JPH07322497A - Current-type active filter for power distribution - Google Patents

Abstract

PURPOSE:To achieve an accurate feedback control when controlling reactor current. CONSTITUTION:High-pass filters (HPF) 27-29 of each phase input currents icuw, icvw and icwu where the fundamental wave component of each phase is eliminated from load current iLi, (i=u, v, w) from a power distribution line. A peak-hold circuit 30 retains only a maximum value icmax out of icuw, icvw, and icwu and inputs it to an operator 31. The peak-hold circuit 30 inputs a reset signal from an oscillation circuit 34 synchronized to the cycle of a load current and updates the maximum value icmax for each cycle. The operator 31 multiplies the maximum value icmax by a coefficient and inputs the result to a hold circuit 33 via a switching circuit 32. The switching circuit 32 performs on/off operations based on a drive signal input for each cycle from the oscillation circuit 34. The hold circuit 33 retains a setting value id*. The reactor current is subjected to feedback control based on the setting value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a power distribution active filter, and more particularly to a current source active filter.

[0002]

2. Description of the Related Art In recent years, power semiconductor switching elements such as diodes and thyristors are often used by directly connecting them to a power source, and a large amount of harmonic current is generated due to these elements, so that a power system and peripherals. Is causing various obstacles to the equipment.

The active power distribution filter compensates a harmonic current generated by a load of an electric device or the like using the power semiconductor element as described above, and allows a sinusoidal current to flow in the system. .

As the active filter for distribution, a voltage type active filter and a current type active filter are known. The current source active filter controls the current of the DC reactor (reactor current) as a DC current source to a predetermined value, and also turns on / off the power switching elements that make up the power converter to change the polarity of the current. , Is a filter that creates the same current waveform as the harmonic current. On the other hand, the voltage-type active filter is a type of filter that produces the same current waveform as the harmonic current by charging / discharging a capacitor as a DC voltage source and turning on / off a power switching element.

Compared with a voltage type active filter, a current type active filter can sufficiently cope with a low speed switching element such as a bipolar transistor, a control circuit and a control calculation of a compensation current are simple, and an operation of a current source is essential. Therefore, it has advantages such as less interference with the system (resonance phenomenon, etc.) and is expected as an active filter for power distribution.

In this current source active filter, the compensation current to be compensated by the active filter (command value to the inverter) is the fundamental current calculated in the circuit from the load current detected by the current transformer provided in the distribution line. It is obtained by subtracting. Then, the active filter compares the compensation current (harmonic waveform) with a triangular waveform called a carrier wave to generate a drive signal for an inverter for turning on / off the switching element of the power converter.

In the conventional current source active filter, the reactor current is feedback-controlled to a predetermined value (set value), but this set value is a fixed value.

[0008]

However, since the fault current changes from time to time, when feedback controlling the reactor current, if it is a fixed value, there is a problem that accurate feedback control cannot be performed. Further, even if the compensation current is small, it is necessary to flow a current corresponding to the rated capacity to the reactor, and there is a problem that loss such as copper loss is large.

An object of the present invention is to solve the problems of the prior art and to provide a current source active filter capable of performing accurate feedback control.

[0010]

In order to achieve the above object, the invention of claim 1 detects a load current in which a fault current generated from a load is superimposed on a fundamental wave current flowing in a distribution line. Detecting means, calculating means for calculating a fault current based on the detection result of the detecting means, and a plurality of switching elements, and on / off of the switching element by a pulse width modulation method based on the calculating result of the calculating means In the active filter for distribution, which includes a plurality of compensation current generating means for generating a compensation current equivalent to the fault current by performing
A filter means that removes the fundamental wave current of each phase from the load current of each phase and passes the residual component of each phase, and the value of the maximum residual component of the residual components of each phase input from the filter means. A setting unit including a setting unit that determines a setting value based on the setting value, a feedback unit that calculates a difference from the setting value determined by the setting unit and the compensation current generated by the compensation current generating unit, and a load current of each phase. Extracting the positive phase component of each phase from the first phase, the first calculating means for calculating the corrected positive phase component based on the positive phase component and the difference from the feedback unit, and the load current of each phase from each phase A second calculation means for calculating a negative phase component of the above, and a corrected positive phase component of each of the phases, which is the calculation result of the first calculation means and the second calculation means, a negative phase component of the each phase, and a load current. Fault current performance to calculate the fault current of each phase based on That an arithmetic unit and means are the subject matter.

