WO1999012243A1

WO1999012243A1 - Method and device for offsetting network voltage distortions

Info

Publication number: WO1999012243A1
Application number: PCT/DE1998/002534
Authority: WO
Inventors: Michael Weinhold; Rainer Zurowski; Martin Sonnenschein
Original assignee: Siemens Aktiengesellschaft
Priority date: 1997-09-01
Filing date: 1998-08-27
Publication date: 1999-03-11
Also published as: BR9812619A; NO20001054D0; JP2001515339A; DE19738125C2; DE19738125A1; CA2302498A1; NO20001054L; EP1010231A1

Abstract

The invention refers to a method and device for offsetting network voltage (14) distortions (uN,v±), involving a pulsating current rectifier (4) and an active filter (2) comprising a coupling device for LC oscillatory circuits. According to the invention, at least one complex amplitude (uN,v±) and at least one distortion of the network voltage (uN,v±), from which a space vector for the partial transfer ratio (üv±) and a space vector for the partial transfer ratio (üh) are determined depending on the real value of the transfer circuit voltage (Vdc) in the pulsating current rectifier (4), while control signals (S1....., S6) intended for the pulsating current rectifier are generated on the basis of a space vector for the partial transfer ratio (ü), obtained from the space vectors for the partial transfer ratio (ü v±, üh). As a result, the active filter (2) operated, in on connecting point (20) on the network (14), as a complex low impedance resistor for higher harmonics to be mitigated (uN,v±), and the capacitive accumulator (6) in the pulsating current rectifier (4) is adjusted according to a set a nominal value for the transfer circuit voltage (Vdcsoll).

Description

description

Method and device for compensating for mains voltage distortions occurring in a network

The invention relates to a method and a device for compensating for mains voltage distortions occurring in a network by means of an active filter having a pulse converter with capacitive memory and an LC resonant circuit coupling.

Passive filters have been used to filter harmonics, which are tuned to the corresponding frequency to be filtered. Due to the strong improvement and approval of power semiconductor components, pulse converters are increasingly being developed and used as active filters. These either completely replace the passive filters or, in combination with passive filters, improve the characteristics of the passive filter system (hybrid filter). The advantage of active filters compared to passive filters is that the filter effect of an active filter is infinitely adjustable and that active filters can protect themselves against overload. They then work along their design limit. In addition, any existing parallel or series resonances with the

Mains impedance can be avoided. In the case of a digital control system, the properties of an active filter can be changed quite simply by re-parameterizing the control parameters.

The status of the active filter is presented in the article "NEW TRENDS IN ACTIVE FILTERS" by H. Akagi, printed in the conference volume "EPE '95", Seville, pp. 0.017 to 0.026. This active filter can be connected in parallel to a load or in series in a connecting line between a network and a load. In the former case, the reactive current components in the load current are determined and by means of the active one Filters are suctioned off at the feed point of the load so that the network is relieved of the reactive currents. When the active filter is connected in series, network harmonics of the network are compensated in such a way that the load voltage is relieved of these network harmonics. Combinations with an active filter and a passive filter and a combination of two active filters with parallel and serial coupling are also presented. This article also shows that the active filter with parallel coupling processes the mains voltage instead of processing the load current when this active filter with parallel coupling is connected in a distribution network.

In the article "HYBRID-ACTIVE FILTERING OF HARMONIC CURRENTS IN POWER SYSTEMS", printed in "IEEE / PES Winter Meeting", 1995, pp. 1 to 7, the topology and control of a hybrid filter, consisting of a passive and active filter , presented. The active filter has a pulse converter with capacitive memory, which is connected to an AC voltage network by means of a rectifier and a transformer. The pulse converter is electrically connected in series with the passive filter by means of a coupling transformer. The passive filter has several LC resonant circuits, each of which is tuned to a harmonic and whose impedance is capacitive. The capacitive memory of the pulse converter is controlled to its setpoint by means of the rectifier. The load current, the mains voltage and the filter voltage are used as input signals of the regulating and control device of the active filter. Various current components of a filter current setpoint are determined from these input signals and are superimposed on the filter current setpoint. The control signals for the pulse converter of the active filter are generated by means of a current control device, at the inputs of which the determined filter current setpoint and a measured filter current actual value are present. From the publication "Shunt-Connected Power Conditioner for Improvement of Power Quality in Distribution Networks", printed in "International Conference on Harmonics and Quality of Power", Las Vegas, Nevada, Oct. 16-18, 1996, a control method for a compensation device with parallel coupling is known. It can be seen from this conference report that the compensator voltage space vector is calculated from the voltage drop across the capacitive memory and a transmission ratio space vector. This report also shows that the transfer ratio space pointer can be composed of several partial ratio space pointers. It also specifies how the partial transfer ratio space pointers are determined.

