Method for feeding electrical power into an electrical power supply network

Abstract

A method for feeding electrical power into a three phase electrical power supply network at a network connection point, in particular by means of a wind power installation, using an inverter, comprising the following steps: detecting an electrical network voltage, in particular at the network connection point, determining a virtual generator voltage using a machine model that emulates a behavior of a synchronous machine, preparing the detected network voltage for comparison with the virtual genera-tor voltage, predefining a setpoint current as predefinition for an infeed current as a function of the virtual generator voltage and as a function of the network voltage prepared for comparison, and generating the infeed current depending on the setpoint current and feeding the generated infeed current at the network connection point into the electrical power supply network, wherein preparing the detected network voltage for comparison with the virtual generator voltage comprises transforming the detected network voltage into a space vector representation.

Claims

1. A method for feeding electrical power into a three-phase electrical power supply network at a network connection point, by a wind power installation, using an inverter, the method comprising: detecting an electrical network voltage at the network connection point, determining a virtual generator voltage using a machine model that emulates a behavior of a synchronous machine, determining, in the machine model, a rotational speed difference between a virtual rotational speed and a reference rotational speed, setting, in the machine model, the reference rotational speed to a filtered value of the virtual rotational speed or a predefined frequency, wherein a difference rotational speed is determined using a difference rotational speed gain with respect to an auxiliary torque, and wherein the auxiliary torque acts on a virtual moment of inertia of the machine model at a summing point and controls the virtual rotational speed to the reference rotational speed, preparing the detected network voltage for comparison with the virtual generator voltage, predefining a setpoint current as predefinition for an infeed current as a function of the virtual generator voltage and as a function of the preparing network voltage, and generating the infeed current depending on the setpoint current and feeding the infeed current at the network connection point into the electrical power supply network, wherein preparing the detected network voltage for comparison with the virtual generator voltage comprises transforming the detected network voltage into a space vector representation.

2. The method as claimed in claim 1, wherein transforming the detected network voltage into the space vector representation comprises a d/q transformation.

3. The method as claimed in claim 1, wherein the network voltage in the space vector representation is filtered and is subjected to inverse transformation, such that the setpoint current is predefined as a function of the virtual generator voltage and as a function of the inverse-transformed network voltage.

4. The method as claimed in claim 1, wherein the machine model takes as a basis a virtual synchronous machine having a stator and a rotor, and wherein determining the virtual generator voltage involves using at least one variable in a list comprising: a virtual angle of rotation of the rotor, the virtual rotational speed of the rotor, a virtual excitation voltage, a virtual stator current, the virtual moment of inertia, a virtual torque of the rotor, or a virtual friction of the rotor.

5. The method as claimed in claim 4, wherein: a value of the infeed current generated is used as the virtual stator current.

6. The method as claimed in claim 4, wherein the virtual moment of inertia is set as a function of a network state or a network property.

7. The method as claimed in claim 1, wherein a virtual impedance is taken into account for predefining the setpoint current, wherein the virtual impedance is taken into account as an impedance between an output of the machine model or a virtual synchronous machine and the network connection point, and wherein a magnitude of the virtual impedance is variable.

8. The method as claimed in claim 7, wherein the virtual impedance is selected depending on whether: infeed is effected in a normal state of the electrical power supply network, or infeed is effected in a recovery mode after interruption or failure of the electrical power supply network, wherein in the recovery mode, the electrical power supply network is run up to a normal operating point.

9. The method as claimed in claim 1, wherein for synchronizing the machine model with the electrical power supply network, a setpoint power has a value zero, and a calculation model is used for calculating at least one of: internal virtual generator voltages, or a virtual torque, wherein the calculation model uses at least one of: a virtual angle of rotation of a rotor, the virtual rotational speed, a virtual excitation voltage, or the infeed current or the setpoint current, wherein no frequency of the electrical power supply network is detected.

10. The method as claimed in claim 1, further comprising: determining a virtual excitation voltage, wherein the virtual excitation voltage is a function of at least one of: a predefined reactive power, or a predefined electrical network voltage at the network connection point.

11. The method as claimed in claim 1, wherein electrical variables of the machine model are calculated in space vector representation in accordance with a d/q transformation.

12. The method as claimed in claim 1, wherein the infeed current is generated by a tolerance band method.

13. The method as claimed in claim 1, wherein in island network operation, if the inverter predefines a network frequency, the virtual rotational speed is dependent on a predefined network frequency such that the auxiliary torque acting in the machine model is controlled as a function of a difference between the virtual rotational speed and the predefined network frequency, and a virtual excitation voltage is dependent on a predefined voltage of the electrical power supply network and a detected voltage of the electrical power supply network such that an auxiliary reactive power value is controlled as a function of a difference between the predefined voltage and the detected voltage, the virtual excitation voltage being dependent on the auxiliary reactive power value.

