Optimized control of an arc furnace using a multilevel converter

Abstract

Electrodes (1) of a three-phase arc furnace (2) are supplied with electric energy by means of a multilevel converter (4) via a furnace transformer (3). Output-side phase currents (I1, I2, I3) flowing on the output side of the multilevel converter (4) are known to a control device (15) of the multilevel converter (4). The control device (15) ascertains first output-side phase voltage components (U11*, U12*, U13*) for each current point in time such that a complex first output-side voltage (U1*) defined by the first output-side phase voltage components (U11*, U12*, U13*) is oriented parallel to a complex output-side current (I) defined by the output-side phase currents (I1, I2, I3) on the complex plane. The control device (15) ascertained second output-side phase voltage components (U21*, U22*, U23*) such that a complex second output-side voltage (U2*) defined by the second output-side phase voltage components (U21*, U22*, U23*) is oriented orthogonally to the complex first output-side voltage (U1) on the complex plane. The control device (15) controls the multilevel converter (4) such that the multilevel converter (4) provides output-side phase voltages which correspond to the sum of the first and second output-side phase voltage components (U11*, U12*, U13*, U21*, U22*, U23*). The control device (15) repeatedly carries out the aforementioned steps with a cycle time (TZ).

Description

[0001] Description

[0002] Title of the invention

[0003] Optimized control of an arc furnace using a multilevel converter

[0004] field of technology

[0005] The present invention is based on a control method for a multilevel converter which supplies electrodes of a three-phase arc furnace with electrical energy via a furnace transformer, wherein a control device of the multilevel converter is repeatedly informed of output-side phase currents flowing on the output side of the multilevel converter with a cycle time.

[0006] The present invention further relates to a control program for a control device for controlling a multilevel converter which supplies electrodes of a three-phase arc furnace with electrical energy via a furnace transformer, wherein the control program comprises machine code which can be directly processed by the control device, wherein the processing of the machine code by the control device causes the control device to carry out such a control method.

[0007] The present invention further relates to a control device for controlling a multilevel converter which supplies electrodes of a three-phase arc furnace with electrical energy via a furnace transformer, wherein the control device is programmed with such a control program so that the control device executes such a control method during operation.

[0008] The present invention further relates to a multilevel converter which supplies electrodes of a three-phase arc furnace with electrical energy via a furnace transformer, wherein the multilevel converter is controlled by such a control device.

[0009] State of the art

[0010] The above-mentioned objects are known, for example, from EP 2 329 684 B1.

[0011] EP 2 947 766 A1 discloses a power supply device for a nonlinear load. The load can be an arc furnace. The power supply device is designed as a multilevel matrix converter. Voltages and currents present on the secondary side can be detected and taken into account when determining current setpoints and / or voltage setpoints for the converter units of the power supply device. Similar, but not quite as extensive, disclosures can be found in EP 3 124 903 A1, FR 2 926 182 A1, and CN 110 957 903 A.

[0012] US 2022 / 0 115 888 A1 discloses an arrangement comprising an electrical storage battery on one side and a three-phase network on the other, which are connected to each other via a converter. The converter can be a multilevel converter. The battery can be charged or discharged alternatively. The voltages and currents present on the output side of the converter to the three-phase network are fed to a control device for the converter. Active power and reactive power are determined based on the voltages and currents. Control signals for the converter are determined based on the differences to corresponding target values, and the converter is controlled accordingly. In particular, to discharge the battery, a target voltage can be determined based on the grid-side voltage, which is to be provided by the converter to the three-phase network.The difference between this target voltage and the mains voltage is orthogonal to the mains voltage. Providing the target voltage generates a current that is parallel to the mains voltage.

[0013] Summary of the invention

[0014] In an electric steel mill, scrap is melted using an electric arc furnace. Typically, three electrodes are immersed in a container filled with scrap, and arcs are then generated between the scrap and the electrodes, heating and melting the scrap.

[0015] The voltages applied to the electrodes typically range from several hundred to just over 1 kV, sometimes up to 2 kV. The power of an industrial arc furnace is often between 50 MW and 300 MW.

[0016] The constantly flickering arc causes voltage fluctuations and—if no countermeasures are taken—significant grid disturbances. Historically, it has been common practice to connect compensators in parallel with the arc furnace, such as a so-called SVC (static VAR compensator) or a so-called STATCOM (static compensator) or passive filters. Such compensators are large, heavy, and expensive. Furthermore, they generate power losses and also affect other units connected to the supply grid.

[0017] EP 2 329 684 B1 discloses feeding the arc furnace via a multilevel converter located on the input side of the furnace transformer. With such a converter, it is possible to regulate not only the reactive power on the input side but also the active power consumed by the electric arc furnace. The cited EP document states that, beyond the symmetrical loading of the phases of the supply network with active power, the effects of the arc furnace phases on the supply network are minimized. The exact method of controlling the multilevel converter is not specified in the cited EP document.

[0018] The object of the present invention is to provide such possibilities.

[0019] The object is achieved by a control method having the features of claim 1. Advantageous embodiments of the control method are the subject of dependent claims 2 to 11.

[0020] According to the invention, a control method of the type mentioned at the outset is created in which the control device of the multilevel converter repeatedly carries out the following steps in addition to the already mentioned receiving of the measured values ​​with the cycle time:

[0021] - the control device determines first output-side phase voltage components for the respective current time, so that a complex first output-side voltage determined by the first output-side phase voltage components is oriented in the complex plane parallel to a complex output-side current determined by the output-side phase currents,

[0022] - the control device determines second output-side phase voltage components, so that a complex second output-side voltage determined by the second output-side phase voltage components is oriented in the complex plane orthogonal to the complex first output-side voltage,

[0023] - the control device controls the multilevel converter in such a way that the multilevel converter provides output-side phase voltages corresponding to the sums of the first and second output-side phase voltage components.