According to a second aspect of the present invention, in addition to the configuration of the first aspect, the compensation current generating means uses a sawtooth wave as a carrier wave used in the pulse width modulation method, and the sawtooth wave and the fault current waveform are large or small. The gist is that the compensation current is output by turning on / off the switching element based on the comparison.

The invention according to claim 3 is claim 1 or claim 2.
In addition to the above configuration, the gist is that it is connected to a distribution line through a band elimination filter. In addition to the structure of claim 3, the invention of claim 4 is characterized in that the band elimination filter is an LC filter.

[0013]

According to the invention of claim 1, the detecting means detects the load current in which the fault current generated from the load is superimposed on the fundamental wave current flowing through the distribution line.

In the calculating means, the filter means of the setting section removes the fundamental wave current of each phase from the load current of each phase and passes the residual component of each phase. The setting means determines the set value based on the value of the maximum phase residual component of the residual components of each phase input from the filter means. The feedback unit calculates the difference from the set value determined by the setting unit and the reactor current generated by the compensation current generating means.

Further, in the calculation means, the first calculation means of the calculation section extracts the positive phase component of each phase from the load current of each phase and calculates the positive phase component and the difference from the feedback section. Based on this, the corrected positive phase component is calculated. The second calculation means calculates the negative phase component of each phase from the load current of each phase. The fault current calculation means determines the fault current of each phase based on the corrected positive phase component of each phase, the negative phase component of each phase, and the load current, which are the calculation results of the first calculation means and the second calculation means. Calculate.

The compensating current generating means generates a compensating current equivalent to the fault current by turning on / off the switching element by the pulse width modulation method based on the calculation result of the calculating means, and supplies the compensating current to the distribution line. Supply.

According to the invention of claim 2, in addition to the operation of claim 1, the compensation current generating means uses a sawtooth wave as a carrier wave used in the pulse width modulation method, and the sawtooth wave and the fault current waveform are large or small. The switching element is turned on / off based on the comparison.

The invention of claim 3 is claim 1 or claim 2.
In addition to the above function, since it is connected to the distribution line through the band elimination filter, the harmonics caused by the pulse width modulation control are removed by the band elimination filter and the compensation current is output to the distribution line.

According to the fourth aspect of the present invention, the LC filter removes the harmonics caused by the pulse width modulation control by the band elimination filter.

[0020]

Embodiments of the present invention will be described below with reference to the drawings. As shown in FIG. 1, a load 2 is connected to a three-phase distribution line 1, and a current source active filter (hereinafter, simply referred to as active filter) 3 is connected to the load 2 in parallel.

The active filter 3 will be described in detail. The active filter 3 is a detection means and a fault current detection device 4 as a calculation means, a compensation current generation device 5 as a compensation current generation means, and an L as a band elimination filter.
And a C filter 6. The fault current detector 4 is connected to each phase of the distribution line 1 via a current transformer CT, and inputs a load current flowing in each phase.

As shown in FIG.
System current i that is composed of a subsidiary circuit and flows from the system_su，
i_sv, I_sw(Commercial frequency 60 Hz in this embodiment)
Folded fault current i_i ^*(_i=_{u, v, w}) To detect
It has become. Fault current i_i ^*Is a power semiconductor device
It is the harmonic current generated by the load 2 such as the electric device.

First, the principle of detecting the fault current i _i ^* ( _i = _{u, v, w} ) will be described. First, if the phase of the load current _{i Li (i = u, v} , w) and detects the load current i _Li, may be taken out therefrom harmonic current i _hi. At this time, the compensation current to be compensated by the active filter 3 is i
_hi = i _i ^* (also referred to as a command value to the converters A and B described later). Then, the fundamental wave component i _il is obtained from the load current i _{Li so} that i _i ^* = i _Li −i _il ( _i = _{u, v, w} ).

The phase order of the phase voltages v _{u to} v _w of the distribution line 1 is u,
As v and w equilibrium voltage, v _u = (2/3) ^1/2 V _l cosωt v _v = (2/3) ^1/2 V _l cos (ωt-2π / 3) v _w = (2/3) ^{It is set to 1/2} V _l cos (ωt + 2π / 3). On the other hand, load current i _Li ( _i = _{u, v, w} )
It can be divided into opposite phase and higher harmonic components.