14. The method as claimed in claim 13, wherein in the island network operation: the auxiliary torque is controlled by way of a PI controller, and the virtual rotational speed results from an integration of a difference torque, which is a difference between the auxiliary torque and a virtual electrical torque, by way of an integration time constant, the auxiliary reactive power value is controlled by way of the PI controller, and the virtual excitation voltage results from an integration of a difference reactive power, which is a difference between the auxiliary reactive power and a detected reactive power, by way of an integration time constant.

15. A wind power installation comprising a controller configured to carry out the method as claimed in claim 1.

16. A wind farm comprising: a plurality of wind power installations, and a farm inverter coupled to the network connection point and configured to carry out the method as claimed in claim 1.

17. The method as claimed in claim 1, wherein the difference rotational speed is set to zero for synchronizing the machine model with the electrical power supply network.

18. A wind power installation for feeding electrical power into a three-phase electrical power supply network at a network connection point, the wind power installation comprising: an inverter for generating an infeed current, a voltmeter for detecting an electrical network voltage at the network connection point, and a controller for controlling feeding electric power into the electrical power supply network at the network connection point, wherein the controller is configured to: determine a virtual generator voltage using a machine model that emulates a behavior of a synchronous machine, prepare the detected network voltage for comparison with the virtual generator voltage, and define a setpoint current as predefinition for the infeed current as a function of the virtual generator voltage and as a function of the network voltage, wherein the inverter is configured to generate the infeed current depending on the setpoint current and to feed the infeed current at the network connection point into the electrical power supply network, wherein determining the detected network voltage for comparison with the virtual generator voltage includes transforming the detected network voltage into a space vector representation, wherein in the machine model, a rotational speed difference is determined between a virtual rotational speed and a reference rotational speed, and the reference rotational speed is set to a filtered value of the virtual rotational speed or a predefined frequency, wherein a difference rotational speed is determined using a difference rotational speed gain with respect to an auxiliary torque, and wherein the auxiliary torque acts on a virtual moment of inertia of the machine model at a summing point and controls the virtual rotational speed to the reference rotational speed.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

(1) The invention is explained in greater detail by way of example below with reference to the accompanying figures.

(2) FIG. 1 shows a wind power installation in a perspective illustration.

(3) FIG. 2 shows a wind farm in a schematic illustration.

(4) FIG. 3 shows an equivalent circuit diagram for power transfer between two AC voltage sources.

(5) FIG. 4 shows a schematic structure diagram of an infeed method in accordance with one embodiment for network operation.

(6) FIG. 5 shows a structure diagram of a method in accordance with one embodiment in the case of island network operation.

(7) FIG. 6 shows a structure diagram for illustrating an adaptation of a virtual impedance in accordance with one embodiment.

(8) FIG. 7 shows a structure diagram for illustrating a proposed voltage filtering process in accordance with one embodiment.

(9) FIG. 8 shows voltage profiles for elucidating a filter effect.

DETAILED DESCRIPTION

(10) FIG. 1 shows a wind power installation 100 comprising a tower 102 and a nacelle 104. A rotor 106 comprising three rotor blades 108 and a spinner 110 is arranged on the nacelle 104. The rotor 106 is caused to effect a rotational movement by the wind during operation and thereby drives a generator in the nacelle 104.

(11) FIG. 2 shows a wind farm 112 comprising for example three wind power installations 100, which can be identical or different. The three wind power installations 100 are thus representative of basically an arbitrary number of wind power installations of a wind farm 112. The wind power installations 100 provide their power, namely in particular the generated current 113, via an electrical farm network 114. In this case, the respectively generated currents or powers of the individual wind power installations 100 are added and a transformer 116 is usually provided, which steps up the voltage in the farm in order then to feed it into the power supply network 120 at the infeed point 118, which is also generally referred to as PCC. FIG. 2 is merely a simplified illustration of a wind farm 112, which for example does not show a controller, even though a controller is present, of course. Moreover, by way of example, the farm network 114 can be configured differently, with for example a transformer also being present at the output of each wind power installation 100, to mention just one different exemplary embodiment.

(12) The proposed method is a current control method of a full converter during island and network operation on the basis of the equation of motion of a synchronous machine. The invention thus relates to a current control method of a full converter, the core control of which is based on the equation of motion of a synchronous machine. This control makes it possible to emulate inter alia the moment of inertia of a synchronous machine in the case of network state changes, in particular changes in the network frequency and the network voltage, in a manner as freely settable as possible. This property is also dependent on an energy stored in the DC link circuit of the full converter. Consideration is also given to an energy fed into the link circuit by way of a link circuit voltage control by means of a primary energy source. Said energy can then likewise be available in the DC link circuit.

(13) The method is also explained in greater detail below by way of example on the basis of a number of examples.