[0024] Controlling the multilevel converter so that it provides predetermined phase voltages on the output side is readily possible. The options for such control are known to those skilled in the art. The decisive factor here is determining the output phase voltages so that the complex first and complex second output voltages are oriented parallel and orthogonal, respectively, to the complex output current, as determined by the output phase currents known to the control device.

[0025] The first output-side phase voltage components set the instantaneous active power supplied to the furnace transformer by the multilevel converter. The second output-side phase voltage components set how quickly the instantaneous operating point of the arc furnace changes, as determined by the output-side phase currents known to the control device and the first output-side phase voltage components determined by the control device. Due to the independent determination of the two phase voltage components for each phase, a completely flexible control of the multilevel converter is achieved, which was previously unattainable.

[0026] The first and second output-side phase voltage components are recalculated in each cycle. This allows for control of the arc furnace that can compensate for disturbances very quickly—within 1 ms or less.

[0027] One could refer to the sum of the product of the second output-side phase voltage components with the output-side phase currents as instantaneous reactive power. However, this term is inaccurate and even misleading, since reactive power, like active power, only yields a meaningful value for operation at a fixed frequency and only for integer multiples of half the period (period = inverse of the frequency). In the present invention, the sum of the product of the second output-side phase voltage components with the output-side phase currents, like the instantaneous active power, is an instantaneous value that is only calculated based on instantaneous values ​​of the respective cycle and is only valid for that respective cycle. Furthermore, the determination of the first and second output-side phase voltage components does not necessarily result in periodic control of the multilevel converter.While it is possible that this result will occur, it is not mandatory.

[0028] The two criteria for determining the first and second output-side phase voltage components can be linked or independent of each other as required.

[0029] It is thus possible for the control device to determine the second output-side phase voltage components taking into account the output-side phase currents, but without taking into account the first output-side phase voltage components. For example, the control device may know maximum values ​​for the output-side phase currents. In this case, the control device can determine the second output-side phase voltage components taking into account the distances of the output-side phase currents from their maximum values. It is even better if the control device knows a current domain in the complex plane within which the complex output-side current may lie, and the control device determines the second output-side phase voltage components taking into account the distance of the complex output-side current from the boundaries of the current domain.By defining the current domain, it is possible to take into account the fact that the complex output current must also have a minimum length to ensure a reasonable division into a first parallel and a second perpendicular output voltage is stable. Furthermore, the limitations are often not constant over time, but are based very specifically on the precise design and current operating state of the multilevel converter. The definition of the current domain can therefore also vary over time.

[0030] It is also possible for the control device to determine the second output-side phase voltage components while taking the first output-side phase voltage components into account, but without taking the output-side phase currents into account. For example, the control device may be aware of maximum values ​​for the output-side phase voltages. In this case, the control device can first determine the first output-side phase voltage components and then determine the second output-side phase voltage components while taking into account the distances between the first output-side phase voltage components and the maximum values ​​of the output-side phase voltages.It is even better if the control device knows a voltage range in the complex plane within which a complex total output voltage determined by the output phase voltages may lie, and the control device determines the second output phase voltage components taking into account the distance of the complex first output voltage from the boundaries of the voltage range. Furthermore, the limitations are often not constant over time, but are based very specifically on the current operating state of the multilevel converter. By defining the voltage range dependent on the respective operating state, it is also possible to take into account the fact that the limitations are based very specifically on the precise design of the multilevel converter.

[0031] Finally, it is also possible for the control device to determine the second output-side phase voltage components by taking into account both the output-side phase currents and the first output-side phase voltage components. In particular, the options explained above for considering only the output-side phase currents and only the first output-side phase voltage components can be combined. A mutual dependency may also exist.

[0032] Preferably, the control device knows the maximum values ​​for the output-side phase voltages. In this case, the control device can determine the first output-side phase voltage components taking the maximum values ​​into account.

[0033] It is even better if the control device knows a voltage range in the complex plane within which a complex total output voltage determined by the output phase voltages may lie. In this case, the control device can determine the first output phase voltage components while taking the boundaries of the voltage range into account. Furthermore, the limitations are often not constant over time, but are based very specifically on the current operating state of the multilevel converter. By defining the voltage range dependent on the respective operating state, it is also possible to take into account the fact that the limitations are based very specifically on the precise design of the multilevel converter.

[0034] It is also possible for the control device to know a flux region in the complex plane within which the complex magnetic flux vector of the furnace transformer may lie. In this case, the control device can determine the first and / or second output-side phase voltage components, taking into account the distance of the complex flux vector from the boundaries of the flux region. The magnetic flux vector can be determined by the control device by integrating the space vector for the primary and secondary voltages remaining after deducting resistive and, if applicable, inductive voltage losses, weighted by the respective number of turns. The space vector must adhere to certain values, because otherwise iron saturation would occur. In particular, the absolute value of the magnetic flux vector must not be too large, because otherwise iron saturation would occur in the furnace transformer.Conversely, the magnitude of the magnetic flux vector must not be too small, as this would lead to an excessively high frequency of the output phase voltages or output phase currents.

[0035] The regions mentioned for the complex space vectors of the output voltages, the output current, and the flux are not necessarily simply connected. Rather, it is possible that they contain "holes," i.e., forbidden zones that are completely surrounded by a permitted region in the complex plane. It is also possible that the permitted region for one of the complex space vectors disappears as a surface, for example, degenerates into a closed line. This can be particularly the case with the magnetic flux vector. Furthermore, it is possible that the regions mentioned vary over time. This can happen, for example, when boundary conditions change or new boundary conditions become known.

[0036] It is possible for the control device to receive the output phase currents—that is, the output phase currents utilized within the scope of the control method according to the invention—as measured values. Alternatively, it is possible for the control device to determine the output phase currents using a model of the arc furnace. Such models are known to those skilled in the art.

[0037] Even in the case of model-based determination of the output phase currents, it is possible for the control device to (additionally) receive measured values ​​for the output phase currents. In this case, the model-based determined output phase currents are also used to determine the first and, if applicable, the second output phase voltage components. However, it is possible for the control device to adapt the arc furnace model based on the deviation of the output phase currents determined using the arc furnace model from the measured values.

