Control method for redundant use in the event of a fault in a polyphase converter with distributed energy stores

The invention relates to a method for controlling a converter having at least two phase modules, which each have an upper and a lower valve branch which each have at least three series-connected two-pole subsystems, in the event of failure of at least one subsystem in one valve branch of a phase module.

One such converter circuit of this generic type is known from DE 101 03 031 A1, and an equivalent circuit of a converter circuit such as this is illustrated in more detail in FIG. 1 . According to this equivalent circuit, this known converter circuit has three phase modules, which are each annotated 100 . On the DC voltage side, these phase modules 100 are each electrically conductively connected to a positive and a negative DC voltage bulbar P 0 and N 0 . There is a DC voltage, which is not annotated in any more detail, between these two DC voltage busbars P 0 and N 0 . Each phase module 100 , which forms one bridge arm of the polyphase converter, has an upper and a lower bridge arm which (since these bridge arm elements each represent one converter valve of the polyphase converter with distributed energy stores) are referred to in the following text as the respective valve branches T 1 , T 3 , T 5 and T 2 , T 4 , T 6 . Each of these valve branches T 1 to T 6 has a number of two-pole subsystems 10 , which are electrically connected in series. Four of these subsystems 10 are illustrated in this equivalent circuit. Each junction point between two valve branches T 1 and T 2 ; T 3 and T 4 and T 5 and T 6 , respectively, of a phase module 100 forms a respective connection L 1 , L 2 or L 3 on the AC voltage side of this phase module 100 . Since, in this illustration, the converter circuit has three phase modules 100 , a three-phase load, for example a three-phase motor, can be connected to their connections L 1 , L 2 and L 3 on the AC voltage side, also referred to as load connections.

FIG. 2 shows an equivalent circuit of a known embodiment of a two-pole subsystem 10 in more detail. The circuit arrangement shown in FIG. 3 represents a functionally completely equivalent variant, which is likewise known from DE 101 03 031 A1. This known two-pole subsystem 10 has two semiconductor switches 1 and 3 which can be turned off, two diodes 2 and 4 and a unipolar energy storage capacitor 9 . The two semiconductor switches 1 and 3 which can be turned off are electrically connected in series, with this series circuit being connected electrically in parallel with the energy storage capacitor 9 . One of the two diodes 2 and 4 is electrically connected in parallel with each semiconductor switch 1 and 3 which can be turned off, such that these diodes 2 and 4 are connected back-to-back in parallel with the corresponding semiconductor switch 1 or 3 which can be turned off. The unipolar energy storage capacitor 9 in the subsystem 10 is either in the form of a capacitor or a capacitor bank comprising a plurality of such capacitors, with a resultant capacitance C o . The connecting point of the emitter of the semiconductor switch 1 which can be turned off and the anode of the diode 2 forms a connecting terminal X 1 of the subsystem 10 . The connecting point of the two semiconductor switches 1 and 3 which can be turned off and of the two diodes 2 and 4 form a second connecting terminal X 2 of the subsystem 10 .

In the embodiment of the subsystem 10 shown in FIG. 3 , this connecting point forms the first connecting terminal X 1 . The connecting point of the collector of the semiconductor switch 1 which can be turned off and the cathode of the diode 2 forms the second connecting terminal X 2 of the subsystem 10 .

In both illustrations of the two embodiments of the subsystem 10 , insulated gate bipolar transistors (IGBTs) are used, as illustrated in FIGS. 2 and 3 , as semiconductor switches 1 and 3 which can be turned off. MOS field-effect transistors, also referred to as MOSFETs, can likewise be used. In addition, gate turn off thyristors, also referred to as GTO thyristors, or integrated gate commutated thyristors (IGCT) may also be used.