(14) In an AC system, the transferred active and reactive powers between two nodes having the voltages V and E are determined by means of the following equations:

(15) $\begin{array}{cc}P=\frac{V*E*\sin \delta }{X}& \left(1\right)\\ Q=\frac{E^2-E*V*\cos \delta }{X}& \left(2\right)\end{array}$
wherein X is the line reactance between the two nodes and δ is the phase shift between the two voltages. That is illustrated in the equivalent circuit diagram in FIG. 3.

(16) Assuming that E is the output voltage of a full converter connected to the voltage V at the network, equations (1) and (2) make it clear that the transferred active and reactive powers can be influenced by the amplitudes and phase angle of the output voltage of the converter.

(17) The control consists of an outer and inner control loop. In the outer control loop, the amplitude and the angle of the rotor voltage of a virtual synchronous machine are varied dynamically such that the active and reactive powers of a full converter that are fed into the network are corrected to predefined setpoint values. The active power control is based on the equation of motion of a synchronous machine, while the reactive power or voltage control is based on a P controller or PI controller.

(18) $\begin{array}{cc}\frac{d{\omega }_r}{\mathrm{dt}}=\frac{T_m-T_e-K_D\Delta {\omega }_r}{J}& \left(3\right)\end{array}$

(19) Equation (3) describes the equation of motion of a synchronous machine, wherein T.sub.m and T.sub.e respectively represent the mechanical and electromagnetic torque, ω.sub.r represents the rotor speed, K.sub.D represents the damping factor, and J represents the inertia of the synchronous machine.

(20) T.sub.m is determined by means of an active power setpoint value, as will also be explained with respect to FIG. 4, while K.sub.d and J are settable parameters. The variable K.sub.D can also be referred to as gain D.sub.P.

(21) The electrical variables of the virtual synchronous machine are transformed into d/q components with respect to the rotor coordinate system and are used for the calculation of Te. The d/q transformation may also be referred to as qd transformation or transformation into qd coordinates and, for further explanations regarding the transformation, reference is made to the literature reference [Lit1] cited below. The angle θ resulting from the solution of (3) and integration of ω.sub.r is used for the transformation.

(22) The abc/qd transformation matrices are defined as follows:

(23) $\begin{array}{cc}\left[T_1\right]=\frac{2}{3}\left[\begin{array}{ccc}\cos \theta & \cos \left(\theta -\frac{2\pi }{3}\right)& \cos \left(\theta +\frac{2\pi }{3}\right)\\ \sin \theta & \sin \left(\theta -\frac{2\pi }{3}\right)& \sin \left(\theta +\frac{2\pi }{3}\right)\\ \frac{1}{2}& \frac{1}{2}& \frac{1}{2}\end{array}\right]& \left(4\right)\\ {\left[T_1\right]}^{-1}=\left[\begin{array}{ccc}\cos \theta & \sin \theta & \frac{1}{2}\\ \cos \left(\theta -\frac{2\pi }{3}\right)& \sin \left(\theta -\frac{2\pi }{3}\right)& \frac{1}{2}\\ \cos \left(\theta +\frac{2\pi }{3}\right)& \sin \left(\theta +\frac{2\pi }{3}\right)& \frac{1}{2}\end{array}\right]& \left(5\right)\end{array}$

(24) In the case of a cylindrical-rotor machine, T.sub.e is calculated as follows:

(25) $\begin{array}{cc}{T}_e=\frac{3}{2}{\mathrm{pL}}_m{i}_f{i}_q& \left(6\right)\end{array}$

(26) In this case, p represents the number of pole pairs, L.sub.m represents the mutual inductance, i.sub.f represents the excitation current, and i.sub.q represents the q component of the stator current, wherein the portion L.sub.mi.sub.f results from the reactive power and voltage control.

(27) The amplitude of the virtual rotor voltage in qd components is calculated from:
E.sub.q=ω.sub.rL.sub.mi.sub.f  (7)
E.sub.d=0  (8)

(28) θ and (5) are used to effect inverse transformation of (7) and (8) into abc. It should be noted that in this case, the voltage angle θ is not obtained from the network voltage separately by means of a method such as a phase locked loop, for example, but rather results from the solution of the above-described system of equations, in particular equation (3) of motion. This is a crucial advantage of the method class of control methods based on synchronous machines, since delays such as are customary with the order of magnitude of tens of ms in the case of phase locked loops and similar methods are avoided in this case.

(29) The inner control loop consists of a current hysteresis controller, which can also be referred to as tolerance band method, the setpoint values of which result from the instantaneous values of the virtual rotor voltages e.sub.abc, measured voltages at the converter terminals v.sub.abc, and a virtual stator impedance Z.sub.s.

(30) The virtual stator impedance can be set such that the control stability is ensured, and consists of the following equation:
Z.sub.s=R.sub.s+sL.sub.s  (9)
wherein R.sub.s and L.sub.s represent the stator resistance and the stator inductance of the virtual synchronous machine.