[0038] Preferably, the control device determines the first output-side phase voltage components such that the sum of the products of the first output-side phase voltage components and the output-side phase currents is equal to an instantaneous output-side target power known to the control device and valid only for the respective cycle. This allows for targeted power control of the arc furnace.

[0039] This approach contrasts with prior art approaches, where control appears to occur over several periods of the grid frequency. This is particularly indicated by the terms "active power" and "reactive power," which are repeatedly used in the prior art—including in EP 2 329 684 B1. The instantaneous output target power can be known to the control device as a corresponding power curve. The power curve can be constant, but can also vary. If it varies, the instantaneous output target power should preferably vary only slowly.

[0040] Typically, the multilevel converter draws input-side phase currents from the input phases of a supply network based on appropriate control by the control device. The supply network operates at a mains frequency. The mains frequency is typically 50 Hz or 60 Hz. Other mains frequencies are also possible. The supply network typically carries a medium voltage of 11 kV, 30 kV, 33 kV, or 110 kV.

[0041] Preferably, the control device repeatedly determines the control of the multilevel converter with the cycle time such that a complex input-side current determined by the input-side phase currents is oriented in the complex plane at a predetermined phase angle relative to a complex input-side voltage determined by the input-side phase voltages of the input-side phases of the supply network, in particular parallel to the complex input-side voltage. As a result, a defined reactive power is drawn from the supply network in addition to the active power, considered over a whole or at least half a period, or – in the case of a parallel orientation – no reactive power is drawn.This makes it possible to draw exclusively active power from the supply grid, each over whole or at least half periods, or to specifically set a draw of reactive power from the supply grid for these periods. Drawing a defined reactive power can, for example, be useful to compensate for the reactive power of other consumers connected to the supply grid. However, over several half periods, the draw of reactive power can also vary over time, preferably slowly. Controlling the multilevel converter so that it draws specified currents from the supply grid on the input side (and, in conjunction with the input phase voltages, also instantaneous power) is readily possible. The options for corresponding control are known to those skilled in the art. The decisive factor in this case is determining the input phase currents.

[0042] Preferably, the control device repeatedly determines the control of the multilevel converter using the cycle time in such a way that the multilevel converter draws instantaneous input-side power from the input-side phases according to the proportions valid for the respective cycle. If the proportions for the input-side phases are specified accordingly over several cycles, a purely symmetrical load on the supply network also occurs. In this case, the proportions for the input-side phases vary with a constant phase offset, in particular without phase offset, with the input-side phase voltages.

[0043] In conjunction with a slowly changing specification of the instantaneous active power on the output side (and possibly also the reactive power on the input side), this enables almost completely flicker-free operation of the arc furnace from the supply grid's perspective. This ultimately makes the arc furnace a manageable—a "tamed" load, so to speak. This applies even if no active or passive compensators are installed on either the input or output side of the multilevel converter. This also applies if extreme operating conditions occur in the arc furnace, for example, if one of the arcs briefly breaks off.

[0044] Preferably, the multilevel converter is designed as an intermediate circuit converter, which has an input-side rectifier connected to a supply network and an output-side inverter connected to the furnace transformer, which are connected to each other via a DC voltage circuit. In this case, the control of the inverter and the rectifier can be determined almost independently.

[0045] The object is further achieved by a control program having the features of claim 12. According to the invention, the processing of the control program by the control device causes the control device to execute a control method according to the invention.

[0046] The object is further achieved by a control device having the features of claim 13. According to the invention, the control device is programmed with a control program according to the invention, so that the control device executes a control method according to the invention during operation. The object is further achieved by a multilevel converter having the features of claim 14. According to the invention, the multilevel converter is controlled by a control device according to the invention.

[0047] Short description of the drawings

[0048] The above-described properties, features, and advantages of this invention, as well as the manner in which they are achieved, will become clearer and more readily understood in connection with the following description of an embodiment, which is explained in more detail in conjunction with the drawings.

[0049] FIG 1 an arc furnace and its power supply,

[0050] FIG 2 an arm of a multilevel converter,

[0051] FIG 3 a submodule,

[0052] FIG 4 another submodule,

[0053] FIG 5 a flow chart,

[0054] FIG 6 a timing diagram,

[0055] FIG 7 a vector diagram for complex output-side electrical quantities,

[0056] FIG 8 a flow chart,

[0057] FIG 9 a complex plane for a complex output current,

[0058] FIG 10 a complex plane for a complex first output voltage,

[0059] FIG 11 a flow chart,

[0060] FIG 12 a flow chart,

[0061] FIG 13 a flow chart,

[0062] FIG 14 a vector diagram for complex output-side electrical quantities and

[0063] FIG 15 an arc furnace and its power supply.

[0064] Description of the embodiments

[0065] According to FIG. 1, electrodes 1 of a three-phase arc furnace 2 are supplied with electrical energy by a multilevel converter 4 via a furnace transformer 3. The arc furnace 2 can be of any type. It can be an electric arc furnace 2 in the narrower sense (EAF = electric arc furnace). However, it can also be a so-called SAF (= submerged arc furnace) or a SMELTER. Iron, steel, aluminum, or another metal can be melted in the arc furnace 2 as required.

[0066] The multilevel converter 4, in turn, is generally supplied with electrical energy from a supply network 5 with several input phases 6 (usually three phases 6). The supply network 5 can, for example, be a medium-voltage network with a nominal voltage of 11 kV, 30 kV, 33 kV, or—in individual cases—110 kV. The supply network 5 operates at a network frequency fN. The network frequency fN is typically 50 Hz or 60 Hz.

[0067] The multilevel converter 4 is arranged between the supply network 5 and the furnace transformer 3, the furnace transformer 3 between the multilevel converter 4 and the electrodes 1. The electrical energy thus flows from the supply network 5 to the multilevel converter 4, from there to the furnace transformer 3 and from there finally to the electrodes 1.