[0025]

[Equation 1]

I _1P is a component in phase with the voltage of the positive phase current, I _1Q
Is a component delayed by 90 °. The same _{applies to the} negative phase currents I _2P and I _2Q . In the above equation, the load current i _Li ( _i =
_{When the} coordinate conversion is performed on _{u, v, w} ), the following equation is obtained. That is, the positive-phase current is converted into direct current, the negative-phase current is converted into alternating current (120 Hz) having an angular frequency of 2ω, and the harmonic current is converted into alternating current of higher frequency. The current after this coordinate conversion is I _1P and I using a low pass filter (LPF).
DC component of _1Q (i ₁ γ ^# ≈ I _1P And i ₁ δ ^# ≈ I _1Q )
Just take out. To obtain the _{antiphase components} I _2P and I _2Q from the load current i _Li ( _i = _{u, v, w} ), the load current i _Li ( _i = _{u,
v, w} ) is coordinate-converted by C ₂ ^T and passed through an LPF to obtain a direct current component (i ₂ γ ^# ≈I _2P and i ₂ δ ^{# at} this time).
≉I _2Q ).

[0027]

[Equation 2]

FIG. 3 shows an electronic circuit diagram of the fault current detecting device 4 for realizing the above principle. In FIG. 3, I _d
Represents the reactor current of the converter to be described later, and I _d ^* is its set value. The setting value I _d ^* will be described later. This reactor current I _d and the set value I _d ^*
The deviation is calculated by the first adder 11 as a feedback unit, the deviation is multiplied by a predetermined coefficient by the coefficient unit 12, and the calculation result is further time-integrated by the integrator 13. The integrated value which is the calculation result of the integrator 13 is input to the second adder 14.

The first coordinate conversion circuit 15 uses the load current i_Li(
_i=_{u, v, w}) Is input, the coordinate conversion operation is performed, and i₁
γ and i₁Input δ to the first low pass filter 16
It The first low-pass filter 16 is i₁γ^#And i₁δ
^#Only pass through. The second adder 14 is i₁γ^#And the above
I and the integrated value are added_1PThe first inverse coordinate transformation time
Enter on path 17. Also, i₁δ^#Is the first inverse coordinate transformation
It is input to the conversion circuit 17. The first inverse coordinate circuit 17 is I_1PWhen
i₁δ^#And the positive phase component i of the u to w phases based on_1i( _i=
_{u, v, w}) Is calculated. The first coordinate conversion circuit 15,
One low-pass filter 16, second adder 14 and first
And the inverse coordinate conversion circuit 17 constitutes the first calculation means.
Has been done.

The second coordinate conversion circuit 21 uses the load current i _Li (
_i = _{u, v, w)} type, and performs an operation of the coordinate transformation, i ₂
γ and i ₂ δ are input to the second low pass filter 22. The second low-pass filter 22 uses i ₂ γ ^# and i ₂ δ.
Only pass ^# . The second inverse coordinate transformation circuit 23, based on i ₂ γ ^# and i ₂ δ ^# , has _{u to w} anti-phase components i _2i ( _i = _{u,
v, w} ) is calculated. The second coordinate transformation circuit 21, the second low-pass filter 22, and the second inverse coordinate transformation circuit 2
A second calculation means is constituted by 3 and.

Subsequently, the third adder 24 outputs i _1u and i _2u.
The command value i _u ^* is generated by calculating the difference between i _Lu and i _Lu . Similarly, the fourth adder 25 outputs i
An addition amount of _1v and i _2v, by calculating a deviation between i _Lv, generates a command value i _v ^*. Similarly, the fifth adder 2
6 generates a command value i _w ^* by calculating a deviation between i _Lw and the addition of i _1w and i _2w . The third adder 24, the fourth adder 25 and the fifth adder 26 constitute a fault current calculating means.

An arithmetic unit is composed of the first arithmetic means, the second arithmetic means, and the fault current arithmetic means. In FIG. 3, the deviation between the reactor current I _d and the set value I _d ^* is time-integrated by the integrator 13, and the integrated value which is the calculation result of the integrator 13 is i by the second adder 14. _The addition to ₁ γ ^{# is} for keeping the reactor current at a predetermined value. That is, I _d <I _d ^*
At the time of, it operates to increase I _1P . Then, the power supplied from the grid to the active filter 3 increases, and as a result,
The reactor current I _d increases. As a result, I _d ≈I _d
^* Control can be achieved.