(31) Consequently, the current setpoint values are calculated from:

(32) $\begin{array}{cc}{i}_{\mathrm{sabc}}=\frac{e_{\mathrm{abc}}-v_{\mathrm{abc}}}{Z_s}& \left(10\right)\end{array}$

(33) The measured terminal voltages of the converter are not necessarily sinusoidal and, depending on the mode of operation or network state, may contain many harmonics, which may have a negative effect on the current setpoint values in accordance with equation (10) and the entire control. This will be the case for example in open-circuit operation, i.e., when a voltage is impressed without a load, or in the event of highly nonlinear loads. In principle, the always symmetrical rotor voltage of the system of equations for a synchronous generator counteracts this asymmetry and harmonic content, but it cannot always completely compensate therefor. In order to counteract this behavior, the converter terminal voltages are transformed into qd coordinates, and their components are filtered by a first-order low-pass filter and subjected to inverse transformation into abc components again. Consequently, no amplitude and phase compensation is required. That is illustrated in the equivalent circuit diagram in FIG. 3.

(34) An explanation is given below of network operation which involves feeding into an electrical power supply network, such as, e.g., the European interconnected grid. That may be different from feeding into an island network, as will also be described below.

(35) Since the control emulates the properties of a synchronous machine, the converter is able to synchronize itself with the electrical power supply network, which may also be referred to simply as network. Before the converter is connected to the network, a synchronization phase takes place.

(36) The active and reactive power setpoint values are set to zero and the control ensures that the conditions according to (11) are met. That means that the amplitude and phase angle of the virtual rotor voltages e.sub.abc are equal to those of the network voltages v.sub.abc at the point of common coupling (PCC). Consequently, the active and reactive powers in the settled state, according to equations (12) and (13), are zero at the PCC.

(37) $\begin{array}{cc}\{\begin{array}{c}E=V_g\\ \theta ={\theta }_g\end{array}& \left(11\right)\\ P=\frac{V_g*E*\sin \left(\theta -{\theta }_g\right)}{X_s}=0& \left(12\right)\\ Q=\frac{V_g\left[E*\cos \left(\theta -{\theta }_g\right)\right]}{X_s}=0& \left(13\right)\end{array}$

(38) Since the converter is not yet connected to the network during the synchronization phase, the currents according to equation (10) are virtual. At the end of the synchronization phase, if equation (11) is satisfied, this results in the following:
e.sub.abc−v.sub.abc=0.fwdarw.i.sub.sabc=0.fwdarw.i.sub.q=0.fwdarw.T.sub.e=0  (14)

(39) When the synchronization phase is concluded, the converter can be connected to the network and P.sub.set and Q.sub.set can be set to a new setpoint value. The structure proposed for network operation is illustrated in FIG. 4.

(40) The setpoint current i.sub.sabc is intended to be predefined by the method. Said setpoint current i.sub.sabc is 3-phase and it results from a difference voltage U.sub.DV present at the virtual impedance Z.sub.V. In the structure diagram in FIG. 4, that is illustrated such that the difference voltage U.sub.DV is input into said virtual impedance.

(41) The difference voltage U.sub.DV results as a difference from the measured voltage V.sub.PCC subtracted from the virtual generator voltage E.sub.gen. An output summing element 10 is provided for this purpose. The voltage V.sub.PCC is the voltage measured at the network connection point, which voltage is filtered, however, namely preferably in the manner as will also be explained with reference to FIG. 7.

(42) The generator voltage E.sub.gen is calculated in the calculation model 12 and output by the latter. For this purpose, the calculation model 12 uses machine equations of the synchronous generator that have been described above. The difference between the generator voltage E.sub.gen and the measured voltage V.sub.PCC at the network connection point thus acts on the virtual impedance Z.sub.V, from which the setpoint current i.sub.sabc is calculated. That basically means that an impedance represented by the virtual impedance Zv is present between the generator voltage and the voltage at the network connection point.

(43) The calculation model 12 has the virtual angle of rotation δ of the rotor and a virtual rotational speed ω of the rotor as input variables. These two variables, which in this respect concern or are mechanical variables, are ultimately dependent on an active power, in particular a predefined active power.

(44) Generating the setpoint current L.sub.sabc as a function of the virtual impedance Z.sub.V and the difference voltage U.sub.DV is based on equation (10) explained above. In this respect, the generator voltage E.sub.gen corresponds to the generator voltage e.sub.abc in equation (10) and the measured and filtered voltage V.sub.PCC corresponds to the voltage v.sub.abc in equation (10). The virtual impedance Z.sub.V corresponds to the impedance Z.sub.S in equation (10).

(45) Moreover, the virtual excitation voltage U.sub.e and the output current I.sub.abc enter into the calculation model 12. The output current I.sub.abc can be a measured current, namely in particular the 3-phase output current that was generated by the inverter in accordance with the setpoint current I.sub.sabc.

(46) It is pointed out moreover that FIG. 4 and FIG. 5 contain a virtual machine model and any variables which relate to a synchronous machine therein are thus variables of this virtual synchronous machine and should thus generally be regarded as virtual variables, even if this is not specifically explained.

