Method for compensating neutral-point voltage of three-stage inverter
By predicting and compensating the current load of the midpoint circuit in a three-stage inverter, and reducing the midpoint voltage change using redundant voltage vectors, the voltage problem caused by unequal load of the midpoint connector is solved, extending the service life of the capacitor and improving the System stability.
Patent Information
- Application Number
- CN202380037702.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-08-09
- Filing Date
- 2023-07-12
- Publication Date
- 2024-12-17
AI Technical Summary
Unequal loads of midpoint connectors in three-stage inverters result in undesirable midpoint voltage changes, reducing the life of the capacitor and potentially damaging the entire inverter.
By determining the specific switching mode of the operating point, a switching state with at least two redundant voltage vectors is used to generate a switching state that has an opposite effect on the midpoint voltage, predicting and compensating the current load of the midpoint circuit, providing a compensation current through the weighted redundant voltage vector Reduce the midpoint voltage change.
It effectively reduces the midpoint voltage change, extends the service life of the capacitor, avoids inverter damage, and improves the stability of the system.
Smart Images
Figure CN119156764A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a method for operating a three-level inverter, the three-level inverter having an intermediate circuit with three connections in the form of a high-side connection, a low-side connection and a midpoint connection. In addition, the three-level inverter has at least three switching sections connected to the intermediate circuit, each of which includes a phase tap for electrical connection to a respective phase of an electric drive machine. Each switching section has three switching devices, which can be switched into at least 27 switching states for modulating an operating point-specific output voltage and in this case connect the phase tap of the corresponding switching section to a respective one of the connections of the intermediate circuit. The invention also relates to a control device, a power module and a motor vehicle. Background Art
[0002] Currently, three-level inverters or three-level inverters are of interest for electrically driven motor vehicles, i.e. electric vehicles or hybrid vehicles. Such three-level inverters are connected between an electrical energy storage device, such as a traction battery, and an electric drive of the motor vehicle and are used to convert the DC voltage provided by the electrical energy storage device into a multiphase AC voltage for the phases of the electric drive. The three-level inverter can also be used as a power factor correction filter for a three-phase charger.
[0003] The inverter usually has three half-bridges connected in parallel and an intermediate circuit connected to each half-bridge. Each half-bridge has two switching devices, wherein the first switching device has at least one high-side switch and is connected to the high-side connection of the intermediate circuit, and the second switching device has at least one low-side switch and is connected to the low-side connection of the intermediate circuit. In addition to the half-bridge, the three-level inverter has three third switching devices, each including at least one midpoint switch, wherein each third switching device is electrically connected to the tap between the first and second switching devices of the half-bridge and to the midpoint connection of the intermediate circuit. Such a three-level inverter can provide three voltage levels for each phase of the motor for each switching section including the half-bridge and the third switching device, and thus has another 19 switching states in addition to the eight switching states of the conventional two-level inverter for modulating the output voltage specific to the working point. With these 27 switching states, the desired sinusoidal shape of the output voltage with a particularly low harmonic content can be generated, thereby reducing the losses in the motor. In addition, the three-level inverter has particularly low switching losses.
[0004] However, a problem with such three-level inverters is the unequal loading of the center point connections, or center points for short, which is caused in particular by the switching states in which at least one center point switch is involved. This results in undesirable center point voltage variations, which lead to a reduction in the service life of the capacitors of the intermediate circuit. If the center point voltage variation exceeds a predetermined threshold value, damage to the entire three-level inverter may occur. Summary of the invention
[0005] The object of the present invention is to provide a solution for reducing the center point voltage variation of a three-level inverter for a motor vehicle.
[0006] This object is achieved according to the invention by a method, a control device, a power electronics module and a motor vehicle having the features according to the respective independent claims. Advantageous embodiments of the invention are the subject of the dependent claims, the description and the drawings.
[0007] The method according to the invention is used to operate a three-level inverter. The three-level inverter has an intermediate circuit with three connections in the form of a high-side connection, a low-side connection and a midpoint connection. In addition, the three-level inverter has at least three switching sections connected to the intermediate circuit, each of which includes a phase tap for electrical connection to a respective phase of the electric drive machine. Each switching section has three switching devices, which can be switched into at least 27 switching states for modulating an operating point-specific output voltage and in this case connect the phase tap of the corresponding switching section to a respective one of the connections of the intermediate circuit while generating a voltage vector.