According to IDE 101 03 031 A1, the subsystems 10 of each phase module 100 of the converter circuit shown in FIG. 1 may be switched to a switching state I, II and III. In the switching state I, the semiconductor switch 1 which can be turned off is switched on, and the semiconductor switch 3 which can be turned off is switched off. A terminal voltage U X21 , which is produced between the connecting terminals X 1 and X 2 , of the subsystem 10 is therefore equal to zero. In the switching state II, the semiconductor switch 1 which can be turned off is switched off, and the semiconductor switch 3 which can be turned off is switched on. In this switching state II, the terminal voltage U X21 that is present is equal to the capacitor voltage U C across the energy storage capacitor 9 . In the switching state III, both semiconductor switches 1 and 3 which can be turned off are switched off, and the capacitor voltage U C across the energy storage capacitor 9 is constant.

In order to allow this converter with distributed energy stores 9 to be operated redundantly as shown in FIG. 1 , it is necessary to ensure that a faulty subsystem 10 is permanently shorted at its terminals X 1 and X 2 . This means that the terminal voltage U X21 of the faulty subsystem 10 is zero irrespective of the current direction through the terminals X 1 and X 2 .

A failure of a semiconductor switch 1 or 3 which can be turned off and is provided in the subsystem 10 , or of an associated drive circuit, results in this subsystem 10 not operating correctly. Further possible reasons for malfunctions include faults in the associated drive circuit of the semiconductor switches, their power supply, communication and measured-value recording. This means that the subsystem 10 can no longer be switched as desired to one of the possible switching states I, II or III. The shorting of the connections X 1 and X 2 of the subsystem 10 means that no more power is supplied to this subsystem 10 . In consequence, consequential damage such as overheating and fire in the event of further operation of the converter cannot reliably be precluded. A conductive connection in the form of a short between the connecting terminals X 1 and X 2 of a faulty subsystem 10 such as this must carry at least the operating current of a valve branch T 1 , . . . , T 6 of the phase module 100 in which the faulty subsystem 10 is connected, safely and without overheating. The prior national patent application with the internal file reference 2005P12103 DE indicates how a faulty subsystem 10 can be safely shorted in order that this known converter with distributed energy stores can still be operated in a redundant form.

The following explanation is based on the assumption that the energy storage capacitors 9 of all the subsystems 10 are each at the same voltage U C . Methods for initial production of this state and for its maintenance during operation are likewise known from DE 101 03 031 A1. FIG. 4 shows a graph, plotted against time t, of a profile of the potential difference U LN between the terminal L of a phase module 100 and a selected reference ground potential N. FIG. 5 shows a graph, plotted against time t, of a profile of the potential difference U PL between the terminal P and a load connection L. These two potential profiles U LN and U PL are normalized with respect to the voltage U C of the energy storage capacitors 9 in the subsystems 10 . One subsystem 10 of the four subsystems 10 in the respective valve branch T 2 or T 1 is in each case switched on and off, or turned off and on, at the respective times t 2 , t 3 , t 4 , t 5 or t 7 , t 8 , t 9 and t 10 in accordance with these normalized potential profiles U LN /U C (t) and U PL /U C (t). Switching on in this case corresponds to a change from the switching state I to the switching state II. Switching off corresponds to a change from the switching state II to the switching state I. These two graphs each show one period Tp of the normalized potential profile U LN /U C (t) and U PL /U C (t). Corresponding components of harmonic or DC voltage components in each of the output voltages U LN of the phase modules 100 of the polyphase converter with distributed energy stores 9 are canceled out in the event of a balanced three-phase voltage system in the difference voltages between any two phase-shifted output voltages U L1N , U L2N or U L3N . These two normalized potential profiles U LN /U C (t) and U PL /U C (t) likewise show that the sum of the normalized potentials at any time is four. This means that the DC voltage between the DC voltage busbars P 0 and N 0 always corresponds to a constant number of subsystems 10 in the switching state II multiplied by the capacitor voltage U C across the capacitor 9 . In the situation illustrated by way of example, this number corresponds to the number of subsystems 10 of the converter in the valve branches T 1 , . . . , T 6 , as shown in FIG. 1 . By way of example, the four subsystems 10 of the valve branch T 1 are all in the switching state II (U X21 =U C ) and the four subsystems 10 in the valve branch T 2 of the corresponding phase module are all in the switching state I (U X21 =0) at the times t 0 and t 1 . One submodule 10 of the valve branch T 1 in each case changes its switching state from II to I at the times t 2 , t 3 , t 4 and t 5 while, in contrast, one subsystem 10 of the valve branch T 2 in each case changes its switching state from I to II. If one subsystem 10 of a respective valve branch T 1 , T 2 ; T 3 , T 4 or T 5 , T 6 of a phase module 100 in the polyphase converter as shown in FIG. 1 now fails because of some fault, then at least one of the three output voltages U L1N , U L2N or U L3N of this polyphase converter with distributed energy stores 9 has harmonic and/or DC voltage components which, in some circumstances, can lead to this converter being turned off, as shown in FIG. 1 .