Next, the set value I_d ^*For setting
The electronic circuit will be described with reference to FIG. Figure 4 shows the fault current
It is a circuit that constitutes a part of the detection device 4. Filter means
High-pass filter (HPF) 27-29 for each phase
Is the load current i of each phase_Li(_i=_{u, v, w}) Of the fundamental component
A filter for removing only the load current i_LiFrom each phase
I in which the fundamental wave component of_cu, I_cv, I_cwEach
Input to the hold circuit 30. Peak hold circuit
30 is the i input from the high pass filters 27 to 29.
_cu, I_cv, I_cwMaximum value of i_cmaxOnly hold and operate
Input to the container 31. Also, the peak hold circuit 30 is
Cycle of load current (commercial frequency 60H in this embodiment)
Input the reset signal from the oscillation circuit 34 tuned to z)
And the maximum value i_cmaxIs updated every cycle. Calculator 31
Is the maximum value i input for each cycle_cmaxGiven coefficient to
(0.8 in this embodiment) is multiplied and the switching circuit
It is input to the hold circuit 33 via 32. Switchon
Input to the oscillation circuit 34 every cycle
ON / OFF operation is performed based on the drive signal. That
The hold circuit 33 outputs the calculation result for each cycle.
Set value I_d ^*And holds the first adder 1 shown in FIG.
Set value I to 1_d ^*Enter. The peak hold circuit
A setting unit is configured by 30 and the arithmetic unit 31.
Further, a setting unit is constituted by the filter means and the setting means.
Is made. In addition, the setting section, feedback
The calculation unit is configured by the unit and the calculation unit.
The reason for multiplying the coefficient in the arithmetic unit 31
Is the reactor current I described later._dCompensation at any instant
Current i_i ^*Because it needs to be greater than. That
Therefore, in this embodiment, 1.5I_d≧ 1.2i_cmaxBecomes
Is set to I _d≧ 0.8i_cmax Tona
The coefficient 0.8 is set so that

Next, the compensation current generator 5 will be described. As shown in FIG. 1, a compensating current generator 5 is a pair of power converters connected in a multiplexed manner (hereinafter, simply referred to as a converter).
It is composed of A and B and a reactor 9.

FIG. 2 is a connection diagram of the converter A. In addition,
Since the converter B has the same structure as the converter A, the converter A will be explained.
The description of the converter B will be omitted. Moreover, in FIG.
For convenience of explanation, only one converter A has an LC filter.
The filter 6 is not connected, but the LC filter 6 is
As shown, connect the AC terminals of converters A and B in common.
There is. As shown in the figure, converter A is a current source PWM inverter.
(Hereinafter referred to as an inverter) 7 and a control circuit 8
Is made. The inverter 7 has a plurality of switching elements.
Bipolar transistor as a child (hereinafter Transis
T)_U ⁺, T_U ^-, T_V ⁺, T_V ^-, T_W ⁺, T
_W ^-, And diodes D1 to D6. I
A predetermined current (reactor voltage) is applied to the DC terminal of the inverter 7.
Flow) I_dIs connected to the reactor 9. Na
The reactor 9 connected to both converters A and B is a common iron.
Keep the phase between reactors (current balance) and smoothing
It is designed to combine.

Transistors T _U ⁺ , T _U ⁻ , T _V ⁺ , T
_The control circuit 8 controls _V ⁻ , T _W ⁺ , and T _W ⁻ so that one of each of the + side and the − side is always on. This control is performed as follows. The control circuit 8 of the converters A and B divides the time axis into minute times T equally as shown in FIG. 5, and outputs the average value of the current i _i at each minute time from the fault current detecting device 4 as a command value. i _i ^* ( _i = _{u, v, w} )
The PWM control is performed so that it becomes equal to that. That is, sawtooth wave modulation is performed.

[0037] When the ^{_{i u * -i v * ≡i uv}} * i v * -i w * ≡i vw * i w * -i u * ≡i wu * to define, the signal wave i _uv ^* ~i _wu ^* and The transistor to be turned on is uniquely determined as shown in FIG. 6A from the magnitude relation with the sawtooth wave i _s having the amplitude of 1.5 I _d .