(47) Moreover, the calculation model 12 outputs a virtual electrical torque T.sub.e. During operation, this virtual electrical torque T.sub.e counteracts a mechanical torque T.sub.m at the torque summing element 14. An effective torque T.sub.w results from the torque summing element 14. The mechanical torque T.sub.m results from a power, which can be predefined here as setpoint power P.sub.s. By way of the torque conversion 16, which substantially takes account of the rotational speed, the mechanical torque T.sub.m results from the setpoint power P.sub.s that is input.

(48) The effective torque T.sub.w is divided by the virtual moment of inertia J, which is realized by the inertia gain 18, and then leads to the rotational speed ω by way of the first mechanical integrator 21. The rotational speed ω leads to the angle of rotation δ of the rotor by way of the second mechanical integrator 22. These two mechanical integrators 21 and 22 thus substantially form again the mechanical behavior of the rotor on which the effective torque T.sub.w acts.

(49) In addition, a torque controller having a torque controller gain 24 is also provided. The torque controller gain 24 leads to an auxiliary torque T.sub.h, which is taken into account with negative signs in the torque summing element 14 and in this respect reduces the effective torque T.sub.w as long as the auxiliary torque itself has a positive value.

(50) For this purpose, the result of the rotational speed summing element 26 acts on the torque controller gain 24. In the rotational speed summing element, in accordance with the configuration shown in FIG. 4, a filtered rotational speed ω* is subtracted from the virtual rotational speed ω. The result is the difference rotational speed Δω.

(51) With this torque control, the rotational speed ω can be influenced in particular for a synchronization. As long as the rotational speed ω changes, a difference between the rotational speed ω and the filtered rotational speed ω* can result. For this purpose, provision is made, in particular, for the rotational speed filter 28 to have a gain of 1 and thus for the filtered rotational speed ω* to correspond to the rotational speed ω in the steady state. In said steady state, the difference rotational speed Δω 0 will then also be 0 and the auxiliary torque T.sub.h will thus also be 0. If, in the synchronization, the predefined power P.sub.s is then also 0 and the virtual electrical machine overall runs in such steady-state open-circuit operation, the virtual electrical torque T.sub.e, then also becomes 0 and thus the effective torque T.sub.w then also becomes 0. The rotational speed ω then no longer changes. A synchronization is then concluded and, e.g., a connection can then be effected for feeding into the electrical power supply network via the network connection point.

(52) For island network operation, in particular, a frequency predefinition can be effected by way of the frequency predefinition block 30, which is provided for island network operation, in particular. Island network operation will also be described below in association with FIG. 5.

(53) The virtual excitation voltage U.sub.e results from the integration of an effective reactive power Q.sub.w taking account of a reactive power gain G, which may also be referred to as integration time constant. The reactive power gain block 32 and the reactive power integrator 34 are provided for this purpose.

(54) The effective reactive power Q.sub.w is the result of the difference between predefined reactive power Q.sub.s and measured reactive power Q.sub.i. The difference is formed at the reactive power summing element 36. The measured reactive power Q.sub.i is the reactive power fed in at that moment by the inverter. The virtual excitation voltage U.sub.e is thus set or influenced by way of the reactive power.

(55) A voltage control can be activated or switched on by way of a voltage control switch 38. A control to a predefined voltage V.sub.PCCS can be performed by said voltage control. At the very least a voltage control can be performed as a function thereof. For this purpose, at the network connection point a voltage V.sub.PCCI measured there is subtracted from said setpoint voltage V.sub.PCCS. The voltage summing element 40 is provided for this purpose. For the voltage control, the difference voltage ΔV thus obtained is passed by way of a voltage gain D.sub.q in the voltage gain block 42. This results in a control reactive power Q.sub.R, which influences the effective reactive power Q.sub.w by way of the reactive power summing element 36 and thus the virtual excitation voltage U.sub.e is then influenced by way of the reactive power gain block 32 and the reactive power integrator 34. This control can be activated by closing the voltage control switch 38. Preferably, the voltage gain D.sub.q of the voltage gain block 42 is also variable in order to influence in particular the dynamic range of this voltage control.

(56) During island network operation, the frequency and the voltage of the network are corrected to desired setpoint values. The frequency control is still based on equation (3) with the difference that T.sub.m is determined by a PI controller, which reacts to a frequency deviation, and is designated as T.sub.R in FIG. 5. In a similar manner to the frequency control, the voltage control is also effected by means of a PI controller, from the output of which the reactive power setpoint value results.

(57) The corresponding control enables the converter to have a black start capability. The property is achieved by a small resistive or inductive load being connected in parallel with the converter terminals if the open-circuit currents flowing through the network filter of the converter or open-circuit currents of other operating equipment are too low for satisfactory use of a hysteresis current controller. A minimum converter output current can thus form, such that the functionality of the hysteresis control is ensured. This results in a voltage at the connecting terminals of the converter, the frequency and amplitude of which are corrected to the desired setpoint values.

