DE102015221926A1 - A method of controlling a semiconductor switch of a power converter, method and apparatus for operating an electrical machine - Google Patents

Abstract

Disclosed is a method for controlling a semiconductor switch (HS1) of a power converter (WR) with a pulse width modulated control signal (SS1), wherein the semiconductor switch (HS1) during a signal period (T) of the control signal (SS1) in two period sections (T1, T2) of Signal period (T), each with one of two different partial duty cycles (D11, D12) is switched activated. Furthermore, a method and a device for operating an electrical machine (EM) by means of at least one semiconductor switch (HS1) are provided, wherein the at least one semiconductor switch (HS1) is controlled by at least one pulse width modulated control signal (SS1) according to the above-mentioned method.

Description

Technical area:

The invention relates to a method for controlling a semiconductor switch with a pulse width modulated control signal. Furthermore, the invention relates to a method and a device for operating an electrical machine with at least one semiconductor switch. Moreover, the invention relates to an electric drive arrangement with a device mentioned above.

State of the art:

Half-bridge converters, in particular inverters, are used to convert an input current of one current type or at a frequency into an output current of a different current type or with a different frequency. Esp. For example, inverters are used to operate electric vehicles of hybrid electric / electric vehicles, which convert direct currents into alternating currents (or phase currents) for the electric machines and / or vice versa. For this purpose, the inverter bridge circuits with semiconductor switches, which are controlled by pulse width modulated control signals alternately switched on / off in a switching frequency and thereby convert the DC currents in the AC / phase currents with a corresponding fundamental frequency.

The object of the present invention is to generate a higher output voltage at the output current of a semiconductor switch of a power converter, in particular an inverter, while maintaining the switching frequency of the semiconductor switches.

Description of the invention:

This object is solved by subject matters of the independent claims. Advantageous embodiments are the subject of the dependent claims.

According to a first aspect of the invention, a method is provided for controlling a semiconductor switch of a power converter, in particular a power output stage of an inverter, with a pulse-width-modulated control signal. According to the method, the semiconductor switch is switched during a signal period of the control signal in two period sections of the one signal period, each with one of two different partial duty cycles.

The invention is based on the following findings:

The conversion of an input current of one type of current or with a frequency by means of a power converter in an output current of another type or with a different frequency is usually done by turning on / off a semiconductor switch a half-bridge of the power converter, eg. An inverter, with a pulse width modulated control signal with a high switching frequency. As a result, the output current (DC or DC or AC / phase voltage or AC / phase current) is generated with a corresponding fundamental frequency. The semiconductor switch is switched on or off at a switching frequency in such a way that a desired voltage curve can be simulated as well as possible at the output current with a requested fundamental frequency. As a result, a higher amplitude of the output voltage of the output current of the semiconductor switch and thus more power to be retrieved from the output current of the semiconductor switch.

In order to map the requested fundamental frequency at the output current, a switching frequency with a frequency value that is at least ten times as high as the required fundamental frequency is generally required, with which the semiconductor switch also has to be switched. The smaller the distance between the fundamental frequency and the switching frequency, the worse the said fundamental frequency can be mapped. This results in increased harmonics in the output current and in the output voltage, respectively.

In order to set the switching frequency ten times as high as the fundamental frequency, the output current must also be mapped with ten duty cycles per fundamental frequency period. Accordingly, the semiconductor switch must be turned on / off at a switching frequency which is ten times the fundamental frequency period. The high switching frequency in turn leads to high switching losses in the semiconductor switch.

In the case of an inverter for providing phase voltages or currents for an electric machine, for example, a hybrid electric / electric vehicle, the increased harmonics in the output current (phase current) or in the output voltage (phase voltage) have negative effects on the power output stage of the inverter and the electrical machine result.

To reduce the harmonics, the switching frequency of the semiconductor switch can be increased. By increasing the switching frequency, however, the (maximum possible) effective amplitude of the output voltage (phase voltage) is reduced due to the dead time in the switching operation of the semiconductor switch. Due to a lower effective amplitude of the output voltage (phase voltage) can in turn from the output current ( Phase current) of the semiconductor switch less power to be retrieved for the electric machine.