[0008] In the method, in order to compensate for an undesired, switching state-specific change in the center point voltage of the intermediate circuit when modulating an operating point-specific output voltage, an operating point-specific switching pattern is determined from at least four voltage vectors, at least two of which are generated redundantly but with opposite effects on the center point voltage by different switching states of the three-level inverter. Based on the determined switching pattern, a current load of the intermediate circuit that causes an undesired change in the center point voltage is predicted. Furthermore, based on the predicted current load, a compensation current is determined that compensates for the resulting current load within a modulation period and in this regard a time weighting of the redundant voltage vectors is determined, by which at least a portion of the compensation current that is to be compensated can be provided when modulating the output voltage.
[0009] The invention also includes a control device and a power electronics module comprising a three-level inverter and a control device according to the invention, the control device being designed to carry out the method according to the invention. The motor vehicle according to the invention has an electrical energy storage device, an electric machine and a power electronics module according to the invention, wherein the three-level inverter is connected to the electrical energy storage device at the input and to the phases of the stator of the electric machine at the output. The three-level inverter is designed to convert a DC voltage provided by the electrical energy storage device into a multiphase AC voltage and to provide it as an output voltage for the phases of the electric machine, the DC voltage being present as an input voltage at the intermediate circuit of the three-level inverter.
[0010] The three-level inverter has an intermediate circuit at the input end, which in particular has at least one series circuit consisting of at least two capacitors. The switching section or the switching bridge respectively has a first switching device including at least one high-side switch, a second switching device including at least one low-side switch and a third switching device including at least one midpoint switch. The first and second switching devices of the switching section are connected to form a half bridge. Each switching section has a phase tap, which is connected to the corresponding phase of the motor. The high-side connection and the low-side connection of the series circuit that form the intermediate circuit are connected to the poles of the energy storage device and to the half bridge of the switching section or the switching bridge. A midpoint connection, referred to as a midpoint, is provided between the capacitors of the at least one series circuit, and the midpoint connection is connected to the third switching device of the switching section. The half bridges here respectively form the vertical bridge arms of the corresponding switching section and the third switching device forms the horizontal bridge arms of the corresponding switching section. The high-side switch and the low-side switch can be configured as IGBTs or power MOSFETs, for example. The third switching device can have two center point switches connected in series in opposite directions or one bidirectional center point switch in order to provide a bidirectional disconnection capability. The center point switches can also be designed as IGBTs or power MOSFETs.
[0011] Through the three switching devices of each switching bridge, the three-level inverter has 27 switching states in the case of three switching bridges, wherein each switching state generates a voltage vector, which can be represented as a space vector in a vector diagram. Here, when the first switching device of the switching section is closed, the phase tap of the switching section is electrically connected to the high-side terminal and the high-side potential is thus applied to the corresponding phase. When the second switching device of the switching section is closed, the phase tap of the switching section is electrically connected to the low-side terminal and the low-side potential is thus applied to the corresponding phase. When the third switching device of the switching section is closed, the phase tap of the switching section is electrically connected to the midpoint terminal and the midpoint potential is thus applied to the corresponding phase.
[0012] For example, three switching states, in which either all high-side switches are closed and therefore all phase taps and phases are connected to the high-side connection of the intermediate circuit, or all low-side switches are closed and therefore all phase taps and phases are connected to the low-side connection of the intermediate circuit, or all midpoint switches are closed and therefore all phase taps and phases are connected to the midpoint connection of the intermediate circuit, produce a zero voltage vector, which is represented as a zero voltage space vector in a space vector diagram. The zero voltage space vector can therefore be generated redundantly.
[0013] Six further switching states without the involvement of the third switching device and thus the midpoint switch produce six voltage vectors, which are represented as long outer voltage space vectors in the space vector diagram and form the corner points of the outer hexagon. These six long outer voltage space vectors correspond to the basic voltage space vector of the two-stage inverter. The six further switching states have the involvement of the midpoint switch and produce six further voltage vectors, which are presented as central outer voltage space vectors in the space vector diagram and are located between two long outer voltage space vectors at the side edges of the outer hexagon.