This compensation device has a pulse converter with at least one capacitive memory, an adaptation filter and a regulating and control device. This compensation device is electrically connected in parallel to a non-ideal consumer that is supplied from a network. A mains voltage space pointer, a mains current space pointer and an intermediate circuit voltage actual value, which drops out at the two capacitive memories of the pulse converter, are fed to the regulating and control device. These room pointers are generated by means of a room pointer transformation device from measured conductor voltages and mains currents. The regulating and control device has a regulating device for determining a transmission ratio space vector and a pulse width modulator. The transmission ratio space vector is the manipulated variable of the pulse converter, which is converted into control signals for this pulse converter by means of the pulse width modulator.

In order for the desired filter effect of the active filter to occur, this active filter must act like a complex resistor with low impedance at the connection point for the harmonics to be damped. The filter thus serves as a sink for the upper floors connected in the neighborhood. Vibrator. It would be advantageous if the filter behaves like an ohmic resistor and thus dampens harmonics. This would avoid resonance effects between the filter and the network.

In order to achieve this, the active filter in its connection terminals must detect the harmonics to be filtered in the mains voltage and absorb a current in phase with the respective harmonics. The size of the current can be adjusted by varying the value of the ohmic resistance. In the ideal case, the active filter then forms a short circuit for the selected harmonics, so that the selected harmonic voltages at the node of the filter become zero.

The invention is based on the object of specifying a method and a device for carrying out the method for an active filter in such a way that a voltage difference between the mains and pulse converter voltage stationary drives a desired filter current through an LC resonant circuit coupling.

This object is achieved according to the invention with the features of claim 1 and claim 3.

Using this method according to the invention, a total transmission ratio space vector is determined in such a way that the voltage difference between the mains voltage and the pulse converter voltage steadily drives a desired filter current through an LC resonant circuit coupling. The total transfer ratio space pointer is composed of several partial transfer ratio space pointers.

A first partial transmission ratio space vector, which is determined as a function of a determined complex amplitude of a mains voltage distortion present in the network, is used as a function of the intermediate circuit voltage actual worth generating a pulse converter voltage that drives a filter current, which causes a line voltage distortion in the line voltage to disappear due to its voltage drop at the line impedances. The active filter thus acts as a short circuit for a line voltage distortion at the connection point.

By means of the other partial transmission ratio space vector, a distortion component is generated in the pulse converter voltage which is equal to the mains voltage distortion component which is determined from the difference between the mains voltage and the mains voltage fundamental oscillation. As a result, no unwanted distortion currents flow into the active filter, so that the entire construction power of the pulse converter is available for filtering predetermined mains voltage distortions.

To save further converter construction work, the LC resonant circuit coupling must keep the mains voltage fundamental oscillation away from the pulse converter and at the same time offer a low-ohmic coupling of the pulse converter for a mains voltage distortion to be filtered. These properties are met by an LC resonant circuit coupling, the resonance frequency of which is, for example, not the same as the frequency of the fifth harmonic but, for example, the same as the frequency of the fourth harmonic. In this way, the pulse converter is not significantly loaded with the mains voltage fundamental oscillation, so that almost its full construction is available for harmonic filtering. As a result of this coordination of the LC resonant circuit coupling, the entire impedance between the pulse converter and the network is inductive for a network voltage distortion to be filtered, so that the method is always stable.

In an advantageous method, a further partial transmission ratio space vector is determined which, if the intermediate circuit voltage of the pulse converter is too low, phase and if the DC link voltage is too high it is in phase opposition to the filter current fundamental. This results in a pulse converter fundamental oscillation voltage, as a result of which energy is supplied or withdrawn from the capacitive memory of the pulse converter. The intermediate circuit voltage is thus regulated to a predetermined target value, so that an additional energy supply device for the capacitive memory of the pulse converter can be dispensed with.