Based on the above-mentioned findings, a solution is sought with which a semiconductor switch, in particular a semiconductor switch of a power output stage of a power converter, especially an inverter for operating an electrical machine of an electric drive is controlled such that at the same switching frequency, a higher voltage utilization ( and thus a higher power utilization) is made possible and the output current of the semiconductor switch also has even lower harmonics.

The semiconductor switches are switched on / off in such a way that the voltage curves of the output current to be generated (for example, a phase current) can be simulated as well as possible. This is to a higher amplitude of the output voltage (or the phase voltage) of the semiconductor switch and thus more power from the output current (the phase current) of the semiconductor switch can be accessed.

In order to achieve the initially mentioned object, the above-mentioned method for controlling a semiconductor switch of a power converter, in particular a power output stage of an inverter for operating an electrical machine, is provided with a pulse-width-modulated control signal in the present application. In the method, the control signal is divided into two period sections during one signal period, one of two different partial duty cycles being used in each of the two period sections. This applies in particular to several or all signal periods of the control signal. The semiconductor switch is then switched under control of one of the two different partial duty cycles during each of the one signal period in each of the two period sections.

In this case, a partial duty cycle means a duty cycle that applies in one of the two period sections of a signal period of the control signal, wherein the semiconductor switch is switched controlled in this period with this partial duty cycle. Thus, in the signal period mentioned, the control signal has two different duty cycle values (namely two partial duty cycles), whereby the semiconductor switch is switched to two different duty cycle values in the said two period sections of the same signal period.

The two partial duty cycles are designed such that with these two Tastgradwerten compared to an undivided, for the entire signal period single duty cycle, which corresponds to the total duty cycle of the corresponding signal period, low harmonics in the output current (ie in the output voltage) result. The total duty cycle corresponds to a desired duty cycle, with which the voltage curves of the output current of the semiconductor switch to be generated can be well simulated.

In this case, the two period sections of one signal period can be of different lengths or have different durations of time. For example. For example, a first of the two periodic portions may be 40% of the total duration of the signal period and a second of the two periodic portions corresponding to 60% of the total duration of the signal period.

Preferably, however, the two period sections of the one signal period are the same length or have the same time duration.

The high level sections of the two period sections of the one signal period may be at the two period ends of the corresponding signal period, so that the low level sections of the two period sections are in the middle of the corresponding signal period and are connected. This is the case, in particular, in the case of a low-side semiconductor switch of a half-bridge (a bridge circuit), since the drive signal for a high-side semiconductor switch of the half-bridge is generally inverted for driving the corresponding low-side circuit located in the same half-bridge. Semiconductor switch is supplied to this.

However, the high-level sections of the two period sections of the one signal period are preferably in the middle of the corresponding signal period, so that these high-level sections are located in the middle of the corresponding signal period and are connected. This is esp. In the case of the aforementioned high-side semiconductor switch of a half-bridge of the case, which is usually switched to the non-inverted drive signal.

By the method described above, the duty cycle of the control signal is almost already changed / adjusted during the respective signal periods of the control signal. As a result, the output current of the semiconductor switch is reproduced the desired output current more accurately, without increasing the switching frequency of the control signal. This in turn allows a higher (maximum possible) effective amplitude (in the fundamental) of the output voltage. By a higher amplitude of the output voltage more power can be retrieved from the output current of the semiconductor switch.

This advantage is all the stronger, the smaller the difference between the switching frequency of the control signal and the required fundamental frequency of the output current of the semiconductor switch (or the required fundamental frequency of the actual rotational speed of the electric machine).

In comparison to the control signal with high-level mid-frequency signal sections, in which only one duty cycle value is mapped per signal or switching period, two duty cycle values can be mapped per signal or switching period per signal or switching period generated in accordance with the method described above become. As a result, the control signal can be adapted more quickly to the output current to be output without having to increase the (switching) frequency of the control signal.

As a result, the switching frequency of the control signal can be kept comparatively low and consequently the semiconductor switch can also be operated with less switching losses. If the semiconductor switch is used as a switch of a power output stage of an inverter for operating an electric machine of an electric drive, then the drive can be operated correspondingly lower power loss.

This provides a possibility for generating more power at the output current of a semiconductor switch of a power converter, in particular an inverter, while maintaining the switching frequency of the semiconductor switch. In addition, the semiconductor switch can be operated with lower power dissipation.