[0038] For ^{_{example, i wu * <i s ≦}} i vw * of T in the period _U ^+, T _W ^- to turn on the. ^{_{i vw * <i s ≦ i}} uv * of T in the period _U ^+, T _V ^- turning on. Also, i _s ≤ i _wu
^{When *} and i _uv ^* <i _s , T _U ⁺ and T _U ⁻ are turned on.

When controlled in this manner, the current i for a minute time is
_i mean the micro period t = 0 to T is substantially equal to the command value i _i ^* for. That is, at t = τ _{1 to} τ ₃ in FIG. 6A, i _u = I _d and at the remaining time i _u = 0, the average value i _u = (τ ₃ −τ ₁ ) I _{d ^{/ T ≒ (i uv * -i}} wu *) becomes the _{I d / (3I d) =} i u *.

[0040] Similarly, the t = τ ₂ ~τ ₃ in i _v = -I
_d , and the average value i _v ≈i _v ^* . Further, in the three-phase circuit, since i _u + i _v + i _w = 0, the average value i _w ≈i
_w ^*

Further, in this embodiment, the converters A and B use saw-tooth wave carriers which are phase-shifted from each other by 1/2 cycle as shown in FIG. Since the converters are multiplexed, the output currents from the converters A and B are combined and the combined current is output to the LC filter 6.

The reason for this multiplexing, each transducer A, the output current is PWM (pulse width modulation), as shown in FIG. 5 controlled waveform of B, and the current (i _i ^* to be supplied to the distribution line In addition to (corresponding to), a high frequency component (hereinafter referred to as sideband) due to a large amount of PWM is included. According to the literature i _i
^* Is a harmonic current having a single frequency f _h , and f _c is a carrier frequency of PWM, f = mf _c + nf _h m = 1,2,3 ..., N = ± 1, ± 2, ± It is known to have sidebands with frequency components of 3 ...

When the power converters are multiplexed, m is an odd number.
The sidebands of are canceled out, leaving only the component with even m.
Are known. Therefore, due to this multiplexing, each converter
The combined current output from A and B and combined as shown in FIG.
(Eg i_u= I_uA+ I_uB, Note that i in the left brackets
_uIs the i output from each converter_uis not. Also, in FIG.
Show i_uAIs the output current i from converter A_uSay i_uB
Is the output current i from converter B_uSay. In the other phase
I _v, I_wIs the same as for)
The ram is as shown in FIG. Due to this multiplexing
Therefore, the output current of the inverter 7 is an even sideband (2f _c
± mf_h, M = ± 1, ± 2 ...) remains.

Next, the LC filter 6 will be described.
The LC filter 6 is connected to the AC terminals of the converters A and B, and the output terminal is connected to the distribution line 1 of each phase. The LC filter 6 with a capacitor C between the respective phases are connected, it is constituted by a coil L is connected to each phase, when the upper limit frequency of the harmonic to be removed and f _0, a f ₀ or more components It is designed to have a characteristic that almost all components are removed and components below f ₀ pass through without attenuation.

Therefore, the LC filter 6 supplies the component of f <f ₀ of the inverter output current to the distribution line side with almost no attenuation. In addition, most of the 2f _c sidebands caused by PWM flow sideways into the capacitor C, and only a very small current flows through the distribution line 1.

In the active filter 3 configured as described above, the fault current detection device 4 causes the load currents i _Lu , i _Lv , which flow in each phase of the distribution line 1 through the current transformer CT to flow in each phase.
Enter i _Lw respectively. Then, the fault current detection device 4
Detects the fault current i _i ^* superimposed on the system currents _isu , _isv , and _isw flowing from the grid.

Then, the fault current detection device 4 is operated by the fault current i.
_i ^* is output to each converter A and B as a command value. As shown in FIG. 7, the converters A and B perform PWM control, that is, sawtooth modulation, using sawtooth wave carriers that are phase-shifted from each other by ½ period so as to be equal to the command value i _i ^* .

It should be noted that this converter alone is shown in FIG.
A frequency spectrum as shown is obtained (Fig. 9 (a)).
Then, for convenience of explanation, f_h= 500Hz, T = 0.5m
s, the amplitude ratio (which corresponds to a sawtooth frequency of 5.0 Hz)
α = I_h/ I_d= 0.5 when the output current i_uThe spectacle
It's a tram. Note that I_hIs the command current, that is, the distribution line
Is the current that should be supplied to. ). In addition, in FIG. 9 and FIG.
And the vertical axis is | I_{m, n}/ I _d| Is shown.