(58) Island network operation is illustrated in FIG. 5, which shows a structure like that in FIG. 4. In so far as the structure or the elements in both figures is/are identical or has/have the same function, identical designations and reference signs are used. For explanation of the functions which have not changed for island network operation, reference is also made in this respect to the explanation concerning FIG. 4.

(59) In island network operation, then, an additional PI power controller portion 50 is provided for a power control, which basically concerns the mechanical movement of the rotor. Proceeding from a difference rotational speed Δω, said portion generates a controller torque T.sub.R, which influences the effective torque T.sub.w by way of the torque summing element 14.

(60) In this case, the difference rotational speed Δω results as the difference between the rotational speed ω of the machine model and a setpoint rotational speed ω.sub.s, which is predefined by the frequency predefinition block 30. The frequency predefinition block 30 predefines the frequency for the island network and predefines here in this respect a setpoint frequency, converted into a setpoint rotational speed ω.sub.s. The difference is formed in the rotational speed summing element 26. A filtered rotational speed ω* in accordance with the structure in FIG. 4 is not used here. However, the structure in accordance with FIG. 4 can also be used in so far as, for island operation, switching from the rotational speed filter 28 to the frequency predefinition block 30 is effected, as is indicated by the frequency changeover switch 29 in FIG. 4. As a result, a temporary island network situation can also be taken into account, i.e., if not just a use in an island network is implemented in principle, such as, e.g., on an actual small geographic island, rather if, on account of switching acts in a larger power supply network, a partial area of this power supply network develops into an island network, that is to say is temporarily decoupled.

(61) By way of switching on the control torque T.sub.R, the rotor movement or virtual rotor movement of the virtual synchronous machine can thus be controlled such that the predefined frequency or the predefined setpoint rotational speed ω.sub.s is established. Taking account of a setpoint power P.sub.s by way of the torque conversion 16 is initially not provided here. It can be selected, if appropriate, by way of the setpoint power switch 52. However, the virtual electrical torque T.sub.e still acts on the effective torque T.sub.w by way of the torque summing element 14. Specifically, in this respect the machine model has remained unchanged, but a different control has been implemented. For the voltage control, island network operation in accordance with the structure in FIG. 5 likewise provides for using a PI portion, namely in the form of PI voltage control portion 54. The PI voltage control portion 54 outputs a control reactive power Q.sub.R, which here is given the same designation as in FIG. 4 for the sake of simplicity. However, the control reactive power Q.sub.R is now the output of the PI voltage control portion 54. Said control reactive power Q.sub.c thus contains an integral portion. A difference between the predefined voltage V.sub.PCCS at the network connection point and a voltage V.sub.PCCI measured there once again forms the input of said PI voltage portion 54. The effective reactive power Q.sub.W is now the result of the difference between the control reactive power Q.sub.R and the measured reactive power Q.sub.i. A predefined reactive power Q.sub.S is not effective because the reactive power switch 56 effects switching to the output of the PI voltage control portion 54.

(62) What is achieved, then, by way of the PI voltage control portion 44 is that this substantially results in a reactive power being predefined which is dependent on the voltage deviation between predefined voltage and measured voltage at the network connection point. The integral portion of the PI voltage portion 54 is provided for achieving a steady-state accuracy of the predefined voltage at the network connection point.

(63) For switching from black start operation to load operation, that is to say in the island network, and vice versa, an adaptation of the virtual impedance is carried out. This is proposed since during operation without significant loads a relatively great low-pass filtering of the difference between the virtual rotor voltage and the measured terminal voltage would be necessary in order to ensure stability.

(64) The adaptation of the virtual impedance is illustrated in FIG. 6. The adaptation when switching from black start operation to load operation is based on the measured output power of the converter, which can concern active and/or reactive power, or alternatively on the measured power gradients dP/dt and/or dQ/dt, and is designated as condition 1, or condition C1, in FIG. 6. In this case, the inductive portion Ls of the virtual impedance Zs is altered and this alteration is limited by a maximum gradient, which can be referred to as rate limiter.

(65) During the transition from load operation to black start operation, the load impedance becomes significantly higher. Since the setpoint value currents do not change during the transition, that leads to a significant transient increase in the terminal voltage. If the difference between the amplitudes of the virtual rotor voltage and the measured terminal voltage exceeds a specific limit, which is depicted as condition 2 or condition C2 in FIG. 6, the virtual impedance of the control is changed.

(66) FIG. 6 shows in this respect an implementation of the virtual impedance Zv of the structures from FIGS. 4 and 5 in that the value of the inverse of said virtual impedance Zv is multiplied by the difference voltage U.sub.DV at the multiplier 60, thus resulting in the setpoint current T.sub.abc. The 3-phase setpoint current I.sub.abc can also be referred to as setpoint currents on account of its 3-phase nature, namely one current per phase. The same applies to other 3-phase variables.