[0014] Twelve further switching states, which also involve the midpoint switch, produce twelve short internal voltage space vectors, wherein two switching states in each case provide two identical voltage space vectors, i.e. voltage space vectors with the same length and the same phase angle. Each of the short internal voltage space vectors can therefore be generated redundantly. In this case, the short internal voltage space vectors define an inner hexagon, wherein each of the two redundant voltage space vectors forms a corner point of the inner hexagon. The mutually redundant voltage space vectors have opposite, complementary effects on the current in the midpoint.
[0015] The outer hexagon is divided into six triangular sectors, wherein the corner points of each sector are formed by the zero voltage space vector and two adjacent long outer voltage space vectors. Each sector is again divided into four triangular segments, wherein two redundant voltage space vectors are located at at least one corner point of each segment. In order to modulate any operating point-specific output voltage with the aid of the voltage vector, it is determined in which sector and in which segment of the sector the operating point-specific output voltage space vector representing the output voltage is located. Subsequently, this operating point-specific output voltage space vector can be generated by pulse width modulation of the voltage space vector that spans this segment. For this purpose, a switching pattern is determined in which the sequence and duration of the segment-specific voltage space vectors used for modulation are specified.
[0016] Since the voltage vector in the switching pattern is also used in the modulation of the output voltage, which is generated by the switching state in which the midpoint switch participates and in which the phase tap of at least one switching section is therefore connected to the midpoint connection, a current load of the midpoint can be generated, which is not equal to zero in terms of the sum over a modulation cycle, because the capacitors of the intermediate circuit are unequally charged and discharged when the midpoint switch is switched. The uncompensated current load of the midpoint over the modulation cycle leads to a midpoint voltage change or a deviation of the midpoint voltage. Since this midpoint voltage change may reduce the service life of the capacitor or may cause damage to the operating limit of the switching device, the midpoint voltage should be balanced and thus compensated for the midpoint voltage change over the modulation cycle. For this purpose, it is first predicted or forewarned what current load in the midpoint of the intermediate circuit is caused by the switching pattern determined for modulating the output voltage specific to the working point.
[0017] In particular, the current load is predicted with the aid of the duration of a non-redundant voltage vector within the switching pattern, which is generated by a switching state with the participation of the midpoint switch. The non-redundant voltage vector generated with the participation of the midpoint switch corresponds to the middle external voltage space vector. In particular, the middle external voltage space vector contributes to the change in the midpoint voltage within a modulation period. For predicting the current load, for example, a current-time area can be determined by determining the current load of the midpoint caused by the middle external voltage space vector and multiplying it by the duration for which the middle external voltage space vector exists.
[0018] Furthermore, the current load which leads to a deviation in the midpoint voltage can be predicted using the currents which are supplied to the phases according to the switching pattern. This embodiment is based on the recognition that the potentials applied to the phases via the phase taps connected to the corresponding connections of the intermediate circuit lead to sinusoidal currents in the phases, wherein this sinusoidal shape of the phase currents can contribute to the midpoint voltage variation. This means that if, for example, the redundant voltage space vectors are each halved within the switching pattern, ), and in this case a redundant first voltage space vector is set at the beginning of a modulation period, and a redundant second voltage space vector having an influence on the midpoint voltage opposite to the first voltage space vector is set at the end of the modulation period, an influence on the midpoint occurs. This influence has the effect that different phase current values are applied to the midpoint at the time of the redundant first and second voltage space vectors.
[0019] In addition, the current load is preferably additionally determined by means of a measured value of the voltage deviation already present at the beginning of the corresponding modulation cycle at the midpoint connection. This measured value can be converted into a current-time-area, which compensates for the difference again with the next modulation cycle. The existing deviation measured at the beginning of each modulation cycle is especially caused by tolerances and asymmetries determined by hardware of the specific three-level inverter and is determined by components. Therefore, the current load can have multiple shares, which, as described, can be formed by component-dependent deviations of the intermediate external voltage space vector, the current curve, and the midpoint current or the midpoint impedance. For example, a corresponding current-time-area can be determined for each current load share, from which a current-time-total area describing the predicted total current load is obtained.
[0020] Based on this predicted current load, a compensation current is now determined, which is applied to the midpoint of the intermediate circuit when modulating the compensation voltage in order to compensate for the predicted current load over the modulation period or switching period. In other words, the current load is selected in such a way that it is compensated over the modulation period and thus balances the midpoint voltage. The compensation current is determined in such a way that it compensates for the current-time-total area. When modulating the output voltage specific to the operating point, the compensation current is provided by means of redundant voltage vectors, i.e. short internal voltage space vectors, because these redundant identical voltage space vectors generated by means of different switching states have opposite effects on the midpoint voltage. For this purpose, the redundant voltage vectors are weighted in time within the modulation period so that at least a part of the compensation current can be provided. Subsequently, the switching pattern that at least reduces the variation of the midpoint voltage is provided for modulating the output voltage, the switching pattern having redundant voltage vectors weighted in time.