In this embodiment, the spectrum of i _u (combined current) when two converters are multiplexed is shown in FIG.
It shows in (a). This is when T and α are the same as those in FIG. 9A, but when f _h = 1.2 kHz.

Then, the LC filter 6 supplies the component of f <f ₀ of the inverter output current to the distribution line side with almost no attenuation. Although a sawtooth wave is used as the carrier in this embodiment, the advantages of using the sawtooth wave in this embodiment will be described below.

Other configurations using only triangular waves for carriers
PWM control with the same active filter as in the above embodiment
The case of performing will be described. Triangle shown in Fig. 6 (b)
When PWM control is performed with the wave carrier, the signal wave i_uv ^*~ I
_wu ^*And 1.5I_dTriangular wave with amplitude i_tRelationship with
Determines the transistor to be turned on. in this case,
The number of commutations (number of times of turning on and off) in one minute period is shown in FIG.
As shown in (b), it is the same three times as in the case of the sawtooth wave.

As a result, the sawtooth wave carrier can have twice the frequency of the triangular wave carrier. From this, the frequency f of the sideband becomes high, and the coil L and the capacitor C of the LC filter can be made small.

Further, FIG. 9B shows the converter shown in FIG.
₄ shows the frequency spectrum of the current i _u when operated so as to output a sine wave current of _h = 500 Hz.
The carrier frequency (triangular wave frequency) is 2.5 kHz. Comparing FIGS. 9A and 9B, m =
1 sidebands are the most, whereas triangular waves are m = 2
There are a lot of it.

FIG. 10B shows the spectrum of the combined current (combined current) at f _h = 1.2 kHz when the converters A and B are multiplexed using triangular wave carriers. The carrier frequency (triangular wave frequency) is 2.5 kHz.
In the sawtooth wave of FIG. 10 (a), I _{2, -2} , I _{2, -1} , I _2,1 , I
_{Although 2 and 2} are conspicuous, these frequencies are several times higher than I _h , and the inflow into these distribution lines 1 can be blocked by a simple LC filter. Note that f _h = 1.2k
Hz corresponds to the 19th-order frequency of the commercial power supply (60 Hz).
In the case of a triangular wave, I _{2, -1} , I _2,1 are close to I _h ,
It is necessary to add a band elimination filter for attenuating this.

When PWM control is performed by the triangular wave carrier and the power converters are multiplexed, the result is as shown in FIG. 10 (b). In FIG. 10B, the converter has f _h = 1.
It is a frequency spectrum when operated so as to output a sine wave current of 2 kHz. The triangular wave carrier frequency f _c is 2.5 kHz. In FIG. 10B, the components of I _{2, -1} and I _2,1 remain, and a band elimination filter circuit is required to remove these components, but a complicated circuit is essential. Furthermore, if the frequency f of the sideband is increased, the coils and capacitors forming the band elimination filter can be reduced, but there is a limit to increasing the carrier frequency f _{c of} PWM.

As described above, since the sawtooth wave is used as the carrier in this embodiment, there is an advantage over the triangular wave. It should be noted that the present invention does not exclude a triangular wave as a carrier because the effect of the present invention can be obtained even if a triangular wave is used as a carrier.

The present invention is not limited to the above-mentioned embodiments, but may be implemented by appropriately modifying a part of the structure without departing from the spirit of the invention. (1) Although the LC filter is used in the above-mentioned embodiment, it may be configured with other filters.

[0058]

As described above in detail, according to the invention of claim 1, when the reactor current is feedback-controlled, the set value is based on the maximum phase value of the residual component obtained by removing the fundamental current from the load current. Since the set value is determined based on the set value, the set value can be changed in accordance with the change in the fault current. Therefore, accurate feedback control can be performed as compared with the fixed set value.

According to the invention of claim 2, in addition to the effect of claim 1, the frequency of the sideband can be increased.
As a result, the sideband removing filter can be simplified,
The performance can also be improved significantly compared to the triangular wave carrier.

According to the third aspect of the present invention, the harmonics caused by the pulse width modulation control are removed by the band elimination filter, and the compensation current can be reliably output to the distribution line. According to the invention of claim 4, the coil and the capacitor can be downsized by the LC filter.

[Brief description of drawings]

FIG. 1 is a circuit diagram of a current source active filter of an embodiment embodying the present invention.

FIG. 2 is a circuit diagram of a converter which also constitutes a current source active filter.