[0068] Specifically, in the embodiment of FIG. 1, the multilevel converter 4 is designed as an intermediate circuit converter, which has an input-side rectifier 7 connected to the supply network 5 and an output-side inverter 8 connected to the furnace transformer 3. The rectifier 7 and the inverter 8 are connected to each other via a DC voltage circuit.

[0069] Multilevel converters are generally known to those skilled in the art. Multilevel converters have arms 9 (see also FIG. 2). Only a few of the arms 9 are provided with their reference symbols in FIG. 1.

[0070] The arms 9, as shown in FIG. 2, in turn consist of a multi-stage series connection of submodules 10. As a rule, there are eight or more such submodules 10 per arm 9. The submodules 10 are constructed identically to one another. As a rule, there is also a choke 11 at the end of the series connection. In an arm 9 of the rectifier 7, the choke 11 is usually located at the end of the respective arm 9 facing an input-side phase 6. Similarly, in an arm 9 of the inverter 8, the choke 11 is located at the end of the respective arm 9 facing an output-side phase 12 (see FIG. 1).

[0071] According to FIG. 3, each submodule 10 comprises a storage capacitor 13 and self-commutated semiconductor switches 14. The term "self-commutated" means that the semiconductor switches 14 can be both switched on and off by control signals supplied externally to the semiconductor switches 14. For example, the self-commutated semiconductor switches 14 can be designed as IGBTs. The term "self-commutated" contrasts with the term "line-commutated." This term means that the respective semiconductor switch can be specifically switched on, but cannot be switched off by an external control signal. An example of a line-commutated semiconductor switch is a "normal" thyristor.

[0072] According to FIG. 3, the submodules 10 each have a single storage capacitor 13 and exactly two semiconductor switches 14. This configuration is the minimum configuration of the submodules 10. Alternatively, the submodules 10 could, for example, have one storage capacitor 13 and four semiconductor switches 14 in a bridge circuit, as shown in FIG. 4. The submodules 10 could also have several individually switchable storage capacitors 13. In this case, at least two semiconductor switches 14 must be present for each independently switchable storage capacitor 13.

[0073] The semiconductor switches 14 of each submodule 10 can be switched independently of the semiconductor switches 14 of the other submodules 10. This applies regardless of whether the other submodules 10 are arranged in the same or in a different arm of the converter 6 as the submodule 10 in question. Depending on the switching state of the semiconductor switches 14 of the respective submodule 10, the storage capacitor 13 of the respective submodule 10 is alternatively bridged or active. If the upper semiconductor switch 14 of a submodule 10 in FIG. 3 is closed and the other semiconductor switch 14 is open, the storage capacitor 13 of the respective submodule 10 is active. Conversely, if the upper semiconductor switch 14 in FIG. 3 is open and the lower semiconductor switch 14 is closed, the storage capacitor 13 of the respective submodule 10 is bridged.When the storage capacitor 13 is active, the voltage drop across the storage capacitor 13 contributes to the resulting voltage of the respective arm 9. A current flowing in the respective arm 9 causes the storage capacitor 13 to be charged or discharged, depending on the direction of the current flowing in the arm 9 and the charge state of the storage capacitor 13. If, however, the storage capacitor 13 is bridged, the voltage drop across the storage capacitor 13 does not contribute to the resulting voltage of the respective arm 9. A current flowing in the respective arm 9 causes neither charging nor discharging of the storage capacitor 13.

[0074] In an analogous manner, the storage capacitor 13 of the submodule 10 of FIG. 4 is also active or bridged depending on the control state of the semiconductor switches 14. However, in the case of the embodiment of FIG. 4, there is a further switching state in which the voltage drop across the storage capacitor 13 contributes with inverse polarity to the resulting voltage of the respective arm 9.

[0075] The multilevel converter 4 is controlled by a control device 15 according to FIG. 1. The control device 15 is used, in particular, to control the semiconductor switches 14 of the submodules 10. The manner in which the semiconductor switches 14 of the submodules 10 are controlled is known to those skilled in the art.

[0076] The control device 15 is programmed with a control program 16. The control program 16 includes machine code 17, which can be directly processed by the control device 15.

[0077] The programming of the control device 15 with the control program 16 or - equivalently - the processing of the machine code 17 by the control device 15 causes the control device 15 to execute a control method which is explained in more detail below in conjunction with FIG. 5. According to FIG. 5, the control device 15 repeatedly executes steps S1 to S6 cyclically. Execution takes place with a cycle time TZ. The cycle time TZ is very short, as shown in FIG. 6. As a rule, the cycle time TZ is less than 1 ms. In particular, the cycle time TZ is considerably shorter than a period time TN = 1 / fN of the supply network 5. Specifically, FIG. 6 shows a single period of one of the voltages present at the input-side phases 6, the associated period time TN and the cycle time TZ.

[0078] Preferably, the inverse of the cycle time TZ is at least twenty times the mains frequency fN, preferably at least fifty times. At a mains frequency fN of 50 Hz, the cycle time TZ is therefore preferably 1.0 ms or less, for example 0.4 ms, 0.2 ms or 0.1 ms. Specifically, for example, a control clock of 8 kHz or 16 kHz can be realized, which corresponds to a cycle time TZ of 125 ps or 62.5 ps. This definition of the cycle time TZ takes into account the fact that for robust and smooth control, several samples are required during which the output phase currents I1, I2, I3, which the multilevel converter 4 supplies to the furnace transformer 3, do not change significantly.

[0079] Analogous statements apply to the ratio of the cycle time TZ to a period time TO of the furnace transformer 3. The period time TO is the inverse of the frequency fO at which the furnace transformer 3 operates. The furnace frequency fO can – purely by chance – have the same value as the grid frequency fN. However, the two frequencies fO and fN are usually different from each other.