To further explain the method according to the invention for compensating for mains voltage distortions occurring in a network by means of an active filter, reference is made to the drawing, in which an exemplary embodiment of the device for carrying out the method according to the invention is schematically illustrated:

1 shows a block diagram of an active filter according to the invention, which has a pulse converter with an LC resonant circuit coupling, which

FIG. 2 shows the structure of a control device for generating an overall transmission ratio space vector, wherein in

Fig. 3 shows the block diagram of a mains voltage distortion controller of the control device according to Fig. 2, the

FIG. 4 shows the block diagram of the mains voltage distortion regulator according to FIG. 3 with a current limitation, FIG. 5 shows the block diagram of a control device for preventing undesired filter currents of the control device according to FIG. 2, FIG. 6 showing a block diagram of a DC voltage regulator of FIG Control device of the pulse converter represents. 1 shows a block diagram of an active filter 2, which has a pulse converter 4 with at least one capacitive memory 6, an adaptation filter 8 and a regulating and control device 10. This pulse converter 4 is electrically connected in parallel in a larger network 14, for example in a ring network, by means of an LC resonant circuit coupling 12. Of this network 14, only two network areas 14ι, 14 ₂ are shown, in which several harmonic generators, not shown, act. The regulating and control device 10 are a mains voltage space vector u _N , a

-.

Filter current space vector i _κ and an intermediate circuit voltage

Actual value V _dc which drops at the two capacitive memories 6 of the pulse converter 4 is supplied. These space vectors u _N and → i _κ are measured from a space vector transformation device, which is not shown in detail

Conductor voltages and filter currents generated. The adaptation filter 8 is shown here as a substitute by an inductance L _κ , whereas this adaptation filter 8 is shown in detail in the conference report from Las Vegas mentioned at the beginning. The regulating and control device 10 has a regulating device 16 for determining an overall transmission ratio space vector u and a pulse width modulator 18, which is represented by a broken line. The total transmission ratio space vector ü → is the manipulated variable of the pulse converter 4, which is converted by means of the pulse width modulator 18 into control signals Si, ..., S ₆ for this pulse converter 4.

The LC resonant circuit coupling 12 consists of a series connection of an inductor L and a capacitor C. In order to

To reduce the performance of the pulse converter 4, this must

LC resonance circuit coupling 12 have the property that they have a high resistance for the mains voltage fundamental oscillation u _N ι ₊ and a low impedance for the harmonics to be filtered u _{N / V +} .

-> Should the fundamental vibration displacement blind power consumption of this coupling circuit 12 should be as low as possible, the LC resonant circuit must be matched to the harmonics to be filtered. Fig. 1 shows an example of the circuit for an active filter that is to filter a 5th order harmonic in the mains voltage u _N. In the event of

- Several harmonics to be filtered, it can be advantageous to connect several suitably tuned LC resonant circuits in parallel. The resonance frequency of this LC resonant circuit coupling 12 is selected such that the total impedance between the pulse converter 4 and the node 20 in the network 14 is inductive for the 5th harmonic to be filtered. In order to achieve a further reduction in the power output of the pulse converter 4, the coupling 12 must keep the basic line voltage direction u _N , ι ₊ away from the pulse converter 4 and at the same time offer a low-impedance coupling of the pulse converter 4 for the harmonics u _{N / V + to be} filtered. In this way, the pulse converter 4 is only slightly with the

Mains voltage fundamental wave → u _N , ι + loaded, and the full

Pulse converter construction is available for harmonic filtering. In practice, an IGBT two-point converter with a DC voltage circuit is selected as the pulse converter 4 due to the high switching frequencies available. Task of measured value acquisition, control device 16 and

It is pulse width modulator 18 that the total transmission ratio space vector u required for the filtering between the

->

DC link voltage actual value V _dc and the pulse converter output voltage space vector u _κ to generate.

If the active filter 2 as a sink for harmonics in the network areas 14ι and 14 ₂ with several connected but not shown harmonic generators u _N , _{v +}

act, the only sensible control strategy is that the filter 2 does not filter a special load current. tert, but acts at node 20 for the harmonics → u _{N / v ±} as a low-resistance resistor.