D1

Gesamt-Tastgrad;

D11

erster Tastgrad;

D12

zweiter Tastgrad;

T

Zeitdauer der Signalperiode;

T1

Zeitdauer eines ersten Periodenabschnitts; und

T2

Zeitdauer eines ersten Periodenabschnitts.

Preferably, the sum of the two partial duty cycles, weighted by the signal duration of the respective period sections, of the two period sections of the one signal period corresponds to a total duty cycle of the corresponding signal period. Thus: D1 = (T1 / T) * D11 + (T2 / T) * D12, where are:

D1

Total duty cycle;

D11

first duty cycle;

D12

second duty cycle;

T

Duration of the signal period;

T1

Duration of a first period section; and

T2

Duration of a first period section.

Preferably, the sum of the two time periods of the two period sections corresponds to one signal period of the entire time period of said signal period.

According to a further aspect of the invention, a method for operating an electrical machine, in particular a permanent / third-excited synchronous machine or an asynchronous machine, is provided with at least one semiconductor switch. According to the method, the at least one semiconductor switch is controlled with at least one pulse width modulated control signal according to a method described above.

The two partial duty cycles are preferably calculated from at least one measured actual angle value (angle measurement value) of a rotor of the electric machine.

Furthermore, the two partial duty cycles are preferably calculated from at least one measured actual rotational speed (rotational speed measured value) of the rotor.

A first partial duty cycle is preferably calculated based on a first reference angle value, which is a sum of a first actual angle value measured at a first measurement time and a rotational angle of the rotor traveled at a first actual rotational speed measured at the first measurement time in a first time period equivalent. In this case, the first time duration corresponds to a time duration between the first measurement time and a first reference time.

The first reference time is a point in time at which the first reference angle value (as described above) is calculated / predicted based on the first actual angle value and the first actual rotational speed. The first reference time is in the first period, preferably between 0% to 50% of the signal period, especially between one sixteenth to seven sixteenths of the signal period, between one eighth to three eighths of the signal period, or between three sixteenths to five sixteenths of the signal period, specifically at one quarter of the signal period.

The first measurement time is preferably before the current signal period. The first measurement time is preferably as close as possible to the current signal period.

Similarly, a second partial duty cycle is preferably calculated based on a second reference angle value corresponding to a sum of the first actual angular value and a rotational angle of the rotor covered by the first actual rotational speed in a second time duration. In this case, the second period of time is a period of time between first measurement time and a second reference time.

The second reference time is a timing at which the second reference angle value (as described above) is calculated / predicted based on the first actual angle value and the first actual rotation speed. The second reference time is in the second period, preferably between 50% to 100% of the signal period, especially between sixteenths to fifteen sixteenths of the signal period, between five-eighths to seven-eighths of the signal period, or between eleven sixteenths to thirteen sixteenths of the signal period, specifically at three quarters of the signal period.

Alternatively, the second partial duty cycle is preferably calculated on the basis of a third reference angle value which is a sum of a second actual angular value measured at a second measurement time and a rotational angle of the second measured at a second measurement instant in a third time period Rotor corresponds. In this case, the third time duration corresponds to a time duration between the second measurement time and the second reference time.

Preferably, the second measurement time is temporally before the second period section. In this case, the second measuring time is preferably after the first measuring time and as close as possible to the second period portion.

The two partial duty cycles for the two period sections of the signal period are thus each calculated from one of the two reference angle values, which are respectively calculated / predicted for the respective reference times in the respective period sections.

Using the above-mentioned reference angle values, a temporal angle correction is performed, whereby the measured actual angle values, which no longer correspond to the current angular positions of the rotor due to the further rotation of the rotor at the respective reference times, are corrected to these current angular positions.

The temporal angle correction is advantageous because the hardware, such. As the microprocessor, to determine the Teiltastgrade and the corresponding signal transmission between the components of the hardware corresponding computing or transmission time need to provide the corresponding partial duty cycles, during which the angular position of the rotor (by the further rotation) changes.

For the calculation of the two partial duty cycles, the angular positions of the rotor at respective reference times in the middle of the respective period sections are advantageous. (The respective middle may also differ).

In addition to the actual angle values and the actual rotational speeds, a rotational angular acceleration of the rotor can also be used for angle correction. As a rule, however, the rotational angular acceleration has only a slight influence on the result of the angle correction and can therefore be neglected.