[0080] According to FIG 5, the output-side phase currents I1 to I3 are known to the control device 15 in step S1. For example, the control device 15 can receive the output-side phase currents I1 to I3 (more precisely: the corresponding measured values) in step S1 from a measuring device 18 (see FIG 1). The output-side phase currents I1 to I3 are the currents currently flowing on the output side of the multilevel converter 4. As a rule, these are the currents on the input side of the furnace transformer 3. Conversely, if these are the currents on the output side of the furnace transformer 3, the transformation ratio of the furnace transformer 3 can be used to easily convert them to the currents on the input side of the furnace transformer 3. The measured values ​​I1 to I3 are only valid for the respective cycle.

[0081] Strictly speaking, it is sufficient if the control device 15 receives two of the three output-side phase currents I1 to I3, since the relationship

[0082] 11 + 12 + 13 = 0 (1) must apply. Therefore, the remaining output phase current 11 to I3 can always be determined using two of the output phase currents 11 to I3.

[0083] In step S2, the control device 15 determines based on the relationship a complex output current I. FIG 7 shows the complex plane and in the complex plane the complex output current I.

[0084] In step S3, the control device 15 determines first output-side phase voltage components U11*, U12*, U13*. The determination is carried out in such a way that a complex first output-side voltage U1*, which is determined according to the relationship is determined by the first output-side phase voltage components U11*, U12*, U13*, is oriented in the complex plane parallel to the complex output-side current I, which is either already explicitly known to the control device 15 or at least already unambiguously determined by the output-side phase currents I1 to I3 known to the control device 15. FIG. 7 also shows the complex first output-side voltage U1*. The complex first output-side voltage U1* is drawn in FIG. 7 at a very small angle to the complex output-side current I. However, the angle only serves to better distinguish between the complex output-side current I and the complex first output-side voltage U1* in FIG. 7. The angle therefore only serves to improve the illustration. In reality, the angle is 0°.

[0085] It is possible for the control device 15 to first determine the first output-side phase voltage components U11*, U12*, U13* and then to determine the complex first output-side voltage U1*. However, it is generally simpler if the control device 15 first determines the complex first output-side voltage U1* and then, using the relationships

[0086] Ull*= Re[Ul*j , (4)

[0087] U12* = Re[Ul* - exp(-2i^ / 3)1 and (5)

[0088] Ul 3 * = Re [Ul* • exp(2i?z / 3)] . (6) the first output-side phase voltage components U11*, U12*, U13* are determined. Furthermore, in step S4, the control device 15 determines second output-side phase voltage components U21*, U22*, U23*. The determination is carried out in such a way that a complex second output-side voltage U2*, which is determined by the second output-side phase voltage components U21*, U22*, U23*, is oriented in the complex plane orthogonal to the complex output-side current I (and thus also orthogonal to the complex first output-side voltage U1*). FIG. 7 also shows the complex second output-side voltage U2*.

[0089] Just as in step S3, it is possible for the control device 15 to first determine the second output-side phase voltage components U21*, U22*, U23* and then determine the complex second output-side voltage U2*. However, it is generally simpler if the control device 15 first determines the complex second output-side voltage U2* and then determines the second output-side phase voltage components U21*, U22*, U23*. The conversions are completely analogous to the conversions between the first output-side phase voltage components U11*, U12*, U13* and the complex first output-side voltage U1*.

[0090] It is of course also possible for the control device 15 to first add the two complex output-side voltages U1*, U2* in the complex plane and then to determine the output-side phase voltages based on this sum.

[0091] In step S5, the control device 15 determines a control C for the multilevel converter 4. The determination in step S5 is based on the sums of the respective first and second output-side phase voltage components U11* and U21*, U12* and U22*, or U13* and U23*. In step S6, the control device 15 controls the multilevel converter 4 according to the control C determined in step S5. The result is that the multilevel converter 4 provides corresponding output-side phase voltages.

[0092] It is possible for the control device 15 to determine the first output-side phase voltage components U11*, U12*, U13* and the second output-side phase voltage components U21*, U22*, U23* independently of one another. However, dependencies may also exist.

[0093] For example, according to the illustration in FIG. 8, it is possible for the control device 15 to determine the second output-side phase voltage components U21*, U22*, U23* taking into account the output-side phase currents I1, I2, I3. This is illustrated in FIG. 8 in a step S11. Alternatively, it is possible for the control device 15 to determine the second output-side phase voltage components U21*, U22*, U23* taking into account the first output-side phase voltage components U11*, U12*, U13*. This is illustrated in FIG. 8 in a step S12. Furthermore, it is possible for the control device 15 to determine the second output-side phase voltage components U21*, U22*, U23* taking into account both the output-side phase currents I1, I2, I3 and the first output-side phase voltage components U11*, U12*, U13*. This is illustrated in FIG. 8 in step S13.

[0094] Steps S11 to S13 are possible embodiments of step S4 in FIG. 5. Only one of steps S11 to S13 is ever present. Steps S11 to S13 are therefore only shown in dashed lines in FIG. 8. Furthermore, the measures taken in steps S11 to S13 are shown in steps S11 to S13 with respect to the complex output-side currents I and voltages U1*, U2*. This procedure is equivalent to the phase-by-phase procedures. Furthermore, the determination only relates to the signed length of the complex second output-side voltage U2*. This is because its orientation is already predetermined by the condition that it must be orthogonal to the complex first output-side voltage U1* or orthogonal to the complex output-side current I.

[0095] To take the output-side phase currents I1, I2, I3 into account, the control device 15 can, for example, know maximum values ​​for the output-side phase currents I1, I2, I3. If simple, independent maximum values ​​are specified, the maximum values ​​correspond to a hexagon as shown in FIG 9. In this case, the control device 15 can, for example, determine the distances of the output-side phase currents I1, I2, I3 from their maximum values ​​in step S11 or in step S13. The distance 11 in FIG 9 is shown purely as an example for the output-side phase current I1. The control device 15 can then take the distances into account when determining the second output-side phase voltage components U21*, U22*, U23*.