(67) The virtual impedance Zv, which can also be referred to as Zs, is composed of an inductive part Ls and a resistive portion Rs and these variables thus form an input for the impedance block 62, in which these two portions are combined and the inverse of the impedance is formed and output for the further calculation.

(68) In accordance with the embodiment shown, the inductive portion Ls is altered, the resistive portion Rs remaining unchanged. However, the latter, too, can be altered in principle.

(69) The change or setting of the inductive portion Ls is then dependent, in particular, on whether black start operation or load operation is present. Accordingly, depending on these operating modes, the operating mode switch 64 can switch over between the two inductive portions, namely the inductive portion L.sub.S1 for black start operation and L.sub.S2 for load operation. For this purpose, the two conditions C1 and C2, respectively, are input into the operating mode switch 64. Since this switchover can be performed in the course of operation, but the virtual inductance Z.sub.V or Z.sub.S, respectively, is intended not to be changed abruptly, a gradient block 66 is provided, which ensures that the change is passed on as a ramp having a maximum gradient or that the change is limited to such a slope having a maximum gradient. In this respect, such a maximum gradient is limited in terms of absolute value. In other words, both a rise and a fall are limited. In other words, if the operating mode switch 64 effects switching between two values of the inductive portions and the output of the operating switch 64 is a jump, then the output of the gradient block 66 is a ramp.

(70) As a result, therefore, the virtual impedance Z.sub.V or Z.sub.S changes gradually and this change can thus be carried out in the course of operation and acts directly on the setpoint current I.sub.abc by way of the multiplier 60.

(71) However, consideration is also given to the fact, particularly in black start operation, that not only is a switchover effected between the two inductive portions L.sub.S1 and L.sub.S2, respectively, but also the inductive portions effective in each case depending on the switch position of the operating mode switch 64 is altered for its part. This alteration, too, can be limited to a maximum gradient, that is to say a maximum edge steepness, by way of the gradient block 66.

(72) FIG. 7 then illustrates the filtering of the voltage V.sub.PCC measured at the network connection point. The measured voltage is designated here as V.sub.PCC′, whereas the filtered result is then the voltage V.sub.PCC. The measured voltage V′.sub.PCC is input into the transformation block 70, which transforms this 3-phase voltage into a fundamentally known representation having a q component and a d component. The transformation is also known as d/q transformation and concerns the same transformation which, e.g., is also known as Park transformation. It converts the 3-phase variables, here namely the voltage of the three phases, into a two-axis coordinate system having the axes d and q. For illustration purposes, a q block 72 comprises the q component and a d block 74 comprises the d component. In order to clarify that these two components still belong to the unfiltered voltage signal V′.sub.PCC, they are likewise illustrated as variables including a prime symbol, namely as q′ and d′. They then respectively enter a filter block, namely the q filter block 76 and the d filter block 78. Both filter blocks 76 and 78 can be identical, but also different. The structure proposed in FIG. 7 uses in each case two identically parameterized first-order linear filters, that is to say in each case a PT1 element or PT1 filter. Both components q′ and d′ are thus passed via a first-order low-pass filter and the result is then respectively the filtered component q and d. These filtered components q and d are then passed to the inverse-transformation block 80 and subjected to inverse transformation into the 3-phase system. The result is the filtered 3-phase voltage V.sub.PCC, which in particular in FIGS. 4 and 5 are passed to the output summation element 10.

(73) FIG. 8 shows two voltage diagrams, each showing a voltage amplitude U in volts against time in seconds. The illustrated voltages thus substantially show in each case a voltage having a sinusoidal profile having a period duration of 20 ms and thus a voltage of a 50 Hz signal.

(74) The upper illustration in FIG. 8 illustrates the voltage V′.sub.1,PCC of a phase that was recorded at the network connection point. That therefore corresponds to a phase of the 3-phase voltage V′.sub.PCC at the input of the transformation block 70 in FIG. 7. It should be noted that the results in FIG. 8 are simulation results. It is also pointed out that an electrical voltage is represented using in some instances the letter U and in some instances the letter V, without this being intended to represent any technical difference.

(75) The upper diagram in FIG. 8 thus shows an unfiltered voltage, which is also evidently severely noisy.

(76) The lower diagram in FIG. 8, the time axis of which diagram is identical to that of the upper diagram, illustrates two filtered voltages. The voltage V.sub.1,PCC shows the filtered profile with respect to the unfiltered voltage V′.sub.1,PCC of the upper diagram, which profile results from the filtering in accordance with FIG. 7. The voltage V.sub.1,PCC is thus the phase—corresponding to the upper diagram in FIG. 8—of the 3-phase voltage signal V.sub.PCC at the output of the inverse-transformation block 80 in FIG. 7. In other words, if a 3-phase voltage signal is input into the transformation block 70 in FIG. 7, with all three phases having approximately a profile as shown in the upper diagram in FIG. 8, a 3-phase voltage signal results at the output of the inverse-transformation block 80 in FIG. 7, all three phases of which voltage signal have approximately the profile V.sub.1,PCC in accordance with the lower illustration in FIG. 8. In this case, the unfiltered voltage V′.sub.1,PCC and the filtered voltage V.sub.1,PCC relate to the same phase.