1 shows that on both sides of the connection point 20 network areas 14ι and 14 _{2 of} the network 14 can be seen, which are each represented by network connection impedances R _N ι, L _m and R _N , L _N2 and the shares a 5th order distortion voltage → u _N ι, ₅ - and → u _{N2 / 5} -. The network

14 contains any harmonic generators, which are also connected via line impedances and influence the distortion voltages → u _N ι, ₅ - and → u _{N / 5} -. At the junction

20 results in a filter voltage → u _{K / 5} _. Since the active filter 2 is regulated, for example, for filtering the 5th harmonic, this active filter 2 is low-inductive for this 5th harmonic and high-impedance for all other frequencies. The filter current i _{K / 5} _, which flows in phase with the filter voltage u _κ , ₅ - and flows through the active filter 2

→ → composed of the harmonic currents im, ₅ - and i _N , ₅ - contained in the mains current in accordance with the connected network impedances R _N ι, L _Ni and _N2 , L _N. The active filter 2, which resembles an ohmic resistance, thus absorbs part of the harmonic currents → i _N ι, ' ₅ _ and → i ₂ , ₅ - occurring in the connected network 14, and the more so, the lower the resistance of this active filter 2 is. If the ohmic resistance of this active filter 2 is equal to zero, the harmonic

Voltage → u _N , ₅ - at node 20 to zero volts and the active filter 2 draws the maximum possible current. This active filter 2 thus forms a short circuit, for example for the 5th harmonic at node 20. The structure of the control device 16 is shown schematically in FIG. 2. This control device 16 has at least one mains voltage distortion controller 22, a control device 24 for preventing undesired filter currents i _κ , _{v +} and a DC voltage controller 26, the outputs of which are linked to a summation point 28. The structure of the controller 22 is shown in more detail in FIG. 3 or FIG. 4, the illustration in accordance with FIG. 4 additionally having a current limitation. The structure of the control device 24 is shown in FIG. 5 and the structure of the DC regulator 26 in FIG. 6. A determined mains voltage space vector → u _{N is} fed to the mains voltage distortion regulator 22, which is also referred to as a harmonic regulator. If this harmonic controller 22 is configured according to FIG. 4, a determined filter current space vector → i _κ and a maximum converter effective current value I _{max are also} supplied, these two signals → i _κ and I _{max being} shown here by means of a broken line . At the output of this harmonic controller 22 there is a partial transmission ratio space vector → ü _v - ₊ . The index V with V = 5, 7, 11, ... stands for the ordinal number of a harmonic, the index + or - identifying the co-system or the counter system. For the example shown in FIG. 1, v = 5, this 5th harmonic occurring in the negative system, so that as the second

Index - appears. The partial transfer ratio space pointer _→ ü _{v +} is then -ü. ₅ _. The control device 24 is a

middle line voltage space pointer → u _N and an intermediate circuit

Actual voltage value V _dc supplied. A partial transfer ratio space pointer is also present at the output. The same Voltage controllers 26 are the determined filter current space vector i _K / the determined intermediate circuit voltage actual value V _dc

- »and a predetermined intermediate _circuit voltage setpoint V _cso ii supplied. At the output there is also a partial transfer ratio space vector _dc . By means of the summation parts 28

these partial transfer ratio space pointers → ü _v - ₊ , → ü _h and ü _dc become a total transfer ratio space pointer ü

-> → added, from which 18 control signals Si, ..., Sβ for the pulse converter 4 are subsequently generated with the aid of the pulse width modulator.

In Fig. 3 the structure of the harmonic controller 22 is shown in more detail. On the input side, this harmonic controller 22 has an identification device 30 for determining a complex amplitude u _{N / V ± of} one present in the network 14

Mains voltage distortion → u _{N / v +} with downstream I controller

32 and multipliers 34 and 36 electrically connected in series on the output side. The identification device 30 has a complex multiplier 38 with a downstream mean value generator 40, an input of this complex

Multiplier 38 is linked to an output of a unit space vector 42. The mains voltage space vector → u _{N is present} at the second input of this multiplier 38.

A complex amplitude u _N , _{v ± of} a Vth network distortion voltage u _{N / v ± is} formed from the product present at the output of the multiplier 38 by means of the averager 40 with respect to a network period T. At the exit of the unit

The space vector 42 is a conjugate complex unit space vector e *, which in the co-system with an angular

- »speed + vω and in the opposite system with an angular speed revolves -vω, where ω is the rotational frequency of the mains voltage fundamental oscillation space vector u _{N #} ι ₊ . By averaging over the network period T, the product of the network voltage space vector u _N and conjugate

plexing unit space vector e * the complex amplitude un u _{Nι Vi}

of the corresponding mains voltage distortion u _{N / v ±} .