Due to the early measurement of the actual angle values and the actual rotational speeds at the first and the second measuring time, the two partial duty cycles are determined in good time before the beginning of the signal period or before the start of the corresponding period sections and thus stand at the beginning of the signal period or already available before the beginning of the corresponding period.

In accordance with yet another aspect of the invention, a computer-readable data store is provided by storing program codes that perform at least one of the previously described methods.

According to yet another aspect of the invention, an apparatus for operating an electrical machine, in particular a permanent / foreign-excited synchronous machine, or an asynchronous machine is provided.

The device comprises at least one semiconductor switch, in particular a bridge circuit, for passing through or providing at least one phase current for the electrical machine.

Furthermore, the device comprises at least one control arrangement for controlling the at least one semiconductor switch with at least one pulse-width-modulated control signal. In this case, the control arrangement via a control signal output to the control terminal of the semiconductor switch is electrically connected and adapted to switch the semiconductor switch during a signal period of the control signal in two period sections of the corresponding signal period with one of two different partial duty cycles.

According to yet another aspect of the invention, an electric drive arrangement, in particular for driving a hybrid electric / electric vehicle, is provided.

The electric drive arrangement comprises at least one electrical machine, in particular a permanent / externally excited synchronous machine or an asynchronous machine, for driving the hybrid electric / electric vehicle. The electric drive arrangement further comprises at least one previously described device for providing at least one phase current for the at least one electrical machine. The at least one device comprises at least one phase-current line and is electrically connected via the at least one phase-current line to at least one stator phase of the at least one electrical machine.

In this case, the load current path of the at least one semiconductor switch (that is to say the drain-source path in the case of a power MOSFET or, in the case of a power IGBT, the collector-emitter path) is preferably electrically connected to the stator phase of the electrical machine via the at least one phase current line connected.

Advantageous embodiments of the two methods described above are, as far as applicable to the above-mentioned device or to the above-mentioned electric drive arrangement, also to be regarded as advantageous embodiments of the device or the electric drive assembly.

Brief description of the drawing:

In the following, exemplary embodiments of the invention will be explained in more detail with reference to the accompanying drawings. Showing:

1 in a schematic representation of an electric drive assembly with a device according to an embodiment of the invention;

2A . 2 B in respective signal flow diagrams in each case one signal period of a pulse width modulated control signal according to the prior art and a signal period of a pulse width modulated control signal according to a method according to an embodiment of the invention.

Detailed description of the drawing:

1 shows a schematic representation of an electric drive assembly EA for driving a hybrid electric vehicle.

The electric drive assembly EA comprises an electric machine EM for driving the hybrid electric vehicle, which in this embodiment is designed as a permanent-magnet synchronous machine having a stator with three stator phases and a rotor rotatable relative to the stator.

The electric drive assembly EA further comprises a device VR for operating the electric machine EM.

The device VR in turn comprises an inverter WR (inverter) for providing phase currents to the electric machine EM, a circuit arrangement SA for controlling the inverter WR, a phase current measuring unit PM for measuring actual phase current values I_ist1, I_ist2 on phase-phase conductors PL of the inverter to be described below WR, a rotational angle measuring unit WM for measuring actual angular values w_ist1, w_ist2 and actual rotational speeds a_ist1, a_ist2 of the rotor, and a data storage unit DS for storing / providing desired phase current values I_soll1, I_soll2 to be described below.

Alternatively, the desired phase current values I_soll1, I_soll2 can also be calculated in a manner known to the person skilled in the art for the current signal period T. In this case, then eliminates the data storage unit DS.

The inverter WR comprises a triple bridge circuit BS as a power output stage for providing phase currents for the electric machine EM.

The bridge circuit BS comprises three half-bridges HB, which are connected in parallel between a positive supply current line V + and a negative supply current line V-. Each of the three half-bridges HB comprises in each case a positive-voltage-side semiconductor switch HS1 and in each case a negative-voltage-side semiconductor switch HS2, the two semiconductor switches HS1, HS2 of the respective half-bridges HB being connected to one another in series. The semiconductor switches HS1, HS2 are designed, for example, as power IGBTs or as power MOSFETs. All semiconductor switches HS1, HS2 are each equipped with a freewheeling diode.

The bridge circuit BS further comprises three phase current lines PL, which electrically conductively connect respective electrical connection points VP of the positive voltage side and negative voltage side semiconductor switches HS1, HS2 of the respective half bridges HB to one of the three stator phases of the electric machine EM.