[0021] In the case where the current load cannot be completely compensated by the time weighting of the redundant voltage vectors, it can be provided that the non-redundant voltage vector that causes the change in the midpoint voltage and is generated by the switching state with the participation of the midpoint switch is at least partially replaced by two non-redundant voltage vectors generated by the switching state without the participation of the midpoint switch. In this case, the short-time average value of the output voltage does not change over a modulation period. Therefore, if the current-time-total area cannot be compensated by means of the short inner voltage space vector, the middle outer voltage space vector is at least partially replaced by the two adjacent long outer voltage space vectors, without changing the short-time average value of the output voltage over the modulation period. By providing the two outer long voltage space vectors with the same time share in the switching pattern, the outer long voltage space vector can at least partially imitate the middle outer voltage space vector. In this case, in order to provide the compensation current, the middle outer voltage space vector is replaced in the switching pattern, in particular, only by the long outer voltage space vector, until the remaining share of the compensation current can be provided by the time weighting of the redundant voltage space vectors.
[0022] It can be provided that in a first modulation cycle half, the two non-redundant voltage vectors generated by the switching states without the participation of the midpoint switch and one of the redundant voltage vectors are provided, and in a subsequent second modulation cycle half, the two non-redundant voltage vectors generated without the participation of the midpoint switch and the other of the redundant voltage vectors are provided. Thus, for example, a redundant first voltage vector is provided in the first modulation cycle half and a redundant second voltage vector is provided in the second modulation cycle half, which redundant second voltage vector has an opposite effect on the midpoint voltage.
[0023] The embodiments and advantages provided with regard to the method according to the invention apply correspondingly to the control device according to the invention, the power electronics module according to the invention and the motor vehicle according to the invention.
[0024] Further features of the invention are apparent from the claims, the drawings and the description of the drawings. The features and feature combinations mentioned above in the description and the features and feature combinations mentioned below in the description of the drawings and / or shown individually in the drawings can be used not only in the respectively indicated combination but also in other combinations or individually. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] The present invention will now be further explained with the aid of preferred embodiments and with reference to the accompanying drawings.
[0026] Figure 1 A circuit diagram showing one design of a three-level inverter;
[0027] Figure 2 A diagram showing a vector diagram of a three-level inverter;
[0028] Figure 3 A diagram showing a switching pattern for modulating the output voltage of a three-level inverter;
[0029] Figure 4a , b shows a diagram of a compensable current load as well as phase currents and switching patterns with and without balancing of the midpoint voltage;
[0030] Figure 5 a diagram showing currently uncompensable current loads and phase currents and switching patterns;
[0031] Figure 6 showing the duration of the voltage vector that resulted in the currently uncompensable current load; and
[0032] Figure 7 Shown for compensation in accordance with Figure 5 The diagram of the optimized switching pattern of the current load and the compensation according to Figure 5 The duration of the voltage vector of the current load. DETAILED DESCRIPTION
[0033] In the figures, identical elements and elements having the same function are provided with the same reference symbols.
[0034] Figure 1 A three-level inverter 1 for an electrified motor vehicle is shown. The three-level inverter 1 has an intermediate circuit 2 at the input, which here has three intermediate circuit branches 2a, 2b, 2c, each of which includes a series circuit consisting of two capacitors C1, C2. The high-side terminal DC+ and the low-side terminal DC- of the intermediate circuit 2 are connected to the poles of an electric energy storage device of the motor vehicle, which supplies a DC voltage V to the intermediate circuit 2. DC In this case, half of the DC voltage V is present on each capacitor C1, C2. DC / 2. The three-level inverter 1 also has three switching sections 3a, 3b, 3c, which are electrically connected to one of the intermediate circuit branches 2a, 2b, 2c. However, it can also be provided that the intermediate circuit 2 has only one intermediate circuit branch comprising two capacitors C1, C2 connected in series, to which the switching sections 3a, 3b, 3c are connected. Each switching section 3a, 3b, 3c has a phase tap u, v, w, which is connected to a phase Phu, Phv, Phw of the electric drive machine of the motor vehicle.