The output signal of the I controller 32 is multiplied by an imaginary unit + j by means of the multiplier 34. The product present at the output of this multiplier 34 is multiplied by means of the downstream multiplier 36 by a unit space vector → e which is present at the output of a further unit space vector 44. The product of this multiplication is a partial transfer ratio space pointer → ü _{v +} . The I controller 32 changes the amount

and the angle of the partial transfer ratio space vector ü _{v ±}

then until the corresponding mains distortion voltage

U _{, V +} vth ordinal number in the mains voltage space vector u _N ele-

→ → is mined.

The task of this harmonic controller 22 is first of all to identify the amplitude of a network distortion u _{N / v ±} from the network voltage u _N and to use this to determine a corresponding partial

-> to generate the transfer ratio space pointer → ü _{v +} , the

filter current i _κ , _{v +} , which due to its voltage drop across the line impedances R _N ι, L _Ni and R _N2 , L _N2 causes the line voltage distortion → u _N , _{v ±} to disappear in the line voltage → u _N

leaves. The active filter 2 thus acts for this distortion Voltage → u _{N / v +} at node 20 like a short circuit. to

Identification of a complex amplitude u _N , _{v ±} one in the network

14 existing distortion voltage → u _{N / v +} is determined using a

discrete complex Fourier transform, which is implemented in the identification device 30, is used. For this purpose, considering the representation according to FIG. 1, the mains voltage space pointer → u _{N is combined} with a conjugate

xen multiplied the unit space vector → e *, the angular velocity of which is 5ω, and then fed to a moving average window. In the case of a co-system harmonic, the conjugate complex unit space pointer → e * must have the angular velocity -vω. The output signal of the I controller 32 is rotated by + 90 ° (multiplication by + j) and multiplied by a unit space vector e rotating in the opposite system direction by the angular velocity -5ω. In the case of a system harmonic, the output signal of the I-controller 32 must be rotated by -90 ° (multiplication by -j) and multiplied by a unit space pointer e rotating in the system direction by an angular velocity + vω. The I controller 32 changes its output until the 5th order negative system in the mains voltage space vector → u _{N has} disappeared.

FIG. 4 shows the structure of a harmonic regulator 22 which is provided with a current limiting device 46.

This current limiting device 46 has a current control circuit 48, a multiplier 50 and a device 52

Formation of a filter current peak actual value i _κ . The current control circuit 48 consists of a comparator 54 and a PI controller 56 which limits on the output side and with a nem input of the multiplier 50 is connected. The non-inverting input of this comparator 54 is linked to the output of the device 52, at the input of which a determined filter current space vector i _κ is present. At the inverting input of the comparator 54 there is a filter current peak setpoint iκsoiι which is specified as a function of a converter effective current value I _max . The facility for

Formation of the filter current peak _actual value i _κ has an absolute value generator 58 and an average value generator 60, which is connected downstream of the absolute value generator 58. At the output of the comparator 54 there is a deviation of the actual filter current peak value i _κ from its target value i _K soiι, from which a manipulated variable VP is generated by means of the PI controller 56. This manipulated variable VP is multiplied by the output variable of the I controller 32 and fed to an inverting input of a further comparator 62, at the non-inverting input of which the output of the mean value generator 40 of the identification device 30 and at the output of which the input of the I controller 32 are connected.

Through this coupling of the current limiting device 46 shown, the I controller 32 of the harmonic controller 22 becomes a PI controller. In the event of an overcurrent, the manipulated variable VP increases the negative feedback of the harmonic controller 22 until the pulse converter 4 operates at its current limit. In this operating mode, the active filter 2 thus always works like the minimum possible ohmic resistance. In order to prevent the manipulated variable VP from becoming negative, that is to say the negative feedback then becomes positive feedback, the PI controller 56 must be limited downwards to zero. If the permissible pulse converter current i _{κ is} not sufficient to

-. a Vth harmonic → u _N , v + in the mains voltage space vector

To eliminate u _N , the current limiting device ensures

46 that the filter current → i _κ , v + in phase to the mains

voltage distortion → u _{N; V ±} and the pulse converter 4 always works at the current limit. Thus, the active filter 2 at node 20 always acts as the smallest possible ohmic resistance for the harmonic to be damped. If there is more than one harmonic to be actively filtered, a harmonic controller 22 is required for each harmonic to be filtered, each of which has a partial transmission ratio space vector → ü _{v +} in the manner described here

to calculate .