The invention is now based on the object of specifying a control method by means of which the balance conditions can be maintained even in the event of a fault in at least one subsystem of a phase module of a converter with distributed energy stores.

According to the invention, this object is achieved by a method for controlling a converter having at least two phase modules, which have an upper and a lower valve branch which each have at least three series-connected two-pole subsystems, in the event of failure of at least one subsystem in one valve branch, with the valve branch with the failed subsystem being determined, and with one subsystem of a valve branch which corresponds to the faulty valve branch in each sound phase module in each case being driven such that their terminal voltages are each zero.

Since, according to the invention in the sound phase modules of the polyphase converter with distributed energy stores, one subsystem of a valve branch which corresponds to the valve branch with the faulty subsystem is in each case driven such that their terminal voltages are zero while the fault is present, all the output voltages of the converter with distributed energy stores are the same again, so that their difference voltages no longer have any harmonic and/or DC voltage components which are divisible by three.

In one advantageous method, a subsystem of a valve branch which corresponds to the faulty valve branch in the faulty phase module is driven such that its terminal voltage is equal to a capacitor voltage in the subsystem. In consequence, the output voltage of this faulty phase module is once again balanced with respect to the mid-value of a fundamental of this staircase output voltage. A corresponding procedure is adopted in the sound phase modules, so that the three output voltages of a three-phase converter with distributed energy stores once again form a balanced three-phase voltage system. In addition, the value of the intermediate-circuit voltage corresponds to the value of the intermediate-circuit voltage when no fault is present, so that the voltage load on the semiconductor switches which can be turned off in the systems likewise corresponds to the voltage load in the sound state.

In a further advantageous method, the switching times of the control signals for the subsystems of the valve branches of the phase modules of the polyphase converter with distributed energy stores are offset in time. This time offset between the control signals for the subsystems of each phase module means that the undisturbed amplitude profile of a fundamental of a phase output voltage is maintained approximately in the event of a fault.

This control method according to the invention allows the output voltages of the phase modules of a polyphase converter with distributed energy stores to be maintained even in the event of a fault. This converter can therefore be operated redundantly. When a polyphase load is connected to this polyphase converter with distributed energy stores, it is not possible to tell whether and how many subsystems in one valve branch of a phase module of this polyphase converter are faulty. All that happens is that this polyphase output voltage system has a reduced amplitude with an unchanged operating point, and apart from this there is no difference from the operating point when no fault is present.

In order to explain the invention further, reference is made to the drawing, which schematically illustrates a plurality of embodiments of a method according to the invention for controlling a polyphase converter with distributed energy stores, and in which:

FIG. 1 shows an equivalent circuit of a known converter circuit with distributed energy stores,

FIG. 2 shows an equivalent circuit of a first embodiment of a known subsystem,

FIG. 3 shows an equivalent circuit of a second embodiment of a known subsystem,

FIGS. 4 and 5 each show a graph plotted against time t of normalized potential profiles on the two valve branches of a phase module of the converter shown in FIG. 1 ,

FIGS. 6 and 7 show the potential profiles as shown in FIGS. 4 and 5 , in each case in a graph plotted against time t, in the event of failure of one subsystem in a lower valve branch of the converter as shown in FIG. 1 ,

FIGS. 8 and 9 show the potential profiles as shown in FIGS. 4 and 5 , in each case in a graph plotted against time t, in the event of failure of a subsystem in an upper valve branch of the converter shown in FIG. 1 ,

FIGS. 10 and 11 show the potential profiles as shown in FIGS. 6 and 7 , in each case in a graph plotted against time t, in which the switching times of the control signals are additionally offset in time, and

FIGS. 12 and 13 show the potential profiles as shown in FIGS. 8 and 9 , in each case in a graph plotted against time t, in which the switching times of the control signals are additionally offset in time.