[0096] In an analogous manner, the control device 15 can, for example, know maximum values ​​for the output-side phase voltages in order to take into account the first output-side phase voltage components U11*, U12*, U13*. If simple, independent maximum values ​​are specified, the maximum values ​​correspond to a hexagon as shown in FIG 10. In this case, the control device 15 can, for example, determine the distances of the first output-side phase voltage components U11*, U12*, U13* from the maximum values ​​for the output-side phase voltages in step S12 or step S13. The distance ÖU* in FIG 10 is shown purely as an example for the first output-side phase voltage component U11*. The control device 15 can then take these distances into account when determining the second output-side phase voltage components U21*, U22*, U23*.In particular, when the distances fall below threshold values, the control device 15 can determine a larger respective second output-side phase voltage component U21*, U22*, U23* than when the distances exceed the threshold values. The representation in the respective complex plane for the complex output-side current I and the complex first output-side voltage U1* was used above to explain how maximum values ​​for the output-side phase currents I1, I2, I3 and for the output-side phase voltages can be taken into account. In particular, FIG. 9 shows the current range corresponding to the maximum values ​​for the output-side phase currents I1, I2, I3, within which the complex output-side current I may lie.In an analogous manner, FIG 10 shows the voltage range corresponding to the maximum values ​​for the output-side phase voltages, within which a complex total output voltage determined by the output-side phase voltages may lie. It is also possible for the control device 15 to know the current range and / or the voltage range directly in the respective complex plane. In this case, the same procedure can be used in principle, although it is not the distances of the output-side phase currents I1, I2, I3 or the first output-side phase voltage components U11*, U12*, U13* from their maximum values ​​that are taken into account, but rather the distance of the complex output-side current I from the boundaries of the current range or the distance of the complex first output-side voltage U1* from the boundaries of the voltage range.This approach is particularly advantageous because it allows the control device 15 to specify permissible current or voltage ranges that are not simply defined by a hexagon in the complex plane. This allows, for example, mutual dependencies between the output-side phases 12 to be taken into account.

[0097] The maximum values ​​for the output-side phase voltages can also be taken into account when determining the first output-side phase voltage components U11*, U12*, U13*. In an analogous manner, the voltage range in the complex plane can also be taken into account when determining the complex first output-side voltage U1*. For example, as shown in FIG 11, the control device 15 can first determine the complex first output-side voltage U1* in a step S21 according to a predetermined criterion. The determination in step S21 still takes place without taking the voltage range into account and is therefore only preliminary. In a step S22, the control device 15 can in this case check whether the complex first output-side voltage U1* determined in step S21 lies within the voltage range for the complex total output voltage.If this is not the case, the control device 15 corrects the complex first output-side voltage U1* determined in step S21 in a step S23. For example, scaling can be carried out with a (real) factor a that lies between 0 and 1. In this case, the factor a is determined such that the complex first output-side voltage U1* determined in step S23 lies within the voltage range for the complex total output voltage. If a phase-by-phase determination is carried out instead of the determination in the complex plane, equivalent results are obtained. Steps S21 to S23 thus correspond to an embodiment of step S3 of FIG 5. Various procedures are possible for determining the complex first output-side voltage U1* in step S3 of FIG 5 or in step S21 of FIG 11.Currently, it is preferred that the control device 15 determines the first output-side phase voltage components U11*, U12*, U13* such that the sum of the products of the first output-side phase voltage components U11*, U12*, U13* and the output-side phase currents I1, I2, I3 is equal to an instantaneous output-side target power P* known to the control device 15. This will be explained in more detail below in connection with FIG. 12.

[0098] Similarly, it is also possible for the control device 15 to know a flux region in the complex plane within which a complex flux vector of the furnace transformer 3 may lie. In this case, the control device 15 can determine the first output-side phase voltage components U11*, U12*, U13* and / or the second output-side phase voltage components U21*, U22*, U23*, taking into account the distance of the complex flux vector from the boundaries of the flux region. This is not shown separately in the figures.

[0099] According to FIG 12, the control device 15 becomes aware of the output-side instantaneous target power P* in a step S31. The output-side instantaneous target power P* is only valid for the current cycle. It is possible that the control device 15 knows the time profile for the output-side instantaneous target power P* in advance for a large number of cycles. In this case too, however, the control device 15 determines the current value P* for the respective cycle in step S31 based on the time profile known to it. Regardless of whether an explicit specification of the output-side instantaneous target power P* is made for the respective cycle or a corresponding time profile is specified, the output-side instantaneous target power P* should not change or should change only slowly in terms of time.

[0100] In a step S32, the control device 15 determines the first output-side phase voltage components U11*, U12*, U13* for the output-side phases 12. The determination is only carried out for the respective cycle and thus only for the current time. The determination in step S32 is carried out in such a way that the sum of the products of the first output-side phase voltage components U11*, U12*, U13* and the output-side phase currents I1, I2, I3 is equal to the output-side instantaneous target power P*. The determination in step S3 is therefore carried out in such a way that the relationship

[0101] I1-U11*+I2-U12*+I3-U13* = P* (7) applies. Equivalent to a determination according to equation (7), as shown in FIG 12, is a direct determination based on the associated complex values ​​I, U1*. In FIG 12, £ denotes the conjugate complex value of the complex output-side current I. Due to the fact that, according to step S32, the real part is no longer explicitly formed, it is required that the imaginary part has the value 0. This also ensures that the first complex output-side voltage U1* is oriented parallel to the complex output-side current I. The determined power also does not change if the second complex output-side voltage U2* is present in addition to the first complex output-side voltage U1*. This is because the second complex output voltage U2* does not contribute to the power due to the fact that it is oriented orthogonally to the complex output current I.

[0102] The reference of the input-side phase currents i1, i2, i3 is generally also determined by the control C of the multilevel converter 4. The input-side phase voltages u1, u2, u3, however, are predetermined due to the operation of the supply network 5. They can be known to the control device 15, for example, due to detection by a corresponding measuring device 19 (see FIG. 1).