The control arrangement SA comprises a calculation unit BE and a control unit SE.

The calculation unit BE is electrically connected on the signal input side to a signal output of the phase current measurement unit PM and receives from the phase current measurement unit PM currently measured actual phase current values I_ist1, I_ist2, which measures the phase current measuring unit PM on the respective phase current lines PL. The actual phase current values I_ist1, I_ist2 are periodically synchronized with the switching frequency of the control signals SS1, SS1 'to be described below; SS2, SS2 '; SS3, SS3 'measured.

The calculation unit BE is also electrically connected on the signal input side to a signal output of the rotational angle measuring unit WM and receives the currently measured actual angular values w_ist1, w_ist2 and the currently measured actual rotational speeds a_act1, a_act2 of the rotor of the electric machine EM from the rotational angle measuring unit WM. The actual angle values w_ist1, w_ist2 and the actual rotational speeds a_ist1, a_ist2 are periodically synchronized with the switching frequency of the control signals SS1, SS1 'to be described below; SS2, SS2 '; SS3, SS3 'measured.

The calculation unit BE is on the signal input side further electrically connected to a signal output of the data storage unit DS and receives from the data storage unit DS the desired phase current values I_soll1, I_soll2 as reference current values, which are stored in advance calculated or iteratively recalculated during operation of the electric machine EM.

Based on the actual and target phase current values I_ist1, I_ist2; I_soll1, I_soll2, the actual angle values w_ist1, w_ist2 and the actual rotational angular velocities a_ist1, a_ist2 calculates the calculation unit BE for individual signal periods of the three control signals SS1, SS1 '; SS2, SS2 '; SS3, SS3 'two partial duty cycles D11, D12; D21, D22; D31, D32 passes the calculated partial duty cycles D11, D12; D21, D22; D31, D32 continue to the downstream control unit SE.

The control unit SE generates based on the transmitted partial duty D11, D12; D21, D22; D31, D32 six control signals SS1, SS1 '; SS2, SS2 '; SS3, SS3 'and controls the six semiconductor switches HS1, HS2 of the three half-bridges HB with the corresponding control signals SS1, SS1'; SS2, SS2 '; SS3, SS3 'in a manner known to those skilled in the art, so that the positive voltage side and the negative voltage side semiconductor switches HS1, HS2 are alternately turned on / off, which are generated from the flowing between the positive and negative supply power line V +, V- DC phase currents, which be supplied via the three phase current lines PL the corresponding stator of the electric machine EM.

The calculation of the partial duty cycles D11, D12; D21, D22; D31, D32 by the calculation unit BE will be described below 2A and 2 B as well as in connection with 1 described in more detail.

2A is a schematic signal flow diagram, a signal period T of a generated according to the prior art pulse width modulated control signal SS0 mapped.

In this control signal SS0, the signal portion having a high level is symmetrically located in the middle of the corresponding signal period T, so that the duration of the signal portion having the high level (D01 * T1 or D02 * T2) in the two equal period portions T1 , T2 (T1 = T2 = 1/2 * T) of the same signal period T are the same length and the two partial duty cycles D01, D02 in the two period sections T1, T2 and the total duty D0 for the entire signal period T equal are: D01 = D02 = D0; and D01 · T1 = D02 · T2 = D0 · T / 2.

2 B is a further schematic signal flow diagram, an analog signal period T of a control signal SS1 with the two determined according to an embodiment of the invention partial duty D11, D12 shown.

In this control signal SS1, the signal portion having the high level is asymmetrical to the middle of the signal period T, so that the time duration of the signal portion having the high level in the two equally long period portions T1, T2 (T1 = T2 = 1/2 × T) are of different lengths and thus the partial duty cycles D11, D12 in the two period sections T1, T2 and the total duty D1 for the entire signal period T are different in size: D11 ≠ D12 ≠ D1; and D11 × T1 × D12 × T2 × D1 × T / 2, but where: D11 * T1 + D12 * T2 = D1 * T.

The above equations continue to hold, even if the two period sections T1, T2 of the signal period T are of different lengths: T1 ≠ T2 ≠ 1/2 · T.