It is now assumed that one subsystem 10 in the valve branch T 2 of the phase module 100 of the converter with distributed energy stores 9 as shown in FIG. 1 is safely shorted because of some fault. This faulty subsystem 10 is identified by means of shading in the equivalent circuit shown in FIG. 1 . Only three subsystems 10 can therefore be used to generate the time profile of the normalized output potential U L1N /U C (t) while, in contrast, four subsystems 10 are, however, still used to generate the time profile of the normalized potential U PL1 . This means that the potential profile U L1N /U C (t) has one potential step less, corresponding to FIG. 4 . This means that this potential profile U L1N /U C (t) corresponds to the potential profile U L1N /U C (t) shown in the graph in FIG. 6 . The faulty system 10 in the valve branch T 2 has changed nothing on the potential profile U PL1 /U C (t), so that this potential profile U PL1 /U C (t) still has a corresponding potential profile to U PL1 /U C (t) as shown in the graph in FIG. 5 . In consequence, the DC voltage between the DC voltage busbars P 0 and N 0 for two time units t 6 -t 5 and t 7 -t 6 in this faulty phase module 100 is equal to 3·U C in comparison to 4·U C in the sound phase modules 100 . Since the DC voltage between the DC voltage busbars P 0 and N 0 is used, a higher valve current flows in the faulty phase module 100 and additionally loads the components of each subsystem 10 in this faulty phase module 100 . If this valve current exceeds the maximum permissible current values of a component of a subsystem 10 in this phase module 100 , the polyphase converter with distributed energy stores 9 is switched off because of overcurrent.

In order to prevent this, the polyphase converter with distributed energy stores 9 is controlled according to the invention. To do this, it is first of all necessary to determine which respective valve branch T 1 , T 2 ; T 3 , T 4 or T 5 , T 6 of a phase module 100 has a faulty subsystem 10 . A subsystem 10 such as this may occur in the respective valve branch T 2 , T 4 or T 6 , or in the respective valve branch T 1 , T 3 or T 5 . If a faulty subsystem 10 occurs in the respective valve branch T 2 , T 4 or T 6 , the associated potential profiles U L1N /U C (t), U L2N /U C (t) or U L3N /U C (t), respectively, correspond to the profile in the graph shown in FIG. 6 . If a faulty subsystem 10 occurs in the respective valve branch T 1 , T 3 or T 5 , the potential profiles U PL1 /U C (t), U PL2 /U C (t) or U PL3 /U C (t) correspond to the profile in the graph shown in a corresponding manner in FIG. 9 .

Once it has been found which of the valve branches T 1 , . . . , T 6 has a faulty subsystem 10 , a respectively corresponding number of the faulty subsystems 10 which have occurred in the respective valve branch T 2 or T 1 are likewise driven, for example, in the respective valve branches T 4 and T 6 , as well as T 3 and T 5 which correspond to this faulty valve branch T 2 or T 1 , such that their terminal voltage U X21 =0. The DC voltage which is applied between the DC voltage busbars P 0 and N 0 is therefore split in all the phase modules 100 of the polyphase converter with distributed energy stores 9 between the same number of subsystems 10 in each case. This control according to the invention of the polyphase converter with distributed energy stores 9 means that the balance conditions are maintained even in the event of a fault, so that the difference voltages between the load connections L 1 , L 2 and L 3 do not have any harmonic or DC voltage components which can be divided by three. This means that it is not possible to tell for a connected load whether the polyphase converter with distributed energy stores 9 is operating with a fault or without any faults. The polyphase converter with distributed energy stores 9 is therefore operated redundantly.