[0103] To (completely) determine the control C of the multilevel converter 4, the procedure of FIG. 5 according to FIG. 13 is preferably supplemented by additional steps S41 to S43. Steps S41 to S43 are also executed anew in each cycle and are always executed before step S5. They are generally executed after steps S1 to S4.

[0104] In step S41, the control device 15 is informed of the proportions a1, a2, a3 for the input-side phases 6, to which a power p currently drawn from the supply network 5, hereinafter referred to as the instantaneous input-side power p, is to be distributed among the input-side phases 6. The proportions a1, a2, a3 are also only valid for the current cycle.

[0105] Analogous to the instantaneous target power P* on the output side, it is possible for the control device 15 to know in advance the temporal profiles for the components a1, a2, a3 for a plurality of cycles. However, in this case, too, the control device 15 determines the current values ​​a1, a2, a3 for the respective cycle based on the temporal profiles known to it in step S41.

[0106] The instantaneous input-side power p is assumed to be equal to the power currently supplied by the multilevel converter 4 to the furnace transformer 3. Only the losses occurring within the multilevel converter 4 need to be taken into account. If the instantaneous output-side target power P* is specified, the control device 15 can therefore determine the instantaneous input-side power p in step S42 based on the instantaneous output-side target power P*. If the instantaneous output-side target power P* is not specified, the control device 15 can determine the instantaneous input-side power p in step S42 based on the determined complex first output-side voltage U1* and the complex output-side current I. The latter represents the general case and is therefore specified in step S42 using the complex variables.In both cases, it is easily possible to determine the instantaneous input power p in step S42.

[0107] In step S43, the control device 15 determines the input-side phase currents i1, i2, i3. The determination is carried out in such a way that the relationships

[0108] 11 - ul = al- p , (8)

[0109] 12 - u2= a2- p and (9) i3 • u3 = a3 • p (10) apply.

[0110] Analogous to the procedure on the output side of the multilevel converter, the input-side phase currents i1, i2, i3 and the input-side phase voltages u1, u2, u3 can also be calculated using the relationships [il + i2- exp(2i7z / 3) + i3 - exp(-2i?z / 3)] (11) and • [ul + u2 • exp(2i?z / 3) + u3 ■ exp(-2i?z / 3)] (12) be converted into respective complex values ​​i, u. Preferably, the components a1, a2, a3 are determined such that the complex input-side current i is oriented in the complex plane parallel to the complex input-side voltage u, as shown in FIG 14. The very small angle shown in FIG 14 between the complex input-side current i and the complex input-side voltage u serves - analogous to FIG 7 - merely to better distinguish between the complex input-side current i and the complex input-side voltage u in FIG 14. The angle therefore merely serves to improve the illustration. In reality, the angle is 0°.

[0111] The parallel orientation of the complex input-side current i and the complex input-side voltage u arises in particular when the components a1, a2, and a3 exhibit curves that each correspond to the square of a sinusoidal curve, with the respective temporal curve being in phase with the corresponding voltage curve of the respective input-side phase 6. This results, viewed over full or half periods, in a load on the supply network 5 exclusively with active power. Under the additional condition that the instantaneous output-side target power P* does not change or changes only slowly over time, a completely flicker-free load on the supply network 5 also results.If necessary, a constant or time-varying phase angle between the complex input-side current i and the complex input-side voltage u can also be set based on predetermined criteria. This allows a specific reference of reactive power from the supply network 5 to be set, again considered over whole or half periods.

[0112] Above, in conjunction with FIGS. 1 and 5, a procedure was explained in which the control device 15 receives the output-side phase currents I1, I2, I3 as measured values. Alternatively, as shown in FIG. 15, it is possible for the control device 15 to determine the output-side phase currents I1, I2, I3 using a model 20 of the arc furnace 2. Corresponding models are known to those skilled in the art. Otherwise, the design of FIG. 1 remains unchanged, and the procedure of FIG. 5 can also be retained unchanged.

[0113] The present invention offers many advantages. In particular, flicker can be almost completely avoided. This applies even if one of the arcs breaks off briefly. Furthermore, the power supply to the arc furnace 2 can be adjusted very flexibly. In particular, the performance limits of the furnace transformer 3 and the multilevel converter 4 can be fully utilized.

[0114] Although the invention has been illustrated and described in detail by the preferred embodiments, the invention is not limited by the disclosed examples and other variations may be derived therefrom by those skilled in the art without departing from the scope of the invention.

[0115] List of reference symbols

[0116] 1 electrodes

[0117] 2 arc furnaces

[0118] 3 Furnace transformer

[0119] 4 multilevel converters

[0120] 5 Supply network

[0121] 6 input-side phases

[0122] 7 rectifiers

[0123] 8 inverters

[0124] 9 arms

[0125] 10 submodules

[0126] 11 Thrushes

[0127] 12 output phases

[0128] 13 storage capacitors

[0129] 14 semiconductor switches

[0130] 15 Control device

[0131] 16 Control program

[0132] 17 Machine code

[0133] 18, 19 Measuring devices

[0134] 20 Model of the arc furnace a1 , a2, a3 parts

[0135] C Control fN Mains frequency

[0136] I. i complex currents r conjugate complex value

[0137] II , I2, I3 output phase currents i1 , i2, i3 input phase currents p* output instantaneous nominal power p input instantaneous power

[0138] S1 to S43 steps

[0139] TN period time

[0140] TZ cycle time complex voltages first output-side phase voltage components second output-side phase voltage components input-side phase voltages a factor oil, ÖU1* distances

Claims

Claims 1. Control method for a multilevel converter (4) which supplies electrodes (1) of a three-phase arc furnace (2) with electrical energy via a furnace transformer (3), comprising the following steps which are repeatedly executed with a cycle time (TZ): - a control device (15) of the multilevel converter (4) is informed of the output-side phase currents (11, I2, I3) flowing on the output side of the multilevel converter (4), - the control device (15) determines first output-side phase voltage components (U11*, U12*, U13*) for the respective current time, so that a complex first output-side voltage (U1*) determined by the first output-side phase voltage components (U11*, U12*, U13*) is oriented in the complex plane parallel to a complex output-side current (I) determined by the output-side phase currents (I1, I2, I3), - the control device (15) determines second output-side phase voltage components (U21*, U22*, U23*), so that a complex second output-side voltage (U2*) determined by the second output-side phase voltage components (U21*, U22*, U23*) is oriented orthogonally to the complex first output-side voltage (U1*) in the complex plane, - the control device (15) controls the multilevel converter (4) in such a way that the multilevel converter (4) provides output-side phase voltages corresponding to the sums of the first and second output-side phase voltage components (U11*, U12*, U13*, U21*, U22*, U23*).

