing the secondary voltage V ₂ or the detection tertiary winding voltage V ₃ by the rectified DC voltage V _DC . This purpose is V _DC
This is to normalize the signal _V'2 to be level checked, taking into account temporal fluctuations in the signal V'2. accordingly
It can be omitted if V _DC is constant, that is, the peak of V ₁ is constant. V' ₂ is compared with the saturation detection level SATLBL, and when V' ₂ <SATLBL, the saturation detection signal SAT is output to the polarized total conduction time calculation circuit shown in FIG.

In the polarized total conduction time calculation circuit, the saturation detection signal
When SAT is received, it is checked which of the combination of switching elements 21, 24, 22, and 23 is conducting from the signal of SW14 or SW23 (or MEM14, MEM23), and the polarity of the voltage applied to the transformer is determined. to judge. Then, from the saturation correction data transfer circuit, for SW14 - polarity saturation correction level SACP (<ULMT) and for SW23 + polarity saturation correction level SACM (>DLMT) is read, and the total conduction time with polarity TN is forcibly set. Set to . This is when detecting saturation.
This is a reset mechanism for the saturation control circuit. After that, the saturation control circuit uses the newly set value of TN.
It operates to return to within the range of ULMT and DLMT, and saturation is quickly released.

The above situation will be explained based on the time chart shown in FIG. When the saturation signal SAT occurs at time B, in this case, SW14 outputs (elements 21, 24
is on), so the transformer is saturated with + polarity, and as a result, +
A polarity saturation correction level SACP is set. The area discrimination circuit outputs the upper limit over signal ULV together with the value of TN as an alternate switching signal.
Reset ALT. As a result, the switching element selection circuit always outputs SEL23 during the period when ULV is output, thereby making the switching elements 22 and 23 conductive and causing the transformer to -
A polar voltage is applied to release the saturation.

Of course, the values of the upper limit ULMT and lower limit DLMT,
It is also possible to perform control by appropriately increasing or decreasing it in accordance with the level of the DC voltage V _DC at that time.

[Effects of the Invention] As described above, according to the present invention, the switching elements of the high-frequency transformer are controlled to turn on and off so as to prevent the saturation of the high-frequency transformer. Even when the transformer begins to saturate, it is quickly operated in the direction of desaturation, and even in that case, the output current of the power converter is continuously controlled to a desired value. Therefore, it becomes possible to design the high frequency transformer at its limits, and the transformer can be made smaller and more efficient, and as a result, the power converter can be made smaller and more efficient.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention;
Fig. 2 is a timing chart for explaining the operation of Fig. 1, Fig. 3 is a qualitative waveform diagram of the voltage and current of a general transformer, and Fig. 4 is a block diagram showing an example of the saturation detection circuit of Fig. 1. 5 is a timing chart for explaining the operation of the present invention, and FIG. 6 is a configuration diagram of a power conversion device to which the present invention is applied. 1...DC power supply, 2...High frequency inverter, 3
...High frequency transformer, 21, 22, 23, 24...
...High-speed switching element, 4...Full-wave rectifier, 7
... Commercial power supply, TN ... Total conduction time with polarity,
ULMT...Upper limit value of conduction time, DLMT...Lower limit value of conduction time, SAT...Transformer saturation signal, SACP
... + polarity saturation correction level, SACM ... - polarity saturation correction level, 100 ... switching pattern generation circuit, 101 ... magnetic flux control circuit.

Claims (1)

[Claims] 1. A high-frequency inverter that converts direct current into high-frequency alternating current, a high-frequency transformer to which the high-frequency alternating current is supplied, a rectifier that converts the high-frequency alternating current isolated by the high-frequency transformer into a second direct current, and a rectifier that converts the second direct current into a predetermined direct current. In a power converter equipped with an inverter for connecting to an alternating current system, which converts the frequency into alternating current and supplies it to an alternating current system, a switch that generates a switch signal that causes the high frequency inverter to output a positive or negative voltage by pulse width modulation control. a switching pattern generating means, and a total conduction command time of the switch signal,
By assigning a positive sign to the conduction command time corresponding to the positive output voltage of the high frequency inverter, and assigning a negative sign to the conduction command time corresponding to the negative output voltage, and taking the sum of the two. magnetic flux simulating means for simulating the magnetic flux of the high frequency transformer and outputting a command to reverse the polarity of the output voltage of the high frequency inverter when the simulated magnetic flux exceeds a predetermined magnetic flux level on the positive or negative side; saturation detection means for detecting an increase in primary current and a decrease in secondary voltage, or saturation of the high frequency transformer using a magnetic flux sensor; A power conversion device comprising: a correction means for replacing the simulated magnetic flux with a value exceeding the predetermined magnetic flux level on the positive side or the negative side.