FIG. 5 shows the structure of the control device 24, which has a device 64 for forming a mains voltage fundamental oscillation co-system space vector → u _{N /} ι ₊ , a subtractor

66 and a multiplier 68, which is connected downstream of the subtractor 66. At the input of the device 64 is the determined mains voltage space vector → u _N , which is also present at the first input of the subtractor 66. The second input of the multiplier 68 is linked to an output of a reciprocal generator 70, at whose input a determined intermediate circuit voltage actual value V _dc is present. At the output of this multiplier 68 of the partial transmission ratio space vector is u to _h. - »

The device 64 for determining a mains voltage basic vibration co-system space vector u _Nι ι ₊ has input and on the output side each have a complex multiplier 72 and 74, which are linked to one another by means of an averager 76. In addition, there are two unit space vector generators 78 and 80, each of which generates a space vector → e * and e, which rotate in opposite directions at the same frequency. At the output of this device 64 there is then a line voltage basic oscillation co-system space vector u _{N /} ι ₊ which

→ is subtracted from the mains voltage space vector → u _N by means of the subtractor 66. The result is all distortion voltage → u _N , _{v ±} present in the mains voltage → u _N. After multiplication by the reciprocal of the DC link voltage actual value V _d c, the partial transfer ratio space pointer ü- → _h results - this partial transfer ratio space pointer → ü causes the same distortion component u _{K / V} + in the pulse converter output voltage → u _K / so that no unwanted

Distortion currents → i _κ , _v - ₊ flow into the active filter 2. The active filter 2 is thus only subjected to a desired filter current i _κ , v ₊ .

, ^-

6 shows the structure of the DC voltage regulator 26 in greater detail. This DC voltage controller 26 has on the input side a device 82 for determining a filter current fundamental oscillation co-system space vector → iκ, ι + and on the output side two multipliers 84 and 86 connected in series. In addition, this DC voltage regulator 26 has an intermediate circuit voltage control circuit 88 and a reciprocal generator 90. The output of the intermediate circuit voltage control circuit 88 is linked to a second input of the multiplier 84, whereas the output of the reciprocator 90 is connected to a second input of the multiplier 84. rers 86 is linked. At the output of the multiplier 86 there is a partial transmission ratio space pointer → ü _d .

The intermediate circuit voltage control circuit 88 consists of a comparator 92 and a PI controller 94. The inverting input of the comparator 92 is connected to an output of a delay element 96 of the first order, at the input and at the input of the reciprocal generator 90 the intermediate circuit determined -Voltage actual value V _dc is present. A predetermined intermediate circuit voltage setpoint Vdc _so ii is present at the non-inverting input of the comparator 92.

The device 82 for determining a filter current fundamental oscillation co-system space vector i _κ , ι + is identical to

→ Device 64 for determining a mains voltage basic vibration co-system space pointer → u _N , ι + set up so that it is no longer necessary to go into this at this point. This device 82 determines the fundamental oscillation of the filter current → i _K, which is mainly caused by the mains voltage

Fundamental vibration u _N , ι ₊ and the impedance of the LC resonant circuit coupling 12 is determined for the fundamental vibration.

The intermediate circuit voltage control circuit 88 compares the smoothed intermediate circuit voltage actual value V _dc with its target value Vcsoii and feeds the control deviation Δv _a c to the PI controller 94. In order to raise the DC link voltage actual value V _dc , the pulse converter 4 on the network side must have a fundamental oscillation voltage → u _κ , ι ₊ in phase with the LC oscillation circuit

Coupling 12 flowing fundamental oscillation currents i _κ , ι ₊ generate

-. gen. Since the mains voltage fundamental _{oscillation N} , ι ₊ much larger will be as the pulse converter output voltage fundamental oscillation -u → _κ , ι +, the fundamental oscillation current -i → _κ , ι ₊ through the LC resonant circuit coupling 12 can initially be a current