This applies analogously to the two further control signals SS2, SS3, which for the two, equally long period sections T1, T2 of the same signal period T different partial duty cycles D21, D22; D31, D32 whose respective average values correspond to the respective total duty cycles D2; D3 of the corresponding signal period T of the respective control signals SS2, SS3 correspond to: D21 ≠ D22 ≠ D2; D21 · T1 · D22 · T2 · D2 · T / 2; and D21 * T1 + D22 * T2 = D2 * T; respectively. D31 ≠ D32 ≠ D3; D31 × T1 × D32 × T2 × D3 × T / 2; and D31 × T1 + D32 × T2 = D3 × T.

The semiconductor switches HS1, HS2 are then controlled by control signals SS1, SS2, SS3 (or inverse control signals SS1 ', SS2', SS3 ') in a manner known to those skilled in the art, offset in time or in opposite directions, their respective total duty cycles D1, D2 or D3 the sums of the respective first and second partial duty factors D11, D12; D21, D22; D31, D32, multiplied by the durations of the signal period T and the respective period sections T1, T2, correspond to: D1 = D11 * (T1 / T) + D12 * (T2 / T); D2 = D21 * (T1 / T) + D22 * (T2 / T); D3 = D31 * (T1 / T) + D32 * (T2 / T).

The determination of the two partial duty cycles D11, D12 of the control signal SS1 takes place as described below. The determination of the respective partial duty cycles D21, D22; D31, D32 of the further control signals SS2, SS2 takes place in an analogous manner:

At a first measurement time tm1, which lies temporally before the beginning of the current signal period T0, the phase current measuring unit PM measures a first actual phase current value I_ist1 and forwards it to the calculation unit BE. At the same time of measurement tm1, the rotational angle measuring unit WM measures or determines a first actual angular value w_ist1 and a first actual rotational speed a_ist1 and forwards the two measured values to the calculation unit BE. The data storage unit DS forwards (at the same time) a first desired phase current value I_soll1 to the calculation unit BE, which has been stored in the data storage unit DS in a pre-calculated manner.

The calculation unit BE first calculates the first two from the first actual phase current value I_ist1 and the first set phase current value I_soll1 in a manner known to the person skilled in the art (for example by means of so-called closed-loop current control (CLS)) Voltage components Ud1 and Uq1 of a d- / q-coordinate system of vector control of the electric machine EM.

From the first actual angle value w_ist1 and the first actual rotational speed a_ist1, the calculation unit BE further calculates a first reference angle value wb1.

In this case, the first reference angle value wb1 is an angular position of the rotor predicted for a first reference time tb1. In this case, the first reference time tb1 lies in the middle of the first period section T1 of the current signal period T, for which the two partial key grades D11, D12 are determined.

The first reference angular value wb1 is calculated as a sum of the first actual angular value w_ist1 and a rotational angle of the rotor covered by the first actual rotational speed a_ist1 in a time duration t1 between the first measuring time tm1 and the first reference time tb1.

The change in the first actual rotational speed a_ist1 in the past time period t1 is neglected because it is negligible and has little effect on the change of the angular position in the time period t1. Thus: wb1 = w_ist1 + t1 * a_ist1 = w_ist1 + (tb1-tm1) * a_ist1.

As a result of the first reference time tb1 being in the middle of the first period T1, the calculated first reference angle value wb1 has an overall smallest deviation from the actual angular position of the rotor at the said first reference time tb1. If necessary, the first reference time tb1 may also lie in the front or in the back half of the first period T1.

From the two first voltage components Ud1 and Uq1 and the first reference angle value wb1, the calculation unit BU calculates three first phase voltages for the three phase currents and from the first three in a manner known to the person skilled in the art (for example based on a "d- / q transformation") Phase voltages the first partial duty D11 and also the first partial duty D21, D31 of the other control signals SS2, SS3. The positive voltage side and the negative voltage side semiconductor switches HS1, HS2 are then controlled for the first period portion T1 of the current signal period T with the respective first partial duty D11, D21, D31 in the manner known to those skilled temporally offset or in opposite directions.

At a second measurement time tm2, which lies ahead of the second period section T2 or before the start t0 of the current signal period T, the phase current measuring unit PM measures a second actual phase current value I_ist2 and forwards it to the calculation unit BE. To the same Measuring time tm2 measures or determines the rotational angle measuring unit WM a second actual angle value w_ist2 and a second actual rotational speed a_ist2 and forwards the two measured values to the calculation unit BE. The data storage unit DS forwards (at the same time) a second desired phase current value I_soll2 to the calculation unit BE, which has also been stored in the data storage unit DS in a pre-calculated manner.