[0035] Each switching section has three switching devices S1, S2, S3, each of which can have at least one switch configured as a semiconductor switch, such as an IGBT. The switching devices S1 and S2 are connected to form the half bridge HB of the corresponding switching section 3a, 3b, 3c. The first switching device S1 is a high-side switching device and is connected to the high-side connection DC+ of the intermediate circuit 2 and the tap A of the corresponding half bridge HB, and the high-side connection is connected to the positive pole of the electric energy storage device. The second switching device S2 is a low-side switching device and is connected to the tap A of the corresponding half bridge HB and the low-side connection DC- of the intermediate circuit 2, and the low-side connection is connected to the negative pole of the electric energy storage device. The third switching device S3 is a midpoint switch and is connected to the tap A and the midpoint connection N of the corresponding intermediate circuit branch 2a, 2b, 2c between the two capacitors C1 and C2. The midpoint switch S3 here has two midpoint switches S3a and S3b connected in reverse series.
[0036] The three-level inverter 1 can assume 27 switching states, by means of which a desired operating point-specific output voltage can be modulated for application to the phases Phu, Phv, Phw. Each switching state generates a voltage vector which can be Figure 2 , β. In the spatial vector diagram 4 with axes α, β shown in FIG. In the spatial vector diagram 4, different switching states (u, v, w) are plotted at discrete points, where u, v, w=0, 1, 2, where u indicates which of the switching devices S1, S2, S3 in the first switching section 3a is closed, v indicates which of the switching devices S1, S2, S3 in the second switching section 3b is closed, and w indicates which of the switching devices S1, S2, S3 in the third switching section 3c is closed. "0" means that the corresponding low-side switching device S2 is closed and thus the corresponding phase taps u, v, w are connected to the low-side connection DC-. "1" means that the corresponding midpoint switch S3 is closed and thus the corresponding phase taps u, v, w are connected to the midpoint connection N. "2" means that the corresponding high-side switching device S1 is closed and thus the corresponding phase taps u, v, w are connected to the high-side connection DC+. The discrete points here mark the arrows of the voltage space vectors generated by the corresponding switching states, which have a certain length and a certain phase angle. It can already be seen here that there are switching states that produce the same voltage vector. For example, the switching states "221" and "110" produce the same voltage vector. The switching states "000", "111", "222" also produce the same voltage vector, that is, each a zero voltage vector.
[0037] The space vector diagram is divided into six sectors K1, K2, K3, K4, K5, K6, wherein each sector K1, ..., K6 is divided into four segments Q1, Q2, Q3, Q4 (only the first sector K1 is shown here). If a specific output voltage for a certain operating point is to be modulated, for example Figure 2 The output voltage vector V a1 、V a2 , then first determine the corresponding output voltage vector V a1 、V a2 In which sector K1, ..., K6 and in which section Q1, ..., Q4 of the corresponding sector K1, ..., K6. For example, the first output voltage vector V a1 In the first sector K1, in the second section Q2, and thus can be modulated or approached via the voltage vector v2p generated by the switching state "221", the voltage vector v2n generated by the switching state "110", the voltage vector v1p generated by the switching state "211", the voltage vector v1n generated by the switching state "100" and the voltage vector v4 generated by the switching state "210", which form the corner points of the second section Q2 and thus surround the output voltage vector V to be generated. a1 The output voltage vector V a2 In the first sector K1, it is in the third section Q3. The output voltage vector V a2 Therefore, modulation can be performed via the voltage vector v1p generated by the switching state "211", the voltage vector v1n generated by the switching state "100", the voltage vector v3 generated by the switching state "200" and the voltage vector v4 generated by the switching state "210", which form the corner points of the third section Q3. The voltage vectors v1p, v1n and v2p, v2n are redundant voltage vectors.