Power source can be considered. The filter current basic vibration co-system space vector i _κ , ι ₊ is analogous to the mains voltage

→ nung-basic vibration-co-system space vector u _N , ι ₊ over one

→ complex Fourier analysis determined and then transformed back. After the inverse transformation, the multiplication of the output variable V _dC y of the intermediate circuit voltage control loop 88 by the filter current fundamental oscillation system space pointer → i _κ , ι ₊ results in the pulse converter output voltage

Fundamental vibration co-system space vector u _κ , ι + calculated that

→ with the reciprocal of the intermediate circuit voltage actual value

V _dc gives the partial transfer ratio space pointer → ü _d .

If the intermediate _circuit voltage actual value Vd _C now deviates from its target value V _cso ii, the partial transfer _ratio space vector ü _{c is} generated, which if the intermediate circuit

→

Actual voltage value V _{dc in} phase and if the intermediate circuit voltage actual value V _{c is} too large in phase opposition to the filter current

Fundamental vibration co-system space vector i _κ , ι + ist. This results in a pulse converter output voltage basic oscillation co-system space vector u _κ , ι ₊ , which makes the capacitive

-.

Storage 6 energy is supplied or withdrawn and thus the intermediate circuit voltage actual value V _{dc is} regulated to an intermediate _circuit voltage setpoint V _dcs oii.

By means of this method according to the invention for a pulse converter 4 of an active filter 2 with an LC resonant circuit coupling 12, it is achieved that the active filter ter 2 at node 20 for the harmonics to be damped acts as a complex resistor with low impedance, whereby this active filter 2 serves as a sink for the harmonic generators connected in the neighborhood. The active filter 2 behaves like an ohmic resistor, which dampens the harmonics. In this way, resonance effects between the active filter 2 and the network 14 can be avoided. This method according to the invention is not only limited to a pulse converter 4 of an active filter 2 with an LC resonant circuit coupling 12, but can also be easily transferred to other coupling variants.

Claims

claims

1. Method for compensating for mains voltage distortions (→ u _N , _{v ±} ) occurring in a network (14) by means of an active filter (2) having a pulse converter (4) with capacitive memory (6) and an LC resonant circuit coupling (12) ) with the following process steps:

a) determining at least one complex amplitude (u _{N / V ±} ) of a mains voltage distortion present in the network (14)

(→ u _N , _{v ±} ), b) Generation of a partial transfer ratio space vector

(ü _{v +} ) depending on these determined complex ampli. ^- tude (u _{N / v ±)} such that this becomes zero, in dependence c) determining from occurring in the network (14) mains voltage distortions (→ u _N,) of a detected chip-voltage space vector (→ u _N) and. a mains voltage fundamental oscillation space system space vector (u _{N / v ±} ); d) Determination of a partial transmission ratio space vector (ü _h ) as a function of this line voltage distortion

( _→ U _N , _{V +} ) and an intermediate _circuit voltage (V _dc ) of

Pulse converter (4) and e) generation of control signals (Si, ..., S ₆ ) for the

Pulse converter (4) from a total transfer ratio space pointer (ü) formed from the partial transmission ratio space pointers (→ ü _v ^_ _{+ /} → ü).

2. The method of claim 1, wherein

f) a filter current fundamental oscillation co-system space vector

(→ iκ, ι ₊ ) is determined from a determined filter current space vector (→ i _κ ), g) an intermediate circuit voltage deviation (ΔVc) of a determined intermediate circuit voltage actual value (V _dc ) of a ne predetermined intermediate circuit voltage setpoint (Vdc _so ii) is determined, h) an intermediate circuit manipulated variable (V _dc y) is generated such that this determined intermediate circuit voltage deviation (ΔVc) becomes zero, i) by multiplying the Filter current basic oscillation co-system space vector (i _κ , ι ₊ ) with the DC link manipulated variable (Vd _c y) a pulse converter output basic oscillation co-system space vector (u _κ , ι +) is determined, and

- »j) where, depending on this pulse converter output voltage basic vibration basic system space vector (→ u _κ , ι ₊ ) and the intermediate circuit voltage setpoint (V _d c _s oii)

Partial transfer ratio space pointer (ü _dc ) generated

- », which is superimposed on the overall transfer ratio space pointer (ü).