The calculation unit BE first calculates the second two from the second actual phase current value I_act2 and the second desired phase current value I_soll2 in the manner known to the person skilled in the art (for example by means of a so-called closed current control ("closed-loop current control (CLS)") Voltage components Ud2 and Uq2 of the d / q coordinate system of the vector electric machine control EM.

From the second actual angular value w_ist2 and the second actual rotational speed a_ist2, the calculation unit BE calculates a second reference angle value wb3.

In this case, the second reference angle value wb3 is an angular position of the rotor predicted at a second reference time tb2.

In this case, the second reference time tb2 is in the middle of the second period section T2 of the current signal period T.

The second reference angular value wb3 is calculated as the sum of the second actual angular value w_ist2 and a rotational angle of the rotor covered by the second actual rotational speed a_act2 in a time duration t3 between the second measuring time tm2 and the second reference time tb2.

The change of the second actual rotational speed a_act2 in the past time period t3 is also neglected for the reason stated above. Thus: wb3 = w_act2 + t3 * a_act2 = w_act2 + (tb2-tm2) * a_act2.

Characterized in that the second reference time tb2 is in the middle of the second period portion T2, the calculated second reference angular value wb3 has a total smallest deviation from the actual angular position of the rotor to said second reference time tb2 tb2. If necessary, the second reference time tb2 may also lie in the front or in the back half of the second period section T2.

From the two second voltage components Ud2 and Uq2 and the second reference angle value wb3, the calculation unit BU calculates three second phase voltages for the three phase currents and the three second voltages in the manner known to the person skilled in the art (for example based on a "d- / q transformation") Phase voltages the second partial duty cycle D12 and also the second partial duty cycles D22, D32 of the further control signals SS2, SS3. The positive voltage side and the negative voltage side semiconductor switches HS1, HS2 are then switched to the second period section T2 of the current signal period T driven with the respective second partial duty D12, D22, D32 further.

Alternatively, the three second partial duty cycles D12, D22, D32 may be from the first actual phase current value I_ist1 measured at the first measurement time tm1, the first target phase current value I_soll1, the first actual angle value w_ist1, and the first actual rotational speed a_ist1 in an analogous manner calculated as described above.

In this case, the second reference angle value wb2 corresponds to a sum of the first actual angular value w_ist1 and a rotational angle of the rotor covered by the first actual rotational speed a_ist1 in a time duration t2 between the first measuring time tm1 and the second reference time tb2: wb2 = w_ist1 + t2 * a_ist1 = w_ist1 + (tb2-tm1) * a_ist1.

The three second phase voltages for the three phase currents and thus the three second partial duty cycles D12, D22, D32 are then calculated by the calculation unit BE from the two first voltage components Ud1 and Uq1 and the second reference angle value wb2 calculated as described above.

By the method described above, the total duty cycles D1, D2, D3 of the control signals SS1, SS1 '; SS2, SS2 '; SS3, SS3 'already in the middle of the respective signal periods T adapted. As a result, the phase voltages provided by the semiconductor switches HS1, HS2 can be adapted more quickly to those desired setpoint voltages, without the (switching) frequency of the control signals SS1, SS1 '; SS2, SS2 '; SS3, SS3 'to increase.

As a result, the switching frequency of the control signals SS1, SS1 '; SS2, SS2 '; SS3, SS3 'are kept comparatively low and consequently the semiconductor switches HS1, HS2 can also be operated with less switching losses. Accordingly, the electric drive EA can also be operated with less power loss.

Claims (14)