If the aim is to maintain the correct DC voltage between the DC voltage busbars P 0 and N 0 and the voltage load on the semiconductor switches 1 and 3 which can be turned off in the subsystems 10 of the phase modules 100 in a polyphase converter with distributed energy stores 9 even in the event of a fault, then, in the faulty phase module 100 , a respective valve branch T 1 , T 3 or T 5 , or T 2 , T 4 or T 6 , which corresponds to the faulty respective valve branch T 2 , T 4 or T 6 , or T 1 , T 3 or T 5 , with a corresponding number of subsystems to the number of faulty subsystems 10 being driven such that their terminal voltage is given by U X21 =U C . A corresponding procedure is used in the sound phase modules 100 of the polyphase converter with distributed energy stores 9 . This additional method step from the method according to the invention results in the number of subsystems 10 which are being used when the phase modules 100 in this polyphase converter with distributed energy stores 9 are faulty and sound being the same again. In consequence, the respective potential profiles U L1N /U C (t), U Pl1 /U C (t) and U L2N /U C (t), respectively and U PL2 /U C (t), U L3N /U C (t) and U PL3 /U C (t), respectively, once again correspond to the profiles in the graphs shown in FIGS. 6 and 7 . If one subsystem 10 in the respective valve branch T 1 , T 3 or 5 fails and if the polyphase converter with distributed energy stores 9 is operated using the advantageous control method according to the invention, then the potential profiles U L1N /U C (t), U PL1 /U C (t) and U L2N /U C (t) and U PL2 /U C (t), U L3N /U C (t) and U PL3 /U C (t) correspond to the profiles in the graphs in FIGS. 8 and 9 . The potential profiles in the graphs in FIGS. 6 , 8 and 7 , 9 therefore correspond to the profiles in the graphs in FIGS. 4 and 5 with the difference that, for example, one subsystem 10 is used to a lesser extent for generation of the potential profiles during faulty operation. This means that the output voltages U L1N , U L2N and U L3N of the polyphase converter with distributed energy stores 9 produces somewhat less amplitude when faulty. The difference from operation in the sound state corresponds to the capacitor voltage U C of a subsystem 10 . The greater the number of subsystems 10 being used in normal operation for generation of the output voltages U L1N , U L2N and U L3N of the polyphase converter with distributed energy stores 9 , the lower is the amplitude reduction when operating with a fault.

If, however, the intention is to maintain the undisturbed amplitude of the fundamental of each output voltage U L1N , U L2N and U L3N that is produced in the polyphase converter with distributed energy stores 9 approximately, then the advantageous control method according to the invention is modified such that the switching times of the control signals for the semiconductor switches 1 and 3 which can be turned off in the subsystems 10 of the valve branches T 1 , . . . , T 6 are additionally offset in time. The potential profiles produced in this way are illustrated in the graphs in FIGS. 10 , 11 , 12 and 13 .

As explained by way of example, this control method according to the invention is not restricted to failure of just one subsystem 10 in one valve branch T 1 , . . . , T 6 in the polyphase converter with distributed energy stores 9 . As described, this control method results in output voltages U L1N , U L2N and U L3N being generated even when a plurality of subsystems 10 in one valve branch T 1 , T 2 , T 3 , T 4 , T 5 or T 6 have failed. Care should be taken to ensure that the remaining number of subsystems 10 in a valve branch T 1 , . . . , T 6 does not become less than three, in order to ensure that the amplitudes of the harmonics in the respective output voltages U L1N , U L2N and U L3N remain low.

This control method according to the invention for polyphase converters with distributed energy stores 9 is particularly advantageous for power supply systems. Converters such as these include converters for power supply system couplings, for power factor correction and for voltage stabilization in power supply systems. Because of the high voltage in conventional power supply systems, a polyphase converter with distributed energy stores 9 has a large number of subsystems 10 in each valve branch T 1 , . . . , T 6 , for example from 10 to more than one hundred.