3. Device for performing the method according to claim 1 for an active filter (2), which has a pulse converter (4) with a capacitive memory (6) and a regulating and control device (10) and an LC resonant circuit coupling ( 12), this regulating and control device (10) having a regulating device (16) for determining an overall transmission ratio space vector (ü) with a downstream pulse width modulator (18), at the outputs of which control signals (Si, ..., S ₆ ) of the pulse converter (4), and wherein this control device (16) has at least one mains voltage distortion controller (22) and a control device (24) for preventing unwanted filter currents, the outputs of which are linked to a summation point (28) and at the inputs thereof a determined mains voltage space vector (u _N ) is present.

4. The device according to claim 3, wherein the control device (16) has a DC voltage regulator (26), which is linked on the output side to the summation parts (28) and at the inputs of which a determined filter current space vector (→ i _κ ), a determined intermediate circuit voltage actual value (Vdc) and a predetermined intermediate circuit voltage setpoint (Vdcsoii) are present.

5. Apparatus according to claim 3, wherein each line voltage distortion controller (22) has an identification device (30) on the input side for determining a complex amplitude (u _N , _{V ±} ) of a line voltage distortion (→ u _{N | V +} ) with a downstream I controller (32) and two multipliers connected in series on the output side

(34, 36), which are connected downstream of the I controller (32), with an imaginary unit (± j) at every second input of the first multiplier (34) and a unit space pointer (36) at the second multiplier (36) e) is present and the mains voltage space pointer (→ u _N ) and a conjugate complex unit space pointer (e *) are present at the inputs of the identification device (30), the unit space pointer (-e> *, → e) rotate in opposite directions with the frequency (ω) of the mains voltage distortion (→ u _{N / v +} ).

6. The device according to claim 3, wherein the control device

(24) has a device (64) for determining a mains voltage basic oscillation co-system space vector (→ u _N , ι ₊ ), which has a comparator (66) with a downstream

Multiplier (68) is connected downstream, a line voltage space vector (→ u _N ) being present at the input of the device (64) and at the non-inverting input of the comparator (66), and the second input of the multiplier (68) having an output of a reciprocal Bildners (70) is linked, at whose input an intermediate circuit voltage actual value (Vd _c ) is present.

7. The device according to claim 4, wherein the DC voltage controller (26) on the input side a device (82) for determining a filter current fundamental vibration Mitsystem-

Space pointer (i _K; ι ₊ ), which is followed by a multiplier (84)

-> is switched, and has an intermediate circuit voltage control circuit (88) and on the output side a multiplier (86), the second input of which is linked to an output of a reciprocal generator (90), and the output of the intermediate circuit voltage - Control loop (88) is linked to the second input of the downstream multiplier (84).

8. The device according to claim 6 or 7, wherein the device

(64, 82) for determining a basic vibration co-system space vector (u _{N #} ι +, iκ, ι +) an identification device for determining a complex amplitude (u _{N / 1 +} , i. _Κ , ι ₊ ) of a basic vibration co-system space vector

(→ u _N , ι ₊ , → iκ, ι ₊ ) with a downstream multiplier (74), with a determined space pointer (→ u _N , → i _κ ) and a conjugate complex at the inputs of the identification device Unit space pointer (e *) and on the second

-

Input of the multiplier (74) is a unit space pointer

(e) are pending, these unit space pointers (e *, e)

→ → - »rotate in opposite directions with the fundamental frequency (ω) of the determined space vector (u _N , i _κ ).

→ →

9. The device according to claim 5 or 8, wherein as the identification device (30) a complex multiplier (38,

72) is provided with a downstream averager (40, 76), with a determined space pointer (→ u _N , → i _κ ) and a conjugate complex unit space pointer (e *) at the inputs of the complex multiplier (38, 72) next

-> hen.

10. The apparatus of claim 7, wherein the intermediate circuit voltage control circuit (88) is preceded by a delay element (96) of the first order and this control circuit (88) has a PI controller (94), the input side with an output of a comparator (92), whose inverting input is connected to the output of this delay element (96), and wherein at the input of this delay element (96) a determined intermediate circuit voltage actual value (V _d c) and at the non-inverting input of the comparator (92) a predetermined intermediate _circuit voltage setpoint V _cso ii are present.

11. The device according to claim 3, wherein a microprocessor is provided as the control device (16).