2. Control method according to claim 1, characterized in that the control device (15) determines the second output-side phase voltage components (U21*, U22*, U23*) taking into account the output-side phase currents (I1, I2, I3) and / or the first output-side phase voltage components (U11*, U12*, U13*).

3. Control method according to claim 2, characterized in that - that the control device (15) knows maximum values ​​for the output-side phase currents (I1, I2, I3) and the control device (15) determines the second output-side phase voltage components (U21*, U22*, U23*) taking into account the distances of the output-side phase currents (I1, I2, I3) from their maximum values ​​and / or - that the control device (15) knows maximum values ​​for the output-side phase voltages and the control device (15) determines the second output-side phase voltage components (U21*, U22*, U23*) taking into account the distances of the determined first output-side phase voltage components (U11*, U12*, U13*) from the maximum values ​​of the output-side phase voltages.

4. Control method according to claim 2, characterized in that - that the control device (15) knows a current domain in the complex plane within which the complex output-side current (I) may lie, and the control device (15) determines the second output-side phase voltage components (U21*, U22*, U23*) taking into account the distance of the complex output-side current (I) from the boundaries of the current domain and / or - that the control device (15) knows a voltage range in the complex plane within which a complex total output voltage determined by the output-side phase voltages may lie, and the control device (15) determines the second output-side phase voltage components (U21*, U22*, U23*) taking into account the distance of the complex first output-side voltage (U1*) from the limits of the voltage range.

5. Control method according to one of the above claims, characterized in that - that the control device (15) knows maximum values ​​for the output-side phase voltages and the control device (15) determines the first output-side phase voltage components (U11*, U12*, U13*) taking the maximum values ​​into account and / or - that the control device (15) knows a voltage range in the complex plane within which a complex total output voltage determined by the output-side phase voltages may lie, and the control device (15) determines the first output-side phase voltage components (U11*, U12*, U13*) taking into account the limits of the voltage range.

6. Control method according to one of the above claims, characterized in that the control device (15) knows a flux region in the complex plane, within which a complex flux vector of the furnace transformer (3) may lie, and the control device (15) determines the first and / or the second output-side phase voltage components (U11*, U12*, U13*, U21*, U22*, U23*) taking into account the distance of the complex flux vector from the boundaries of the flux region.

7. Control method according to one of the above claims, characterized in that the control device (15) receives the output-side phase currents (11, I2, I3) as measured values ​​or that the control device (15) determines the output-side phase currents (11, I2, I3) by means of a model (20) of the arc furnace (2).

8. Control method according to one of the above claims, characterized in that the control device (15) determines the first output-side phase voltage components (U11*, U12*, U13*) in such a way that the sum of the products of the first output-side phase voltage components (U11*, U12*, U13*) and the output-side phase currents (I1, I2, I3) is equal to an output-side instantaneous target power (P*) known to the control device (15).

9. Control method according to one of the above claims, characterized in that the multilevel converter (4) draws input-side phase currents (i1, i2, i3) from input-side phases (6) of a supply network (5) due to a corresponding control by the control device (15), and in that the control device (15) repeatedly determines the control of the multilevel converter (4) with the cycle time (TZ) in such a way that a complex input-side current (i) determined by the input-side phase currents (i1, i2, i3) is oriented in the complex plane at a predetermined phase angle relative to a complex input-side voltage (u) determined by the input-side phase voltages (u1, u2, u3) of the input-side phases (6) of the supply network (5), in particular is oriented parallel to the complex input-side voltage (u).

10. Control method according to one of the above claims, characterized in that the multilevel converter (4) draws input-side phase currents (i1, i2, i3) from input-side phases (6) of a supply network (5) due to a corresponding control by the control device (15) and that the control device (15) repeatedly determines the control of the multilevel converter (4) with the cycle time (TZ) in such a way that the multilevel converter (4) draws input-side instantaneous powers from the input-side phases (6) corresponding to components (a1, a2, a3) valid for the respective cycle for the input-side phases (6).

11. Control method according to one of the above claims, characterized in that the multilevel converter (4) is designed as an intermediate circuit converter which has an input-side rectifier (7) towards a supply network (5) and an output-side inverter (8) towards the furnace transformer (3), which are connected to one another via a DC voltage circuit.

12. Control program for a control device (15) for controlling a multilevel converter (4) which supplies electrodes (1) of a three-phase arc furnace (2) with electrical energy via a furnace transformer (3), wherein the control program comprises machine code (17) which can be processed directly by the control device (15), wherein the processing of the machine code (17) by the control device (15) causes the control device (15) to carry out a control method according to one of the above claims.

13. Control device for controlling a multilevel converter (4), the electrodes (1) of a Three-phase arc furnace (2) is supplied with electrical energy via a furnace transformer (3), wherein the control device is programmed with a control program (16) according to claim 12, so that the control device carries out a control method according to one of claims 1 to 11 during operation.

14. Multilevel converter which supplies electrodes (1) of a three-phase arc furnace (2) with electrical energy via a furnace transformer (3), wherein the multilevel converter is controlled by a control device (15) according to claim 13.

Images