[0038] The future modulated first output voltage vector V a1 , can be used in Figure 3 The first switching mode SM1 is shown, and the second output voltage vector V a2 , can be used in Figure 3 In particular, the voltage vector v4 influences the midpoint voltage at the midpoint connection N of the intermediate circuit 2, since this voltage vector v4 is generated with the participation of at least one midpoint switch S3 and therefore at least one of the phase taps u, v, w is electrically connected to the midpoint connection N. In addition, as with the voltage vector of the first switching mode SM1, the current i v1n 、i v2n 、i v4 、i v1p 、i v2p exist Figure 4a As shown in FIG. 1 , the switching process of the midpoint switch S3 results in unequal current loads i at the midpoint N. B The current load share i B1 The redundant voltage vectors v1p, v1n or v2p, v2n have an opposite effect on the midpoint voltage and can thereby neutralize the current load i B However, complete neutralization cannot be achieved by this mechanism if the current-time area to be compensated exceeds the time-weighted potential of the redundant vectors v1p, v1n or v2p, v2n, which is derived from the middle long vector v4, from the sinusoidal curve of the current within a modulation period, and from the measured midpoint deviation, which causes the current load i B The other current load share i B2 , the current load i caused by the midpoint deviation B The other current load share i B3 .
[0039] In order to compensate for the uncompensated current load i B And thereby balance the midpoint voltage on the modulation cycle, such as Figure 4b As shown in the figure, determine the compensation current i A , the compensation current corresponds to the current load i B In order to provide the compensation current i A , determine the optimized time weighting t of the redundant voltage vectors v1p, v1n and v2p, v2 v1n *、t v1p * and t v2n *、t v2p * The first optimized switching mode SM1*. In this way, the voltage vector v1n and the redundant voltage vector v1b are no longer as Figure 4a The switching mode SM1 has the same time weighting t v1n and t v1p , but with different time weights t v1n * and t v1p *. The voltage vector v2n and the redundant voltage vector v2b are no longer Figure 4a The switching mode SM1 has the same time weighting t v2n and t v2p , but with different time weights t v2n * and t v2p *. This results in an optimized voltage vector-specific current i of the redundant voltage vectors v1p, v1n, v2p, v2n v1n *、iv2n *、i v1p *、i v2p *, the current load i can be fully compensated by the current B .
[0040] exist Figure 5 The current load i for the second switching mode SM2 is shown in B and voltage vector-specific optimized current i with redundant voltage vectors v1n, v1p v1n *、i v1p * Partially optimized switching mode SM2*, only the compensation current i can be generated through the redundant voltage vector A The compensation current share i A1 This means that if Figure 6 The current load i is also caused by the duration d4 for which the voltage vector v4 is present in the switching mode SM2. B Full compensation is not possible with the redundant voltage vectors v1n, v1p, since only the duration d1 is available for these redundant voltage vectors v1n, v1p, during which the redundant voltage vectors v1p, v1n can be weighted in time. With the redundant voltage vectors v1p, v1n, the current load caused by the voltage vector v4 can only be compensated for the duration d4*, so that the midpoint voltage variation can only be reduced with the redundant voltage vectors v1p, v1n, but not eliminated. For the duration d4*, the current load caused by the voltage vector v4 can only be compensated for the duration d4*, so that the midpoint voltage variation can only be reduced with the redundant voltage vectors v1p, v1n, but not eliminated. red , it is impossible to provide compensation current i through redundant voltage vectors v1p and v1n A .exist Figure 6 Furthermore, the duration d3 is shown, for which the voltage vector v3 is present in the switching pattern SM2 .
[0041] As in Figure 7 As shown in FIG. 1 , a fully optimized switching mode SM2** is thus provided in which the voltage vector v4 is constant for a duration d4. red The voltage vectors v3 and v5 are used instead. For this purpose, for the duration d3*=d3+0.5*d4 red Apply voltage vector v3 and for duration d5=0.5*d4 red The voltage vector v5 is applied. By means of this optimized switching pattern SM2**, the midpoint voltage can be completely balanced. Neutralization over one modulation cycle can be achieved by reducing the value of the midpoint current by partially replacing a voltage vector that generates a midpoint current, such as the voltage vector v4, with such a voltage vector that does not generate a midpoint current, such as the voltage vector v3, v5.