(77) It is evident that practically no phase shift occurs between the unfiltered voltage V′.sub.1,PCC and the corresponding filtered voltage V.sub.1,PCC.

(78) For the filtering, a PT1 element having the same gain, namely 1, and the same time constant, namely 10 milliseconds, was used in each case for the two filter blocks 76 and 78 in accordance with FIG. 7.

(79) As comparison, the same noisy signal V′.sub.1,PCC from the upper illustration was filtered directly via a PT1 filter which was parameterized in exactly the same way as the two filter blocks 76 and 78. The result of this direct filtering by means of a PT1 filter is illustrated in a dashed manner in the lower diagram and designated as V.sub.PT1R. Besides the reduced amplitude, which could certainly be adapted by adapting the gain of the PT1 filter, a phase shift of almost 90 degrees is also clearly discernible. That actually also reflects the phase behavior known from a PT1 filter.

(80) It is thus evident that the proposed filtering, which not only is adapted to the 3-phase signal in a targeted manner but also in particular takes account of the fact that a sinusoidal signal is expected, yields a very good filter result. In particular, the phase fidelity should be emphasized. With the PT1 filter used internally here, a direct filtering with similar filter quality, i.e., similar noise suppression, would be possible only with a large phase shift or extensive phase lagging. Alternatively, a significantly more complex filter of higher order, in particular, could be used to reduce in particular the problem of the phase delay. However, this filter would then have to be configured with much greater complexity, and be adapted very accurately to the expected noise behavior, and would then probably not be very robust vis-à-vis changes in the signal to be filtered, in particular vis-à-vis superposed interference variables.

(81) A power control of a converter in network operation, for both strong and weak networks, and a converter having black start capability in island network operation are thus proposed. A voltage and frequency control during island network operation has also been presented here.

(82) In principle, the control method is suitable for use with various primary energy forms, thus, e.g., wind power installations, battery storage devices, flywheels and further forms. Furthermore, the self-commutated design of an installation power supply network such as, e.g., in the case of an uninterruptible power supply (UPS) is also possible.

(83) A possibility of emulating a synchronous machine using a converter has thus been afforded. That is also based on the following insights. Synchronous machine emulation is particularly important with regard to increasing the proportion of loads controlled by power electronics and generator units which can lead to an energy supply system based on power electronics. Such a system can entail network instabilities since, e.g., in the event of network frequency changes, the moment of inertia or the stored rotational energy of conventional power plant generators is no longer present but can be emulated by means of targeted control strategies. Consequently, given suitable parameterization of the virtual synchronous machine, this has a positive effect on various phenomena related to the frequency, such as, e.g., a frequency dip after the failure of a relatively large power plant unit or an HVDC transmission line. Frequency oscillations may also be relevant.

(84) After loss of the electrical power supply network, in particular after loss of an interconnected grid, it is possible to achieve the formation of an island network system by means of black start capability by virtue of the fact that the network voltage and frequency can be predefined by the converter.

(85) A use of the core equations of a synchronous machine using a current-controlled converter for simple implementation of current limiting is made possible.

(86) A use of the method for network formation of an island network in the case of black start and switchover from load-free to island network operation and back is a preference.

(87) It is also proposed that, in order to generate the sinusoidal setpoint value currents, use is made of instantaneous values of the voltages, as represented in equation (10). The measured terminal voltages of the converter are not sinusoidal and may contain many harmonics; therefore, problems might arise in the calculation of the setpoint value currents and lead to instability in an unfavorable case. The process of filtering the terminal voltages into qd components, in comparisons with the abc coordinate system, leads, in the settled state, to a better voltage quality with a small filter time constant, without having to compensate for the amplitude and phase shift.

(88) The virtual impedance according to equation (9) is adapted adaptively depending on the operating mode, such that the control stability is ensured, at least in accordance with one embodiment.

(89) In particular, it is also possible to achieve stable operation of a current-control converter with the following properties: Black start by impressing a voltage at the converter terminals without a connected load. Switchover to island operation with load and in the process correction of voltage and frequency to selectable setpoint values. Synchronization with an existing network and rapid control of active and reactive power as a function of network frequency and voltage. Network operation even with very small short-circuit ratios.

(90) Reference is made, in particular, to the following literature: [Lit1] P. C. Krause, O. Wasynczuk and S. D. Sudhoff, “Analysis of Electric Machinery and Drive Systems”, 2nd Edition, New York, 2002, John Wiley & Sons.