FIG. 3 is an electronic circuit diagram of a main part of the fault current detection device.

FIG. 4 is an electronic circuit diagram of the main part of the fault current detection device for calculating the set value I _d ^* .

FIG. 5 is an explanatory diagram showing an output current waveform of a current source converter.

FIG. 6A shows P of the current source converter in the case of sawtooth wave modulation.
FIG. 3B is an explanatory diagram of a WM control pattern, and FIG. 6B is an explanatory diagram of a PWM control pattern of the current source converter in the case of triangular wave modulation.

FIG. 7 is an explanatory diagram of carriers when multiplexed in the case of sawtooth wave modulation.

FIG. 8 is an explanatory diagram of carriers when multiplexed in the case of triangular wave modulation.

9A is a frequency spectrum of a converter output current in the case of sawtooth modulation, and FIG. 9B is a frequency spectrum of a converter output current in the case of triangular wave modulation.

10A is a frequency spectrum of a converter output current in the case of sawtooth modulation by multiplexing the converter, and FIG. 10B is a frequency spectrum of a converter output current in the case of triangular wave modulation by multiplexing the converter. Is.

[Explanation of Codes] 1 ... Distribution line, 2 ... Load, 3 ... Active filter, 4 ...
Fault current detecting device as detecting means and calculating means, 5 ... Compensating current generating device as compensating current generating means, 6 ... LC filter, 7 ... Inverter, 8 ... Control circuit, 9 ... Reactor, 11 ... First adder , 12 ... Coefficient adder, 13 ... Integrator (which constitutes a feedback section together with the first adder 11 and the coefficient adder 12), 14 ... Second adder, 15 ... First coordinate conversion circuit, 16 ... First low pass filter, 1
Reference numeral 7 ... First inverse coordinate transformation circuit (first computing means is configured with each of the circuits 14 to 16), 21 ... Second coordinate transformation circuit, 22 ... Second low-pass filter, 23 ...
Second inverse coordinate transformation circuit (second arithmetic means is configured together with the circuits 21 and 22), 24 ... Third adder, 25 ... Fourth adder, 26 ... Fifth adder (A fault current calculating means is configured with each of the devices 24 and 25, and a computing section is configured with the first computing means and the second computing means), 27 to 29 ... HPF (filtering means) , 30 ... peak hold circuit, 31 ... arithmetic unit (setting means is configured with the peak hold circuit, and setting section is configured with the filter means, and
The setting unit, the feedback unit, and the calculation unit constitute the calculation means), CT ... Current transformers, A, B ... Power converters.

Claims (4)

[Claims] 1. A detection means for detecting a load current in which a fault current generated from a load is superimposed on a fundamental wave current flowing through a distribution line, and a calculation means for calculating the fault current based on a detection result of the detection means. And a plurality of switching elements, the switching element is turned on / off by a pulse width modulation method based on the calculation result of the calculation means to generate a compensation current equivalent to the fault current, and the compensation current to the distribution line. In the active filter for distribution having a plurality of compensating current generating means for supplying, the calculating means removes the fundamental current of each phase from the load current of each phase and passes the residual component of each phase. A setting unit that includes a setting unit that determines a setting value based on the value of the residual component of the maximum phase among the residual components of each phase input from the filter unit; and the setting unit determines the setting value. A feedback unit for calculating a difference from the set value and a compensation current generated by the compensation current generating means, a positive phase component of each phase is extracted from a load current of each phase, and the positive phase component and the feedback unit The first arithmetic means for calculating the corrected positive phase component on the basis of the difference from, the second arithmetic means for arithmetically operating the negative phase component of each phase from the load current of each phase, the first arithmetic means and the first arithmetic means. A correction unit for each phase, which is the calculation result of the second calculation unit, a calculation unit having a fault current calculation unit for calculating the fault current of each phase based on the reverse phase component of each phase and the load current; Current source active filter for distribution. 2. The compensating current generating means uses a sawtooth wave as a carrier wave used in a pulse width modulation method, and turns on / off a switching element based on a magnitude comparison between the sawtooth wave and a fault current waveform.
The current source active filter for electric power distribution according to claim 1, which outputs a compensation current when turned off. 3. The current source active filter for distribution according to claim 1, which is connected to the distribution line via a band elimination filter. 4. The current source active filter for distribution according to claim 3, wherein the band elimination filter is an LC filter.