Claims
1. A method for operating a three-level inverter (1), the three-level inverter having an intermediate circuit (2) with three connections in the form of a high-side connection (DC+), a low-side connection (DC-) and a midpoint connection (N), and at least three switching sections (3a, 3b, 3c) connected to the intermediate circuit (2), the switching sections each comprising a phase tap (u, v, w) for electrical connection to a respective phase (Phu, Phv, Phw) of an electric drive machine, in, Each switching section (3a, 3b, 3c) has three switching devices (S1, S2, S3) for modulating an operating point-specific output voltage (Va) 1 、Va 2 ) can be switched into at least 27 switching states and the phase taps (u, v, w) of the corresponding switching sections (3a, 3b, 3c) are connected to each of the terminals (DC+, DC-, N) of the intermediate circuit (2) while generating voltage vectors (v1p, v1n, v2p, v2n, v3, v4, v5), Wherein, for the method, in order to compensate for the output voltage (Va) specified in the modulation working point 1 、Va 2 ), an undesired, switching state-specific change in the midpoint voltage of the intermediate circuit (2) when - determining an operating point-specific switching pattern (SM1, SM2) from at least four voltage vectors (v1p, v1n, v2p, v2n, v3, v4), at least two of which (v1p, v1n, v2p, v2n) are generated redundantly but with opposite effects on the midpoint voltage by different switching states of the three-level inverter (1), - predicting the current load (i) of the midpoint connection (N) of the intermediate circuit (2) which would cause an undesired midpoint voltage change based on the determined switching pattern (SM1, SM2); B ), - Based on the predicted current load (i B ) determines the compensation of the current load (i B ) of the compensation current (i A ),and - Determine the time weighting of the redundant voltage vectors (v1p, v1n, v2p, v2n) by which the output voltage (Va) is modulated 1 、Va 2 ) provides compensation current (i A ) at least a portion of.
2. The method according to claim 1, It is characterized in that The current load (i B ) is predicted with the aid of the duration of a non-redundant voltage vector within the switching mode (SM1, SM2), the non-redundant voltage vector being generated by a switching state in which a midpoint switch having the switching device (S3) connected to the midpoint terminal (N) of the intermediate circuit (2) participates.
3. The method according to claim 1 or 2, It is characterized in that The current load (i B ) is predicted using the currents which are supplied to the phases (Phu, Phv, Phw) according to the switching patterns (SM1, SM2).
4. The method according to claim 2 or 3, It is characterized in that The current load (i B ) is additionally determined with the aid of a measured value of an already existing deviation of the mid-point voltage at the beginning of the respective modulation period.
5. The method according to any one of the preceding claims, It is characterized in that At the current load (i B ) cannot be fully compensated by the time weighting of the redundant voltage vectors (v1p, v1n, v2p, v2n), the non-redundant voltage vectors (v1p, v1n, v2p, v2n) causing the midpoint voltage change, which are generated by a switching state in which a midpoint switch of the switching device (S3) connected to the midpoint terminal (N) of the intermediate circuit participates, are at least partially replaced by two non-redundant voltage vectors (v3, v5) generated by switching states in which the midpoint switch is not involved.
6. The method according to claim 5, It is characterized in that In the first modulation cycle half, one of the two non-redundant voltage vectors (v3, v5) generated by the switching state without the participation of the midpoint switch and the redundant voltage vector (v1p, v2p) is provided, and in the subsequent second modulation cycle half, the other of the two non-redundant voltage vectors (v3, v5) generated without the participation of the midpoint switch and the redundant voltage vector (v1n, v2n) is provided. 7 . A control unit for a power electronics module of a motor vehicle, the control unit being designed to carry out the method according to one of the preceding claims.
8. A power electronic module for a motor vehicle, the power electronic module include: - A three-level inverter (1), comprising an intermediate circuit (2) with three connections in the form of a high-side connection (DC+), a low-side connection (DC-) and a midpoint connection (N), and at least three switching sections (3a, 3b, 3c) connected to the intermediate circuit (2), each of the switching sections comprising a phase tap (u, v, w) for electrical connection to a respective phase (Phu, Phv, Phw) of an electric drive machine, wherein each switching section (3a, 3b, 3c) comprises three switching devices (S1, S2, S3) for modulating an operating point-specific output voltage (Va) 1 、Va 2 ) can be switched into at least 27 switching states and in this case the phase taps (u, v, w) of the corresponding switching sections (3a, 3b, 3c) are connected to each of the terminals (DC+, DC-, N) of the intermediate circuit (2) while generating a voltage vector (v1p, v1n, v2p, v2n, v3, v4, v5); and - A control device according to claim 7.
9. A motor vehicle comprising an electrical energy storage device, an electric drive machine and a power electronics module according to claim 8, in, The three-level inverter (1) is connected at the input to the electrical energy storage device and at the output to the phases (Phu, Phv, Phw) of the electric drive machine.