 Method for controlling a semiconductor switch (HS1) of a power converter (WR) with a pulse width modulated control signal (SS1), wherein the semiconductor switch (HS1) during a signal period (T) of the control signal (SS1) in two period sections (T1, T2) of the signal period (T ) is switched in each case with one of two different partial duty cycles (D11, D12). Method according to Claim 1, wherein the sum of the two partial duty cycles (D11, D12) of the two period sections (T1, T2) of the signal period (T) weighted by the signal duration of the respective period sections (T1, T2) corresponds to a total duty cycle (D1 ) corresponds to the signal period (T), where D1 = (T1 / T) * D11 + (T2 / T) * D12. Method according to claim 1 or 2, wherein the sum of the two time periods of the two period sections (T1, T2) corresponds to the total duration of the signal period (T). Method for operating an electrical machine (EM), with at least one semiconductor switch (HS1), esp. A bridge circuit (BS), for passing at least one phase current to the electric machine (EM), wherein the at least one semiconductor switch (HS1) with at least one pulse width modulated Control signal (SS1) is controlled according to a method according to any one of the preceding claims. Method according to Claim 4, wherein the two partial duty cycles (D11, D12) are calculated from at least one measured actual angular value (w_ist1, w_ist2) of a rotor of the electric machine (EM). Method according to claim 4 or 5, wherein the two partial duty cycles (D11, D12) are calculated from at least one measured actual rotational speed (a_ist1, a_ist2) of the rotor. The method of claim 6, wherein a first of the two partial duty cycles (D11) is calculated based on a first reference angle value (wb1), wherein the first reference angle value (wb1) is a sum of a first actual angle value measured at a first measurement time (tm1) (w_ist1) and one of a first, at the first measurement time (tm1) measured actual rotational speed (a_ist1) in a first time period (t1) rotational angle of the rotor corresponds, wherein the first time period (t1) between the first measuring time (tm1) and a first reference time (tb1), wherein the first reference time (tb1) in a first period section (T1), preferably between 0% to 50% of the signal period (T), in particular between one sixteenth to seven sixteenths of the signal period (T) , between one-eighth to three-eighths of the signal period (T), or between three sixteenths to five-sixteenths of the signal period (T), specifically at one quarter of the signal period (T). The method of claim 7, wherein the first measurement time (tm1) is prior to the signal period (T). A method according to claim 7 or 8, wherein a second one of the two partial duty cycles (D12) is calculated based on a second reference angular value (wb2), the second reference angular value (wb2) being a sum of the first actual angular value (w_ist1) and one with the the first actual rotational speed (a_ist1) corresponds to the rotational angle of the rotor covered in a second time duration (t2), the second time duration (t2) lying between the first measuring time point (tm1) and a second reference time point (tb2), the second reference time point (tb2) in a second period portion (T2), preferably between 50% to 100% of the signal period (T), in particular between nine sixteenths to fifteen sixteenths of the signal period (T), between five eighths to seven eighths of the signal period (T), or between eleven Sixteenth to thirteen sixteenths of the signal period (T), specifically at three quarters of the signal period (T). A method according to claim 7 or 8, wherein the second partial duty factor (D12) is calculated based on a third reference angle value (wb3), wherein the third reference angular value (wb3) is a sum of a second actual angular value measured at a second measurement time point (tm2) (w_ist2) and one with a second, measured at the second measurement time point (tm2) actual rotational speed (a_ist2) in a third time period (t3) rotational angle of the rotor corresponds, wherein the third time duration (t3) between the second measurement time (tm2) and the second reference time (tb2), wherein the second reference time (tb2) in the second period section (T2), preferably between 50% to 100% of the signal period (T), esp. between nine sixteenths to fifteen sixteenths of the signal period (T) , between five eighths to seven eighths of the signal period (T), or between eleven sixteenths to thirteen sixteenths of the signal period (T), specifically at three quarters of the signal period (T), l ying. The method of claim 10, wherein the second measurement time (tm2) is prior to the second period period (T2). Computer-readable data storage in which program codes are stored, with which a method according to one of the preceding claims can be carried out. Device (VR) for operating an electrical machine (EM), comprising: - At least one semiconductor switch (HS1), esp. A bridge circuit (BS), for passing at least one phase current for the electric machine (EM); - At least one control arrangement (SA) for controlling the at least one semiconductor switch (HS1) having at least one pulse width modulated control signal (SS1), which is arranged, the semiconductor switch (HS1) during a signal period (T) of the at least one control signal (SS1) in two period sections (T1, T2) of a signal period (T) with one of two different partial duty cycles (D11, D12) controlled to switch. Electric drive arrangement (EA), in particular for driving a hybrid electric / electric vehicle, comprising: - At least one electrical machine (EM); - At least one device (VR) according to claim 13, wherein the at least one device (VR) at least one phase current line (PL) and via the at least one phase current line (PL) with at least one stator phase of the at least one electric machine (EM) is electrically connected ,

Images